THE DEVELOPMENT OF CONVECTIVE INSTABILITY DURING SESAME,
1979
by
Frank Parker Colby, Jr.
B.S.,
M.S.,
University of Michigan
(1976)
Massa-chusetts
Institute of Technology
(1979)
Submitted to the Department of
Meteorology and Physical Oceanography
in Partial Fulfillment of the
Requirements of the Degree of
DOCTORATE OF PHILOSOPHY
at
the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
February,
c
Mas-s-a=chuset.ts
1983
Institute of Technology,
1983
Signature of Aut.ho-r
Department
of Meteorology and Physical
Oceanography. 6 January 1983
Certified by
Snders,
Frede
Accepted
Thesis Supervisor
by
Ronald G.
Prinn, Chairman, Departmental
Committee on Graduate Students
Undgreh
MASL
STITUTE
MAR 22 1983
I IRAARIR
PAGE 2
DEVELOPMENT OF CONVECTIVE INSTABILITY DURING SESAME, 1979
by
FRANK PARKER COLBY, JR.
Submitted to the Department of
Meteorology and Physical Oceanography
on 25 January 1979 in partial fulfillment of the
requirements for the Degree of Doctorate in Philosophy
ABSTRACT
Convection in t-he Central United States is assumed to require
the presence of convective instability and a triggering
mechanism to release the instability. Often, a stable layer
caps the PBL, preventing or delaying the release of convective
instability. The development of both convective instability
and convective inhibition (from the stable layer) is studied
with data from three cases from the SESAME field project of
1979. The cases are: 19-20 April, 9-10 May. 6-7 June. The data
are analyzed and both convective instability and inhibition
are quantified.
A one-dimensional thermodynamic model which includes
radiation, a surface energy balance, and routines to predict
soil and boundary layer characteristics is used as a tool to
understand the imp-ortant physical factors involved in the
development o-f convective instability and inhibition.
The results show that some convective instability was present
before dawn in all three cases. The boundary layer heating
during the day added to the initial instability. Soil
moisture, clouds, and changes in atmospheric structure above
the PBL were all important Pactors controlling the PBL
evolution.
The modelled convective instability grew during the day as a
result oF the boundary layer heating. Increased soil moisture
sometimes exerted a positive influence on the growth of
instability, but in other cases limited the growth by keeping
the PBL from heating and deepening. Clouds gene-rally reduced.
the convective instability
growth, but in the June case,
clouds had the opposite effect. The influence of changes above
the PBL was stronger on the reduction o- convective inhibition
than on the growtn of convective instability. For these cases,
the inPluence on the growth of convective instability from
changes above the PBL was stronger than the presence of clouds
or the increase of soil
moisture, but all
o
these Factors
were able to modify the development of convective instability
and inhibition.
I.
PAGE 3
The results of the modelling and the observations show that
the convection occurred where and when the inhibition was
reduced to low values. The convection began when the available
forcing was sufficient to overcome the remaining inhibition.
Therefore, the forecasting of convective outbreaks requires
the ability to measure and predict the convective inhibition
within the larger region of cbnvective instability.
Thesis Supervisor:
Title:
Dr. Frederick Sanders
Professor of Meteorology
PAGE 4
TABLE OF CONTENTS
ITEM
Abstract
PAGE
...........................................
Table of Contents ...............................
List of Figures
.................................
L ist of Tables ........
1.
Introduc-tion
2.
Model
............
.............
15
................................
Introduction .............
Atmospheric Structure...
Conceptual Model Run ....
Radiation ................
A.
Incident Radiation .
IR Emission .........
B.
2.5 Radiation Data Comparison
Surface Energy Balance ..
2.6
A.
Sensible and Latent Hea
B.
Soil Heat Flux .....
2.7
Ground Variables ........
2. - Boundary Layer Variables
2.9 PBL Temperature and Moist ure
2.10 Initialization Procedure
2.11 Sensitivity Tests .......
A.
TBAR Test ..........
B.
VS Test ............
C.
WMAX and GWB Test ..
D.
GWO Test ...........
2.12 Comparison Runs .........
2.13 Model Comparison Summary
Tables for Chapter 2 .........
Figures for Chapter 2 ........
2.1
2.2
2.3
2.4
...
•
...
..
3.
19 April
Case St-udy I:
3.1
Introduction ......
Synoptic Analysis .
3.2
Mesoscale Analysis
3.3
3.4. Soundings .........
3.5 Hybrid Modelling ..
Summary ...........
3.6
lables for Chapter 3 ...
Figures for Chapter 3
14
.
°
•
es
...
•
...
•
27
28
29
30
30
32
34
36
36
38
38
40
42
42
44
46
46
48
50
51
54
57
61
74
74
75
76
80
98
104
111
PAGE 5
4.
5.
6.
Case Study II:
9 May
4. 1
Introduction ........
4.2
Synoptic Analisis.
......
4. 3
Mesoscale Analysis ...........
4.4
Soundings .......
4.5 21 GMT Modelling ...
........
4. 6
Summary for 21 GMT Modelling
4. 7
23 GMT Modelling ............
4.8 Summary for 23 GMT Modelling
Tables for Chapter 4 .............
Figures for Chapter 4 ............
145
145
146
151
155
165
168
172
177
183
Case Study III:
6 June
5. 1
Introduction ......
5.2
Synoptic Analysis
5. 3
Mesoscale Analysis
5.4
Soundings .........
5. 5
Hybrid Modelling ..
5. 6
Summary ...........
Tables for Chapter 5 ...
Figures for Chapter 5
225
225
226
230
237
243
247
250
Conclusions
........... ..
280
of
of
of
of
of
290
295
298
300
304
Appendices
7. 1
Derivation
7.2
Derivation
7.3 Derivation
7.4
Derivation
7.5
Derivation
References
Radiation Parameterization ...
Ekman Layer Similarity Equations
Soil Heat Flux Parameterization
Ground Variable Equations ....
Inversion Equations ..........
..................
............
Acknowledgements ............................
.....
305
.....
309
PAGE 6
LIST OF FIGURES
2. 1
Q.2
2.3
Schematic diagram of model atmosphere.
61
.Comparison of model net radiation and soil fluxes with
observations and modelling by Wetzei (1978) for O'Neill
day number 2.
62
Comparison of model IR cooling rates with
of Rogers and Walshaw (1966).
calculations
62
2. 4
Comparison of model IR cooling rates with calculations
of Brooks. (1950), Elsasser (1942), and ECMWF (1979). 63
2.5
Model IR cooling rates
coverages.
2.6
Schematic diagram illustrating process of finding new
PBL potential temperature from new inversion
65
calculations.
2. 7
Schematic diagram showing process of
A~
from initial sounding.
2.8
2. 9
2. 10
2. 11
2. 12
only for cloud
layer of variable
64
initialization of
66
Sounding plotted on a pseudoadiabatic diagram from
O'Neill, Nebraska at 1200 GMT.
Time variation
model run.
of pressure level
Net radiation and sensible heat
model run.
Time variation
run.
Latent
run.
heat and
of PBL top
67
for standard
67
fluxes for standard
68
of PBL mixing ratio for
standard model
68
soil
heat fluxes for standard model
69
2. 13
Time variation of ground temperature and
temperature for standard model run.
2.14
Comparison of model sensible heat flux with
observations and modelling by Wetzel (1978)
day number 2.
heat
surface layer
69
flux with
2. 15
Comparison of model latent
for O'Neill day number 2.
2.16
Comparison of model output and observed
temperature for O'Neill sounding.
for O'Neill
70
observations
70
ground
71
PAGE 7
2. 17
2. 18
2. 19
2.20
2.21
Comparison of.model output and observed
temperature for O'Neill sounding.
surface layer
Comparison of model output and
for O'Neill sounding.
growth of PBL
observed
71
72
Sounding plotted on a pseudoadiabatic diagram from,
initial sounding from Barnard (1977).
72
Comparison of Barnard's (1977) model output with
present model output for PBL moisture at 0700 LST.
73
Comparison of Barnard's (1977) model output with
present model output for PBL moisture at 1000 LST.
73
3. 1
Map of SESAME region showing sounding stations in April,
and two surface observation stations mentioned later in
the text.
111
3.2
500 mb analysis for
3.3
500 mb
3.4
Synoptic-scale analysis for 12 GMT,
19 April.
114
3.5
Synoptic-scale analysis for 00 GMT, 20 April.
115
3.6
Photo.of low-elevation angle radar screen display at
Garden City, Kansas, 2102 GMT, 19 April.
116
Photo of low-elevation angle radar screen display at
Garden City, Kansas, 2122 GMT, 19 April.
116
3.7
19 April.
112
analysis-for 00 GMT, 20 April.
113
12 GMT,
3.8
Photo of low-elevation angle radar screen display at
117
Garden City, Kansas, 2140 GMT, 19 April.
3.9
Photo of low-elevation angle radar screen display at
Garden City, Kansas, 2200 GMT, 19 April.
117
3. 10
Mesoscale surface analysis
for 12 GMT,
19 April.
118
3. 11
Mesoscale surface analysis
for 16 GMT,
19 April.
119
3. 12
Visible satellite
3. 13
Mesoscale surface analysis for 21 GMT,
19 April.
121
3. 14
Mesoscale surface analysis for 22 GMT, 19 April.
122
3. 15
Sounding plotted on a pseudoadiabatic diagram from
Dodge City, Kansas for 1115 GMI, 19 April.
123
3. 16
Sounding plotted
photo for
1601
OMT,
19 April.
on a pseudoadiabatic diagram from
120
PAGE 8
Dodge City,
3. 17
Kansas for
1415 GMT,
19 April.
123
Sounding plotted on a pseudoadiabatic diagram from
Dodge City, Kansas for 1715 GMT, 19 April.
124
Sounding plotted on a pseudoadiabatic diagram from
Dodge City, Kansas for 2015 GMT, 19 April.
124
Sounding plotted on a pseudoadiabatic diagram from
Goodland, Kansas for 1124 GMT, 19 April.
125
Sounding plotted on a pseudoadiabatic diagram from
Goodlandi, Kansas for 2007 GMT, 19 April.
125
Sounding plotted on a pseudoadiabatic diagram from
model output for 20 GMT from GLD initial sounding.
126
Sounding plotted on a pseudoadiabatic diagram from
Concordia, Kansas for 1108 GMT, 19 April.
126
Sounding plotted om a pseudoadiabatic diagram from
Concordia, Kansas for 2008 GMT, 19 April.
127
Sounding plotted on a pseudoadiabatic diagram from
model output for 20 GMT from CNK initial sounding.
127
3. 25
Change in surface potential
GMT to 22 GMT, 19 April.
12
128
3.26
Sounding plotted on a pseudoadiabatic diagram for
initial h.gbrid sounding at 11 GMT, 19 April.
129
3.27
Map
130
3.28
Time variation of model output for HYB sounding, 19
April, 50-50 soil parameters with no extra factors.
131
Time variation of model output for HYB sounding,
April, 50-50 soil parameters with morning clouds
imposed.
19
132
Time variation of model output For CNK sounding,
April, 70-80 soil parameters with no clouds.
19
3. 18
3. 19
3. 20
3. 21
3. 22
3. 23
3. 24
3. 29
3. 30
3. 31
3. 32
3. 33
of rainfall
in Kansas on
temperature and dewpoint
18 April.
133
Time variation of model output for CNK sounding, 19
April, 70-80 soil parameters with "all day" clouds.
134
Portion of sounding data from Dodge Citu, Kansas
plotted on a pseudoadiabatic diagram for ii GMT and
GMT, 19 April.
14
135
Time -variation of model output for HYB sounding, 19
April, 50-50 soil parameters with inversion changes
PAGE 9
imposed.
136
3.34
Time variation of model output for. HYB sounding, 19
April, 50-50. soil parameters with DDC and inversion
13"7
changes imposed.
3.35-
Time vari'ation oF model output for HYB sounding, 19
April, 50-50 soil parameters with GLD and inversion
138
changes imposed.
3.36
Sounding plotted on a pseudoadiabatic diagram for model
output for 21 GMT from HYB initial sounding with GLD
139
and inversion changes imposed
3.37
Time variation of model output for HYB sounding, 19
April, 50-50 soil parameters with DDC and inversion
.140
changes and morning clouds imposed.
3.38
Time variation of model output for HYB sounding, 19
April, 50-50 soil parameters with GLD and inversion
141
changes and morninrg clouds imposed.
3.39
Sounding plotted on a pseudoadiabatic diagram for model
output for 21 GMT from HYB initial sounding with GLD
142
and inversion changes and morning clouds imposed.
3.40
Sounding plotted on a pseudoadiabatic diagram for model
output for 21 GMT from HYB initial sounding with
modified GLD- and inversion changes and morning clouds
142
imposed.
3.41
Mesoscale analysis of convective instability and
convective inhibition at 17 GMT, with 21 GMT radar
echoes superimposed.
143
Mesoscale analysis of convective instability and
convective inhibition at 20 GMT, with 22 GMT radar
echoes superimposed.
144
3.42
9 May to
12
183
4. 1
Severe weather events during period 12 GMT,
10 May 1979.
GMT,
4.2
Synoptic-scale
500 mb analysis for
12 GMT, 9 May.
184
4. 3
Synoptic-scale
500 mb analysis for
00 GMT,
10 May.
185
4. 4
Synoptic-scale
surface
12 GMT,
9 May.
186
4. 5
Synoptic-scale surface analysis for 00 GMT,
10 May.
187
4.6
Locations
4.7
Mesoscale 500 mb analysis -For
of
radiosonde
analysis for
launch- sites for May.
188
G h T,
189
11
9 May.
PAGE 10
4.8
Mesoscale 500 mb analysis
4.9
Change in temperature
12 to 20 GMT, 9 Mau.
190
for 20 GMT, 9 May.
and mixing
4.10
Change in temperature and mixing
20 to 23 GMT, 9 May.
4. 11
Mesoscale 700 mb analysis for
4. 12
4.13
ratio at 500 mb
from
191
500 mb
Prom
192
ratio at
11 GMT, 9 May.
193
Mesoscale 700 mb analysis
for 20 GMt, 9 May.
194
Change in temperature and
12 to 20 GMT, 9 May.
mixing ratio at 700 mb from
195
4.14
Change in temperature and mixing ratio at 700 mb from
20 to 23 GMT, 9 May.
196
4. 15
Mesoscale
surface analysis for
12 GMT) 9 May.
19 7
4. 16
Mesoscale surface analysis for
18 GMT,. 9 May.
19 8
4.17
Mesoscale surface analysis for 21 GMT, 9 May.
4. 18
Change in potential temperture and
21 GM-T,. 9 .May.
4. 19
Mesoscale surface analysis for 23 GMT, 9 May.
4.20
Photograph of lowu-elevation angle display from radar
screen at -Amarillo, Texas at 2242, 2247, and 2254 GMT,
9 May.
20 2
4.21
Sounding plotted on a pseudoadiabatic
Shamrockr Texas for 1143 GMT, 9 May.
diagram from
Sounding plotted on a pseudoadiabatic
Shamrock,. Texas for 1705 GMT, 9 May.
diagram from
Sounding plotted on a pseudoadiabatic
Amarillo, Texas for 2300 GMT, 9 May.
diagram from
Sounding plotted
Childress, Texas
diagram from
4.22
4.23
4.24
19'9
dewpoint from 12 to
20 0
on a pseudoadiabatic
for 2006 GMT, 9 May.
20)oi
2C)4
2C 5
2C 6
20)7
4.25
Depth of nearly dry adiabatic
K isentropes, 11 GMT, 9 May.
layer between 319 and 31 -7
2C 8
4.26
Depth of nearly dry adiabatic
K isentropes, 20 GMT, 9 May.
layer between 319 and 31 7
2C.9
4.27
Sounding plotted on a pseudoadiabatic diagram from
Oklahoma City., Oklahoma for 2000 GMT, 9 May.
21 0
PAGE 11
4.28
Sounding plotted on a pseudoadiabatic diagram from
MAYHYB hybrid sounding for 1100 GMT, 9 May.
211
4.29
Time. variation of model output for MAYHYB initial
sounding, 9 May, 5-70 soil parameters, with no extra
factors modelled (P).
212
4.30
Time variation of model output for MAYHYB initial
sounding, 9 May, 5-70 soil parameters, with morning
-clouds imposed (C).
213
4.31
Time va-riation of model output for MAYHYB initial
sounding, 9 May, 5-70 soil parameters, with imposed
changes (H).
214
4.32
Time variation of model output for MAYHYB initial
sounding, 9 May, 5-70 soil parameters, with both
morning clouds and imposed changes (HC).
215
4.33
Sounding plotted on a pseudoadiabatic diagram from
model output at 2L00 GMT, 9 May from MAYHYB initial
sounding, 5-70 soil parameters, and no extra factors
modelled (P).
216
4.34
Sounding plotted on a pseudoadiabatic diagram from
model output at 2100 GMT, 9 May from MAYHYB initial
sounding, 5-70 soil parameters, with both clouds and
imposed changes (HC).
217
4.35
Comparison between 5-70 HC model run and
observatiGns ta-ken from analyses.
4.36
4.37
surface
218
Sounding--plotted on a pseudoadiabatic diagram from
Amarillo, Texas at 1700 GMT, 9 May.
219
Mesoscale analysis of convective instability and
convective inhibition at 17 GMT with 21 GMT radar
echoes superimposed.
220
4.38
Time variation of model output for MAYHYB initial
sounding, 9 May, 5-70 soil parameters, with clouds,
imposed changes, and surface moisture advection (GCH).
221
4.39
Sounding plotted on a pseudoadiabatic diagram from
model output at 2300 GMT, 9 May from MAYHYB initial
sounding, 5-70 soil parameters, with clouds, imposed
222
changes, and surface moisture advection (OCH).
4.40
Sounding plotted on a pseudoadiabatic diagram from
model output at 2300 GMT, 9 May from MAYHYB initial
sounding, 5-70 soil parameters, with clouds, imposed
changes, surface moisture advection, and modified for
PAGE 12
surface temperature advection
4.41
223
(OCH modified).
Mesoscale analysis .of convective instability and
convective inhibition at 20 GMT with 23 GMT radar
echoes superimposed.
224
5.1
Synoptic-scale 500 mb analysis for
12 GMT, 6 June.
250
5.2
Synoptic-scale 500 mb analysis for 00 GMT, 7 June.
251
5.3
Synoptic-scale surface analysis for
12 GMT, 6 June.
252
5.4
Synoptic-scale surface analysis for 00 GMT, 7 June.
253
5.5
Sounding network
254
5.6
Mesoscale 500 mb analysis for
15 GMT, 6 June.
255
5.7
Mesoscale 500 mb analysis for
18 GMT, 6 June.
255
5.8
Mesoscale 700 mb analysis for
15 GMT, 6 June.
256
5.9
Mesoscale 700 mb analysis for
18 GMT, 6 June.
256
for June 6-7 Case..
5. 10
Mesoscale
surface analysis for 12 GMT, 6 June.
257
5.11
Mesoscale
surface analysis for
15 GMT, 6 June.
258
5. 12
Mesoscale s-u-rface analysis for
18 GMT, 6 June.
259
5. 13
Mesosca-l--e-surfac-e analysis for
19 GMT, 6 June.
260
5. 14
Change of potential temperature and
and 19 GMT, 6 June.
5. 15
Sounding plotted on a pseudoadiabatic diagram fror
Oklahoma City-,
Oklahoma for 12 GMT, 6 June.
dewpoint betweeen
12
261
262
Sounding plotted on a pseudoadiabatic diagram
Hennesse,
Oklahoma for 1312 GMT, 6 June.
from
Sounding plotted on a pseudoadiabatic diagram
Elmore City, Oklahoma for 15 GMT, 6 June.
from
Sounding plotted on a pseudoadiabatic diagram
Elmore City, Oklahoma for 18 GMT
6 June.
from
5. 19
Sounding plotted on a pseudoadiabatic diagram
Sill, Oklahoma for 15 GMT, 6 June.
from Fort
266
5.20
Sounding plotted on a pseudoadiabatic
Sill, Oklahoma for 1
GMT, 6 June.
5. 16
5. 17
5. 18
263
264
265
diagram from Fort
267
PAGE
5.21
13
Sounding plotted on a pseudoadiabatic diagram from
Clinton Sherman AFB, Oklahoma for 18 GMT, 6 June.
268
Sounding plotted on a pseudoadiabati
diagram from
Wichita Falls, Texas for 17 GMT, 6 June.
269
Sounding plotted on a psuedoadiabatic diagram for.
JUNHY3, 12 GMT, 6 June.
270
5.24
Rainfall
271
5.25
Time variation of model output for JUNHYB sounding, 6
June, 30-60 soil parameters, with no clouds or imposed
changes aloft (plain).
272
5.26
Time variation of model output for JUNHYB sounding, 6
June, 30-60 soil parameters, with clouds.
273
5.27
Time variation of model output for JUNHYB sounding,
June, 30-60 soil parameters, with imposed changes
aloft.
5.22
5. 23
for Oklahoma
for 5 June in
inches.
6
274
5.28
Time variation of model output for JUNHYB sounding, 6
June, 30-60 soil parameters, with clouds and imposed
changes aloft.
275
5. 29
Comparison -of TS and CS values from 30-60 model run
with clouds and imposed changes aloft with values taken
from Elmore City, Fort Sill, and Chickasha, Oklahoma
276
sound ing s.
5.30
Sounding plotted on a pseudoadiabatic diagram for model
output at 15 GMT from JUNHYB initial sounding, 30-60
soil parameters, with clouds and imposed changes aloft.
277
5. 31
Sounding plotted on a pseudoadiabatic diagram for model
output at 18 GMT from JUNHYB initial sounding, 30-60
soil parameters, with clouds and imposed changes aloft.
278
5.32
Sounding plotted on a pseudoadiabatic diagram for model
output at 19 GMT from JUNHYB initial sounding, 30-60
soil parameters, with clouds and imposed changes aloft.
278
5.33
Mesoscale analysis of convective instability and
convective inhibition at 18 GMT with 19 GMI radar
echoes superimposed.
280
PAGE
14
List of Tables
2.1
Schedule of
2.2
Model
results for sensitivitu
tests.
2.3
Model
results for sensitivity
tests
3. 1
Advection
3.2
Model results at 21 GMT, 19 April.
3.3
Clouds
3.4
Imposed changes from Dodge City, Kansas including
inversion changes, for April 19.
107
Imposed changes from Goodland, Kansas
inversion changes, for April 19.
108
3.5
model
calculations.
calculation
imposed
57
for April
59
(continued).
case.
104
105
in model runs for April
results at 21 GMT,
60
case.
106
including
3.6
Model
19 April.
3.7
Sensitivity
4. 1
Model
4.2
Clouds
4.3
Imposed changes on model
4.4
Sensitiviztq values for 21 GMT, 9 May model runs.
179
4. 5
Model results at 23 GMT, 9 May.
180
4. 6
Model results at 23 GMT, 9 May,
4.7
Sensitivity values for 23 GMT, 9 May model runs.
5. 1
Clouds
5.2
Imposed
5. 3
Model
5. 4
Sensitivity
5. 5
Negative area for 30-60 model run with
and clouds in June case.
values for 21
GMT,
109
19 April
model
runs.
results at 21 GMT, 9 May.
imp-os-ed in model
imposed
177
runs for May
case.
runs for May
178
case.
178
continued.
181
in model runs for June case.
182
247
247
changes on model runs for June case.
results at
110
19 GMT, 6 June.
248
values for 23 GMT, 6 June model
runs.
imposed
249
changes
249
PAGE 15
INTRODUCTION
The term "convection"
•distinguish
flows.
As
such,
it includes a very
contains both
defined and will
overturning
broad class
convection will
in the-atmosphere.
The intent
be much more
is to include
only
severe weather in the central
section of the United.States.
Although most severe weather is
to buoyant overturning, the association
Carbone
of atmospheric
be used to denote only buoyant
convection associated with
due
to
buoyant and non-buoyant motion.
purpose of this thesis,
narrowly
in meteorology
overturning motion in the atmosphere from laminar
motion , which
For the
is used
is not strict.
(1982) reported on a case of nearly neutral
stratification which produced heavy rain and tornadoes.
Convecti-on in
the- central
to require the...ol-lowing:
stra.tif-ication and
an
U. S.
in the spring
un-stable thermal vertical
initiating
(trigger) mechanism.
expected sequence is approximately as follows:
scale or mesoscale motion and/or heating and
mechanisms, a
unstable.
boundary
layer
atmosphere.
The
Through
large
cooling
section of the atmosphere becomes convectively
This means that air
positively
is assumed
(PBL),
if
parcels from the planetary
lifted
sufficiently, would
become
buoyant, and would continue to rise in the
This requires abundant moisture, as the latent
heat of condensation released
in the parcel
is needed to
PAGE 16
(The atmospheric
maintain the parcel's buouancy.
is almost always more stable than
over any appreciable layer.)
the PBL which
rest of
separates the potentiallybuoyant air
the atmosphere.
central U.S.
In the
buoyant air
potentially
where it will actually
is
sequence of events
lifted
"through"
that a
part o-f the PBL
from the
this stable
inversion.
by a temperature
cooling
incorrect.
Actually, no
layer
this expected
parcels are
What really happens
is destabilized
of the inversion.
by
is
heating and/or
This destabilization
spat-ial scale, much
convection "breaking through"
the stable
In a sense,
layer.
observational -etwork.
differences
through
be buoyant.
the stable
occurs on a small
of the
above
(trigger mechanisms) must act to
forcing
external dynamic
adiabatic
lapse rate
for convection to begin, a combination of heating
Therefore,
lift the
adiabatic
exists a stable layer
There
is very often characterized
layer
and
the dry
lapse rate
Hence,
smaller than the scale
the appearance of the
the stable layer
is
due to
in s-cales.
Initiation of this type of convection has been studied
varying emphases.
Early
for many
years with
focussed
on the problems of forecasting this
Convection is
small
presents very real
Work
scale in both
space and
work
type of weather.
time, which
forecasting and observational
by Fawbush et al. (195i)
scale patterns which were
generally
diffi-culties.
sought to characterize
favorable to
large
convective outbreaks.
PAGE 17
Their aids
and
included vertical
wind shear, low level
temperature
moisture advection, and mid-level vorticity patterns.
Darkow et al.(1958) found a convincing statistical
between a. particular surface
occurrence of tornadoes.
often occurred
temperature pattern and the
They showed
that severe weather
to the east and near the axis of a
tongue-shaped warm region.
vertical
correlation
Other work sought to look at
stratification, with the development of various
indices to quantify
instability
or potential
the atmosphere.
Thus, the Showalter Index
the Lifted
(Galway, 1956),
Index
(Miller, 1972) were spawned as
an-d- Moore
in
(Showalter, 1953),
and the Total Totals Index
forecast aides.
analysis continues to the present, with
et al. (198)
instability
This type of
such work as Carlson
(unpublished) in which new ways of
quantifying vertical stratification have
been developed which
try to take into account more details in
the structure.
instance,
simply compares the temperature
the-Shoimalter Index
of air at 850 .rb when
lifted adiabatically to 500 mb,
ambient temperature at 500 mb.
an index called Lid Strength
Index
inversion which
with
(LSI) which
includes in
a measure of the
caps the PBL.
This index
would show a difference between places which have equal
amounts of parcel b-uoyancy
resistance to initiating
the
Carlson et al. (1980) developed
addition to the Shawalter type of buoyancy,
strength of the
For
but have differing amounts of
the overturning
motion.
PAGE 18
Much work has
mechanisms
focussed
on determining what triggering
are operating i.n various situations.
The result
has been the
identification of several mechanisms,
ageostrophic
frontogenetical
triggering,
convergence,
sea breeze convergence,
gravity
including
wave
and outflow boundary
convergence.
As described
last of three
groups of convection on 8 June 1974 seemed
forced
by
by Koch and McCarthy
frontogenetical motion along
a cold front.
instance the convergence and deformation along
pre-existing
front provided
(1982)
the
to be
In this
the
the vertical motion to set off the
convection.
Work
(1930)
by Uccellini
implicated
thunderstorms.
Miller and
tornadoes
(1975) and
gravity wave
Uccellini
Sander-s !o-k-ed
on 3 :April
later by Miller and
forcing
studied
at
1974.
for severe
data from 18 May
Both
these packets with
or
waves in
their
Koch
the
study
and McCarthy
of the 8 June
gravity wave
radar
initiate
they moved across the
also implicated
second
1974 case.
packets triggered
large
They
the packets seemed to
(1982)
formation of the
a
tornado occurrence and
enhance convective activity as
country.
the
showed that
of
identified wave packets which
section of the east and central United States.
echo activity, and
1971 while
the so-called super-outbreak
seemed to have coherent surface signatures through
correlated
Sanders
gravity
convective outbreak
in
In all of these examples
convergence
in
the PBL
PAGE 19
which
provided
the lifting
required
Sea breeze circulations
in
triggering and
Florida.
have been shown to
Cooper et al. (1932)
convection.
sea breeze
showed clearlu
type circulations
can be
Oklahoma on 8 June
line
1966.
in the vicinity
emerged along
breeze.
This
A very
strong
sea breeze" was a
in
examined
several
lifting
cases
They
Sun and Ogura
the same way
as a sea
factor in the
to trigger more
Maddox
tornadic thunderstorms
in the vicinity
theorized
of thermal
that convergence and cyclonic
the boundaries, although
could not show this clearly.
Matthews
where the outflow from a large
cloud
embedded.
formation
Cooper
in which
et al. (198
current
et al. (1980)
vorticity were enhanced along
arc
from
temperature gradient
in the manner of a density
in which
intensified when they were
boundaries.
places where
this case.
lighter air up.
pushing the
late
of Norman, Oklahoma, and a convergence
Outflow- from other storms can act
bodily
peninsula
type circulation using data
triggering of the convection
convection, through
only
important.
this gradient in much
"inland
that
in-
sea breeze triggered
Coastlines are not the
(1979) modelled a sea breeze
formed
be important
organizing convection, particularly
scale convergence due to the daily
morning
to trigger the convection.
(1981)
convective
they
discussed a case
storm triggered an
small convective cells were
),
in
their
examination
of
PAGE 20
Florida convection,
convergence
cell
on a
growth.
showed
that outflow contributed
storm scale which was responsible
A particular
(1980).
They
showed that a
tornado formed along an
intersection
cells,
on the order of 10**-3 1/s.
convergence
of outflows from old
In comparison to the work on dynamic
creation of the
convective
instability
attention
in the published
literature.
generally
quoted or referred
(warm low level,
(BLH).
Modahl
Of
cold
upper
to are
level)
the two,- more work
(1979) studied
from 1972 -
1974,
for new
case from 1973 in Florida was
analyzed by Holle and Maier
with
to
forcing, the
has received
much
The mechanisms
differential advection
and boundary
layer heating
on advection has been done.
National Hail Research Experiment data
for the occurrence of hail.
two factors were most important:
He showed
increased southerly
increase moist-u-re and heat in the PBL, and easterly
sustain the storms
Carlson e
amount
of potential
above).
For
instability
Scoggins
(1981)
Variability Experiment
examine
the creation of
fields
used
winds to
winds to
give upslope triggering).
capping
inversion and
(as discussed with
10 April
their
that
to correlate convective
strength of the
the case of
correlation between
Davis and
(and perhaps to
al. (1980) attempted
outbreaks with the
less
1979, they
of LSI and
the
the LSI
found good
the convection.
data from Atmospheric
IV (AVE IV,
convective
24-25 April
instability,
1975)
wind
to
shear,
PAGE 21
and vertical motion.
They
considered
convective
instability except BLH.
discovered
that all
motions)
contributed
surface -
850 mb
residual,
could
low level
and
This
1979),
but some
residual,
included
the
important
(as above
in
the Big Thompson Storm in
moisture was crucial
to the
lifting.
convection over an
that the mountain was a large
storms sustenance,
Raymond and Wilkening
isolated mountain, and
source of
heat which
Although their
case was
speculated that an adequate moisture supply would
certainly have led
mountain.
1977.
to thunderstorm formation
Cotton et al. (1982)
thunderstorm which
level
part of
in the
BLH as a factor too.
helped drive an upslope circulation.
July
especially
in mountains suggested that
al. (1979) discussed
low level
they
large scale
implies that BLH,
of convection
(1980) examined dry
dry,
(eg.
they
have been the dominant term.
upslope winds provided the
found
sources
than the
sources of
Interestingly,
advection of moisture was
Caracena et
which
less
layer.
Some studies
Modahl,
of the other
all
formed
They concluded
over the
studied a quasi-steady
in the mountains
in Colorado on
that a combination of BLH,
moisture advection, and
upslope winds
19
low
initiated the
convection.
These
studies of convection in mountains have
be more complete
than
those over
the plains,
seemed
to
although the BLH
PAGE 22
contribution was
still
not
clearly
delineated.
Ogura et al.(1982) analyzed a convective
1.979 in
the
central United States.
that BLH, inland sea breeze
instability
A study
system on 9-10 May
Their results
suggested
circulation and perhaps symmetric
were responsible for the convection, a
of thermodynamic and
by
dynamic
combination
factors
The above discussion shows that initiation of convection
by
dynamic
forcing
occurs
in many waus,
the same place, within a few hours
It
(Koch and McCarthy,
is apparent that the thermodynamic
convection se-quence has not been well
speculate as
complex
to the reasons for the
large
soil
spatial variation
Additionally,
as soil
some
lack of work
between BLH and
in the physical
constants
The approach used
(such as
The problem is
standpoint.
observed,
such
moisture.
this thesis
the
in this area:
layer characteristics, plus
from the theoretical
temperature and
One can
surface
of the factors are not well
The intent of
study
quantified.
composition, albedo, vegetation, etc.).
certainly difficult
1982).
contribution to the
interactions between radiation,
characteristics, and boundary
sometimes many ways in
development
is a combined
is to quantify
of
the relationship
convective instability.
observational and modelling
of convective outbreaks occurring
in
three case studies
from the Severe Environmental Storms and Mesoscale Experiment
PAGE 23
(SESAME).
The cases of 19-20 April;
9-10 May, and 6 June 1979
are examined to determine the role of BLH in the creation of
convective
instability.
instability will
The PBL air
For this purpose,
be measured by
convective
the PBL lifted
of the PBL..
adiabatically to 500 mb, and
its temperature
500 mb
When
temperature.
The parcel
instability
is.lifted
is compared with
lifted,
will be warmer than the ambient atmosphere.
positive for
(PLI).
is taken as a parcel with the mean potential
temperature and mixing ratio
the ambient
index
unstable air
The PLI
unlike the operational
is
Lifted Index
(Galway, -1956).
The PLI
(SLI,
is- not the same thing as
from Sanders, personal
the surface lifted
communication)
are defined by the surface observations.
in which
index
parcels
The PLI measures a
mean PBL convective instability, above any superadiabatic
layers which may
PLI
be present.
For a clear,
can be as much as 4 degrees C
Sill,
Oklahoma sounding at
although generally
lower than the SLI
1800 GMT, 6 June
the difference
is responsible for the
modellers generally
question:
convection?
which
On
which
initiate
the other
is difficult to envision that parcels can rise
parcel
Cloud
large perturbations to
convection, even with unstable soundings.
it
1979, Fig.5.20),
convective instability
initiation of
need very
(Fort
is closer to 2 degrees C.
This difference raises an interesting
really measures the cloud-scale
well-mixed PBL, the
hand,
from the
PAGE 24
surface
layer through the entire turbulent PBL without
dilution.
Observations at
the top
show such large perturbations.
that parcels with
(mean PBL and
I suspect that the answer
characteristics
surface layer),
of the clear PBL do not
in between the two
surface layer
this question,
measure convective
determine how much
The
Without attempting
in
stable layer is studied as well
inhibition exists
inhibition is
is the region between the
the ambient sounding
The energy
done by
equal
to
this area
T(This
quantified by
must be surpassed
convective
instability.
unit mass,
is
prior to the
This negative
on the chart, and
is negatively
found by
force
to
integrating the
diagram..
parcel's path
can
.
well which
this negative area)
immediately
while the parcel
the negative buoyancy
S
to settle
this thesis.
"negative area" on a pseudoadiabatic
area
being the result of
use the mean PBL parcel definition to
instability
The role of the
convection.
parcels.
I shall
extremes
are more representative of the
cloud-scale convective instability,
diluted
buoyant.
calculating
the work
per unit mass, =
be regarded as an energy
for the parcel
to realize its
If a PBL parcel were to rise through
it must have sufficient
or an updraft velocity equal
to
kinetic
energy per
(2*Negative
area)**1/2.
The model
is used
is
to
identify and
quantify
the role
PAGE 25
various
physical paramenters have
one dimensional,
and
is designed to. model
solar radiation on BLH.
divided
into
soil
on the BLH.
heat,
The model
is
only effects of
Incident radiation on the surface is
sensible heat and
latent heat
fluxes.
The effect of soil moisture plays a prominent, role in this
part.the
The. fluxes into
top of the
the PBL drive
turbulent entrainment at
(assumed) well-mixed PBL , and the temperature
height and moisture
the PBL, atmospheric
sounding data and
content of the PBL are computed.
changes are
imposed
inferred
from observed
on the model.
The term PBL will be used extensively
of this thesis.
atmosphere.
part
As
Above
It is not a well-defined
used in the
of the atmosp-hee which
following,
throughout the rest
term for the
it will refer to the
obtains most of
its
characteristics. rom its proximity to the ground.
Physically,
this will meant-hat part of the atmosphere which is heated
during the day by
adiabatic
the fluxes
from the ground surface.
chart,- the- well-mixed PBL will be
dry adiabatic
mixing ratio
In each
nearby
identifiable by a
temperature lapse rate and a nearly constant
(q)
in the air nearest the ground.
of the three cases studied, the
to convection are analysed, and the
pinpointed.
On an
The model
location of the outbreak
is then run using
soundings and varying
conditions prior
initial
data from
the important parameters.
PAGE 26
Various
combinations of physical factors are applied and
effects on convective
physical
instability are determined.
changes in the atmosphere above te PBL.
factors are
tried separately to determine
importance.
Then combinations are tried,
including all
of the relevant factors.
runs can reproduce
moisture content,
the sense that the
the observed
it will
individual
with the
last runs
one of these last
be regarded as a "correct" run
included effects
is described
If
in detail in
is a summary of the
The
surface temperature and PBL
studies follow in chapters 3,4,and
chapter 6,
The
factors which are varied are ground wetness, presence
of clouds, and
The model
the
in
are modeled correctly.
chapter 2.
5.
The
findings.
The three case
last chapter,
PAGE 27
THE MODEL
2.1:
Introduction
One of the tools -used in this research
one-dimensional boundary layer model.
used
to determine and
heating
The model
illustrate the role of -boundary layer
without various physical
importance
determine
The model
is.also used to
the vertical structure of the atmosphere at times
sounding data
between stations and/or between sounding
SESAME case stud-u days- were characterized
soundings on a
often sig.nificantly
Because of this,
statements about
by 3 hourly.
in some way
(time or space or
it was not possible to make
the vertical
The convection primarily
of the BLH took
itself had a
Although
smaller, the convection often
structure of the atmosphere
some kind of supplementary
after much
times).
(i.e.
rid spacing no larger than the normal synoptic
broke out "between" soundings
without
is run
instability and/or
locations which did not have real
both).
It
effects to quantify their
in generating convective
removing convective inhibition.
grid and
is primarily
(BLH) in the time evolution of soundings.
with and
and
is a
information.
studied
place.
in this thesis occurred
This means that the PBL
simple structure, and the
fluxes at the top and
PAGE 28
bottom of the PBL could
functions.
Although
present by
with
strong horizontal
the outbreak time,
throughout most of
that the
be represented
initial
the
thermal
initial
conditions at any
The model
then used
layer is not of crucial
of the atmosphere.
is characterized by
G
constant q)
its height
and
inversion which
),
temperature
all of the
This is
( A&
the
surface layer
(well-mixed in
(well-mixed in moisture =
The PBL is capped by
) and a
depth.
taken from the input
inversion is held
enough to absorb
to exist.
structure, but the
in meters.
has a strength
is
in heat and moisture.
The PBL. above the
its moisture
inversion depth
of the
be specified
importance in affecting
its potential
heat = constant
top
so
these initial
is assumed
an ad hoc representation of the real
initial
to be small,
Structure
surface la-uer of five mb depth
stability
gradients
the model run.
Atmospheric
surface
thermal
changes due to BLH for
The PBL is assumed to be well-mixed
A
gradients were
point could
conditions and predicted the thermal
2. 2:
known structure
the SESAME area, tended
some certainty.
the rest of
by
The
sounding.
The
constant until the PBL.grows
inversion layer.
The bottom of
the
inversion rises as the PBL grows by entraining air from
the
inversion
is
an
layer into the PBL.
taken directly from the data
The
structure above the PBL
in the original
sounding.
The
PAGE 29
assumed
2. 3:
structure
is illustrated
Conceptual Model
Detailed
in Fig.
1.
Run
later are the various parameterizations used
effect changes in
outlined here.
the initial atmospheric
Table 2.1
operations in the
structure.
to
They are
contains a schedule of the
order actually
used by
Radiative transfer is computed
the computer.
first.
The
incoming
radiation at the top of the atmosphere is a function of time
of day and
geographic
water vapor, COV
scattered,
emit
some
Radiation is absorbed
location.
and liquid water
is reflected.
(clouds).
Some
The atmosphere and
infrared radiation, and the net radiation
for each
laye-r in the atmosphere and
is
the ground
is calculated
the ground
surface. .The
net absorbed radiation in the
ground is partitioned
parts:
sensible heat flux
1)
soil
heat flux, 2)
atmosphere, and 3)
soil
surface energy balance.
change
heat
The air temperature
the PBL characteristics, and
determined by
changed
changes and virtual
in the PBL
the moisture
the sounding are
The
inversion strength
sensible
is determined from
(q)
of the PBL is
a budget calculation for the PBL.
changes made to
to the
to reflect the new
The PBL height and
due to surface temperature
flux.
into.three
Latent heat flux to the atmosphere.
temperature and moisture are
by
The last
imposed changes above the
PAGE 30
PBL.
These can be derived
trom known data
hourly SESAME soundings) or predicted
analysis.
The model
different
is
on the basis
three
of current
can be stopped and restarted to allow
imposed rates of change or changes
When a whole stable
layer
(such as
lay-er
incorporated
is absorbed
by
in cloud
cover.
the PBL, a new stable
from the next higher level
in the
sounding.
2.4:
Radiation
The radiation
parameterization
is taken
(1972) and
is a routine
originally designed
UCLA GCM.
The model
described
is
details are available
and
in
in Appendix 7.1.
layer.
gives a mean temperature
The model
for specific
content.
The
incident radiation
The net flux
incorporates an exponential
included
data of Yamamoto (I952),
in a fixed
and
its
and more
change for each model
humiditq to allow simple
CO is
for use in the
brief here,
IR emission ar-e calculated separately.
divergence
from Katayama
fit to the data
integration of water
form based
contribution
on experimental
is
then a
constant.
A.
Incident Radiation
The
influx of radiation
solar constant and
modifying
is computed by
starting with
it for albedo at the top
the
of the
PAGE 31
atmosphere.
separately,
Scattered and absorbable radiation are computed
the fraction being
for absorption, 65% scattered
part of the
assumed constant
to the ground).
incident radiation is corrected
(35% available
The
scattered
for multiple
reflection between the atmosphere and the ground.
Z
.WDSt SZT*-
GL
S
where
S o = solar constant = f(day of year) -units of
mcal/sqcm min
XT = zenith angle for time of day and location
(radians)
cS = scattering albedo for atmosphere--if clouds
are present they determine the scattering
aledo
= albedo of ground surface = f(hour angle)
If a
cloud
layer
is present, its presence
scattered and abs-orbable components.
enough, and
off.
covers enough
sky,
If the cloud
of cloud
is allowed, but
atmospheric
does not
layer expressed
incident radiation can be shut
as a percentage.
it may be composed
in each
one layer
of one or several
The model
include any feedback mechanism to the cloud amounts,
stoppped and restarted.
water vapor is calculated,
specific humidity.
a method
Only
layers of various percentage coverage.
hence these must be manipulated
is
both
is thick
The model allows for variable amounts of cloud
atmospheric
run
is felt by
from Wetzel
using
manually whenever the model
Atmospheric absorption by
the
sounding data for
Albedo at the surface
(1978) w-hich allows
is parameterized by
for
the change in
PAGE 32
albedo depending
absorbed in the
energy
at
balance.
the ground
on sun angle.
soil
part of the
=
.391
= absorption
So., IT -
in atmospheric
is
(.)
ABS
layer i
then
IR Emission
The
the
incident radiation
is
Total absorption at the ground
B.
is
becomes one component of the surface
The absorbed
GLW
where ABS(i)
The radiation which finally
equation of radiative transfer is
boundary
the top of
co-d4-itions .that
the atmosphere
the earth's surface is
temperature.
corrected
downward
is zero,
solved subject to
infra-red
and the
(IR)
flux at
upward IR flux at
t-he black body radiation at
the surface
We-ighted transmission functions are used,
for the pressure dependence of absorption by
defining an effective amount of an absorber.
The total
transmission function is assumed to be the product of the
individual
ones for CO and H O.
upward and
downward
obtained
The following
expressions for
fluxes at a particular height z are
(see Appendix 7.1
for details).
Downward flux
is
PAGE 33
IR
where
d
= 7B
z
rnBL
- rB tn(u
c
*
c
*
Z ,T)
*
*
- 7B )T(u-u
, T ) (B
c
o
c
z
*
*
B
+ fi
oT(u - u
, T) d (rB)
iiB
z
u
-
('LO
TT
S=
Stefan-Boltzmann constant
" = mean total transmission function for effective
absorber u# at temperature T
TC =-critical temperature which divides-the region of
weak temperature dependence of T
to that of strong
dependence of T
The weak
region is 210 to 320 K for water vapor.
= 220 K,
the weak dependence region need only
temperature specified = T .
Similarly,
So
letting Tc
have a mean
the upward flux
is
7B
IRu
and
=
irB z + f
the net upward
T
-- d('B)
T)
(W'5)
flux
IR
The only
g((u
z- U,
-B
z
= IR
diff-iculty
- IRd
*()
is determining the
transmission function near the particular
varies exponerrtially.
The model uses an
proper
level, where tau
interpolation factor
which
is an empirical function of pressure, mixing ratio and
layer
thickness.
This allows proper calculation of tau
without a fine vertical mesh.
are defined by
The mean transmission functions
empirical formulae at T,= 220 K and T = 260 K.
Temperature dependence of tau for CO, is neglected,
tau
for CO is
used based
on pressure and
so a mean
amount of COi.
The
PAGE 34
distribution
empirical
Yamamoto
of CO,at
functions
(1952).
each pressure
For thick
temperatures, with no net
the top
fit to data from
and
bottom are
body radiation at their
flux
of the
respective
inside.
expressions used
atmospheric absorption were
After
a)
clouds,
The
Radiation Data Comparison
All
and
is a constant.
for both absorbers were
assumed to radiate black
2.5:
level
the model was assembled,
for transmission
derived
from empirical
it was run
on sounding
allow comparison with radiation measurements,
comparison with other more complex
variation
models, and
of parameters such as aibedo and
ensure reasonable
in August and reported
were available
observations
model.
in Wetzel
plotted
The agreement
forecast near
data.
data to
b) allow
c) allow
cloud amount to
behavior.
Data for radiation measurements
Nebraska
coefficients
taken at O'Neill,
in Lettau and
(1978).
Davidson (1957),
Fig.2.2 shows
the
over the radiation calculations
is quite good,
with a
from the
slight over
1200 LST of about eight percent.
Three models were compared with the Katayama radiation
routine.
classic
Rogers and
Walshaw
parameterization
(1966) has been regarded as a
for many
years,
so the IR
PAGE 35
calculations were compared with it first.
Fig. 2.3 shows the
comparison for a sounding taken from Rogers and Walshaw.
Notice that the agreement
is very close.
On Fig.2.4, three
calculations are compared with the mod.el for an equatorial
sounding from London (1952).
Although differences exist above
8 km, below that level, the present model
is nearly in the
middle of the scatter of the rest of the calculations.
A
comparison is also made with a one-dimensional model.from
Wetzel
(1978).
Wetzel's parameterization was run on the
O'Neill data, and the comparison is shown for net radiation on
Fig.2.2.
Again, the agreement is quite satisfactory.
The radiation model's cloud routine was tested for
qualitative behavior for thin and thick clouds.
for the IR cooling- only appear on Fig.2. 5.
cloud, the top
warms strongly.
The results
For a very thick
-f th-e cloud cools rapidly, while the bottom
The cooling occurs because the flux for the
top of the cloud only co-mes from relatively cool
layers aloft,
while no contribution comes through the thick clouds below.
The cloud top radi-ates strongly to all layers above, so it has
a net flux divergence.
The opposite is true on the bottom of
the cloud, which gives a net warming effect.
The clear and
partly cloudy cases deviate from the extreme in the expected
way, with both the warming and cooling peaks losing
intensity.
Notice, however, that the sensitivity of the warming peak is
much greater than that for the cooling peak.
The partly cloud
PAGE 36
condition still
little warming
removed, all
gives
below.
of the
lagers above.
Not
radiation which
cooling
Presumably,
cloud tends
shown
at cloud
as the cloud
is the absorption
the cloud
but very
barrier
is
of incident
extreme values of cooling,
top of much
lower magnitude.
Surfa-ce Energy Balance
The surface energy
balance at the
surface- has the form
NR = SH + LH + GS
where NR
latent heat
heat flux
Fig.2.1
(7)
is the net radiation incident on the
the sensible heat flux upward
NR
top,
to radiate strongly to the
greatly reduces the
giving net cooling at
2. 6:
a strong
flux upward
from the surface, LH is the
from the surface and GS is the soil
downward into the ground which
heats the soil.
illust-rates the various fluxes and
is already
surface, SH is
their directions.
known via the radiation routine.
The rest are
parameterized as follows.
A.
Sensible and Latent Heat Fluxes
The sensible heat
parameterized
PBL.
layer,
The
flux
and
latent heat
flux
theory for the
using Monin-Obukhov similarity
fluxes depend
the depth
upon
the gradients in
of the boundary
layer, and
(SH,LH) are
the surface
the
incident
PAGE 37
radiation .
The theory assumes that the structure of
temperature and moisture .in the PBL have focrms which can be
described by universal structurTe functions when scaled
equations are used.
There are actually two structures
involved, since the PBL contains at least two distinct layers:
the surface layer and the well-mixed
layer.
If the functions
are required to be matched at their common boundary, the
following form results:
where
z = height above the ground
zo
ft
= roughness
length
h = t-he .d-pth of the boundary layer
L = Monir/bukhov length
fl
= uni-versa-! functions
The form of the temperature function is taken from Arya
(1975).
Details are in Appendix 7.2.
For stable boundarg layers, much scatter results when
data are compared with theory .
However, the present model
used only for unstable, well-mixed conditions.
is
These
conditions give quite good agreement between theory and
observations
(Businger,
moisture function, f.,
and
et al.,
1971).
Furthermore, the
is not well defined in the literature,
is usually assumed to be the same as the temperature
PAGE 38
structure function.
For-a well-mixed PBL this is
decent assumption, since both @ and
likely a
q are nearly constant with
height.
B.
Soil Heat Flux
The soil
(1975).
heat flux
(GS) is
Assuming vertical
variation of
surface soil
parameterized after Bhumralker
heat flux only,
and that the
temperature from an average
temperature
is
equation to
give eventually an expression for
flux
sinusoidal, one can
solve the heat conduction
the soil heat
Evaluated at the
(see Appendix 7.2 for details).
surface,
LC
C VI_
T
c = volumetric heat capacity
where
)
T BA
of the soil
TG = ground temperature
TBAR = some suitable average ground temperature
I1 = thermal conductivity of soil
S=-fre"uency of oscillation (= 2%/ 1 day)
2. 7:
Ground Variables
Two parameters are crucial to the calculation of all
the components of radiation:
parameterized
(1975)
by
TG and
"force-restore":
and Deardorff
(1977).
q(ground).
of
These are
methods from Bhumralker
Details of the derivations are
PAGE 39
in Appendix 7.3.
solved
for a
surface.
where
To
find TO the heat conduction equation
lager between 50 cm and
1 cm below ground
This gives a prediction equation
t3
c,
and
k
are as previously
The soil moisture is
for TO
defined.
found by assuming
that
moisture responds to three main processes:
evaporation,
(GWB)
and
is assumed
DeardorffP
flux
surface soil
precipitation,
The bulk soil
moisture
to be constant over the period.
(1977) the
scale of a few weeks,
for a
from below.
bulk
soil
so GWB
12 hour period with
According to
moisture changes over a time
can certainly
little loss of
surface soil maisture is changed
be assumed
accuracy.
according
constant
The
to:
where
GWB
GW
di
X
9s
WMAX
'
c Jc,
=
=
=
=
=
=
=
% bulk soil saturation (top 50 cm)
" soil saturation
depth of diurnal cycle (=10 cm)
latent heat of evaporation
density of H20 = 1 gm/cc
field capacity soil moisture
period of cycle
are non dimensional constants
Deardorff's values
Jackson
(1973),
for c and
is
c 2were computed
measurements taken over
bar-e
from data of
soil
near
PA(GE 40
Phoenix, Arizona
in March.
Cwt
.5
Notice the middle value of c
the
two
2.8:
is a
linear interpolation between
extreme values.
Boundaru
Laer Variables
The depth
of the boundary
inversion
(1977).
(A4)
layer
(h) and
strength
Their method
assumes that the
PBL depth
No allowance is made for the
The energy
the
surface, and
strength
energy
of the
changes due
inversion into the
late afternoon collapse of
the PBL, since the convective outbreaks occurred
time.
of the
are predicted according to Zeman and Tennekes
to turbulent entrainment of air above the
PBL.
7/o
prior to this
comes from the virtual SH flux at the
change of
inversion.
depth with
They
time depends upon the
use the turbulent
kinetic
budget to develop a simple set of equations to describe
this process.
The equations which result are:
w
*lr
=
T
9
h
where
TS
h
g
w.
=
=
=
=
surface temperature (top of surface
height of inversion
acceleration of gravity
convective velocity scale
layer)
PAGE 41
VSH = virtual
sensible heat
flux
at ground
and,
Wbv
where
W
S
T = temperature gradient above the inversion
= Brunt-Vaisala frequency in the air above
inversion
\SH
C
the
- Cd bh
c wZ T
t
S
gh AG
1+
5
where VSH k = heat flux at the inversion (C 0 to give
entrainment from above)
are dimenrsionless coefficients which are
c&,cf ,c
Zeman, 1975)
(1&)
= 3. 55
c
c& = O. 024
c. = O. 50
In the case where
atmosphere
6
=
O0 ,
no
inversion exists and
presents no barrier to inversion rise.
case, the model
A
assumes a very small value fort8
(from
the
In this
since the
inversion must rise at a rapid but finite rate due to the
turbulent
entrainment.
Finally,
These equations will allow the
PBL characteristics.
calculation of the necessary
See Appendix
7.4 ?or a discussion of
PAGE 42
their derivation.
2.9:
PBL Temperature and Moisture
The final calculation to be performed
values of T(
i) and Q.
is that for PBL
These are found using the simple
assumed structure in the PBL, and a budget for
0
and Q.
pressure level of the top of the inversion is known.
The
The
change in the height of the. inversion has just been
calculated, so the amount of entrainment is known.
This
entrainment comes from the inversion layer, so the new depth
of the inversion layer can be found.
Ae,
Using the new value for
the potential temperature at the top of -the inversion, and
the lapse rate of potential temperature in the inversion, the
potential temp-era-t-ure -of the PBL can be calculated.
gives the new TS value,
the
inversion hattom.
This then
and the hydrostatic pressure level of
Since the pressure depth of the surface
layer is fixed,-- the height of the surface layer can be
computed hydrostatically.
The process is illustrated
schematically in Fig. 2.6.
2.10:
Initialization Procedure
The model requires the initial sounding
specific form, to allow the model
reasonably.
to have a
procedures to operate
In particular, a thin surface layer and a
PAGE 43
well-mixed PBL above are
atmosphere
later
is
uniikely
in the morning,
Nevertheless,
immediately
the
both assumed.
to possess
especially
initial
to the model
Realistically, the
layers until
either of these
in
the presence of
sounding must be
forced
clouds.
to
conform
requirements to prevent model
collapse.
initial
sounding
down to near the
surface.
The
SESAME
The
tapes at 25mb
(ie.
within
So
10 mb
if
level w hich
this first
GS in
cannot handle very
d ata point above
pressure.
it
The
C higher than
to be 0. 1 deg
is
to the surface value.
set equal
initialized
at 90 m, measured
above 90 m,
up to
layer.
is
If this
ground
the first
first
set to a minimal
at the top
ground
stable layer
value to allow
temperature
The PBL depth
is
of the
the surface
from the ground.
data point,
is
initial TS
The moisture
temperature to produce positive SH flux.
PBL
the
thin
is rejected.
These values are set
above the ground.
surface layer, 5mb
is always
integral multiple of 25
observation is used to determine
the model.
initialized
is an
The model
of the surface
The surface
is
is available from the
levels ! plus the surface-observation.
875mb or 950-Gb).
layers.
and
The data
first data point above- the surface on the tapes
taken at a pr-essure
mb
i s input as available from 400 mb.
the
is neutral,
for
the
is
The atmosphere
first stable
the stability
finite but fsst
growth.
PAGE 44
The delta theta
extending
the
calculating
on
the
potential
and
temperature.
air used
This
for T
procedure seems
enough
detectable when plotted
to be barely
chart.
15 minutes
In addition:
of ground
with this
For a typical
changes are
small
on a
results from a
illumination,
sounding,
the change
in the
sounding
study
by
layer appears
so the
errors
initialization procedure should
the adjusted portion of the
dry
is less
static
be small.
energy
of
than 0. 1%.
Sensitivity Tests
2. 11:
There are a number of constants used
of
in the
a major
but the actual
(1982) suggest that a small mixed
associated
layer are
These changes allow the mixed layer
to work properyi.
within
this new
like
sounding,
Whiteman
layer above
lapse rate
modification to the
model
is shown
in the model.
initialization
pseudoadiabatic
level, and
isothermal
the
,
by
This process
the characteristics of
to determine the value
stab.le
is derived
to the 90 m
This operation creates an
initial PBL,
used
iteration
first data point down
its
Fig. 2. 7.
for the first
which are
tests are
not
easily evaluated.
performed to
insure that
in
A number
this model,
of sensitivity
these arbitrary
are not
controlling the results.
The sounding
purpose
is
project
taken
from a PBL
Field
many
constants
used for this
at O'Neill,
Nebraska
PAGE 45
on
13 August
1953.
(This is the same sounding which
already been mentioned
above. ) The sounding
against which
values
cc
in the radiation comparison discussion
appears in Fig. 2.8.
the others are
TBAR = 296.5 K,
soil,
addition, initial
standard value.
GW
PBL are
and VS =
The results of the
imposed
No
1000 cm/s.
using
(Fig. 2. 13
the other
c.onditi6ns.
rises rapidly
1400 local solar time
(LST).
(Fig. 2.10) is strongest at
afterwards.
The lag
temperature peak
soil
to heat.
is
due to the
incident solar radiation
begins to fall
in thermal
surface up
to about 825 mb
equilibrium with the
near midday,
the PBL.
is very
reaching 9325 mb very
the ground
late
the
The PBL moisture
The LH flux
to strongly moisten
off
so continues to heat under
strong radiation.
sufficient
silowly,
peaking at
finite time required for the
is not
(Fig. 2. 12).
under
The temperature of the
1200 and
rises rapidly until
latent heat flux
strong heating
during the day,
The
incident radiation at noon, and
(Fig. 2. 11)
in
sensitivity runs.
between the radiation peak and
The soil
decreasing but still
55% as the
standard run appear
The standard run is characterized by
ground
In
clouds or other changes above the
in this or any of
moderately moist soil
standard run
GWB = 90%, WMAX = 0.80
(GWO) is varied,
Figs.2.9 through 2.13.
The
compared uses the following
for the various constants:
H20/cc
has
in response to
in this run
The air above
stable, and
in the run.
the
is
the
the PBL grows
The ground
PAGE 46
temperature reaches 314 K,
while the surface air temperature
TS reaches 308 K shortly after.
nearly
A.
The moisture at this time
14 g/kg.
TBAR Test
TBAR
is the
constant with
diurnal
time.
diurnal average soil temperature, assumed
depth.
variation
At a depth
(about 50 cm),
equal to the
TBAR = T(soil)
constant for the -iodel run.
the standard of 296.5 K.
The timing
It
The results appear
of the peak values does not
greatest response, particularly
TBAR
the
for QS which
not have a
B.
changes at all
is
change in
to the
ground temperature
show little response,
hardly
(almost +/-
soil heat flux
responds noticeably, showing a response of +/other parameters
4 K from
in Table 2.2.
heat flux
in the
Furthermore, the
+/-
The fluxes show the
quantity TG-TBAR, the model's response
is expected.
temperature,
change over the range
soil
Since -one of the two factors
independent of
is varied
of TBARs, only the values themselves do.
20%).
L.imit of the
So TBAR can be regarded as a 50 cm soil
assumed
the
is
2
to 4%.
particularly
(+/- 0. 1 g/kg).
The
the value
TBAR does
great effect on the model output.
VS Test
VS is
the magnitude of the wind
speed at the
top
of the
PAGE 47
surface
layer.
It
is used
in the calculations
(SH) and
sensible heat flux
heat flux
latent
standard run uses a value consistent with
The test varied
speed.
expected,
observations.
of the observed
As
in Table 2.2.
the SH and LH change substantially,
the VS= 200 run
heat flux
of the test appear
The
(LH).
local
this value from 20% to 150%
The results
for the
about 20% for
However, the soil
(20% of the observation).
(GS) changes more. than either LH or
SH.
The maximum
value for GS for the VS=200 run
is almost double the VS=1500
(150% of the observation)
Since GS depends
for
its calculation,
the
soil heat flux.
The ground
is
linked
run.
on SH and LH
the changes in SH and LH are additive for
temperature also responds strongly,
to the so-il h-eat
for the weakest wind runF
VS=200.
temperature rises rap-idly as
it
The changes are pronounced
flux.
The others cluster much
the maximum-ground
the wind •speed drops,
As
closer togetr-er.
since
does the soil heat
real PBL with we-ak winds, the effective
In a
flux.
surface layer
in which
the eddies remove the heat and moisture from the surface
likely becomes smaller
eddies
(less frictional
less vigorous and
efficient.
is
left
in the soil,
contributing
to a
larger soil
radiation
effects are part of
changes
in wind
both
turbulence),
So more
heating
heat flux.
the parameterization.
speed
(+/- 50%)
and the
incident
it and
In the model,
these
For moderate
the effect is not
large.
For
PAGE 48
a factor of 5,
in VS by
a reduction
the change becomes very
important.
The variation in
The surface temperature
is significant.
the VS=200 run, and
for
is anomalous.
other.
a
critical
Usually,
parameter,
strength.
C
is 851 .mb, 30
but only
Again this
the standard run.
closer to
for very
each
speed becomes
that the surface wind
small values.
deep PBLs, the wind
in strong BLH and
moderate in
I deg
lower by
The other runs are all much
true.. then,
is
It
the PBL depth
the surface moisture higher by nearly
the atmosphere than
higher in
run
is
final PBL depth.for the VS=200 run
The
1 g/kg.
mb
the surface parameters and
So the VS parameter
is at
least
does not require
beyond choosing a value consistent with the
extreme care,
observations.
C.
WMAX and GiWB Test
The
bulk
soil
effects of the soil
moisture
(GWB) are discussed
shows the results.
of these two
As
(WMAX) and the
together.
one can see on the table,
Table 2.3
the variation
parameters produces results of similar magnitude.
The runs with
wide variation
WMAX
moisture capacity
standard GWB
(=90%
the wettest GWB value) show
over the range of WMAX.
(=0.80 cc HIO/cc
over the GWB values.
soil)
The
The runs with
standard
show nearly the same variation
only run which
is not
similar
is the
PAGE 49
wetter WMAX value (=1.00 cc
H O/cc
corresponding wetter GWB value;
than the standard.
cooler and
The general
wetter PDBLs,
and
soil).
There is no
all
the GWB values are dryer
trend
for both parameters is
lower inversions for wetter
parameter values.
The variation within each of the tests
the TBAR test.
the LH flux
change
is larger than in
The SH flux almost doubles from wet to dry and
drops by 30%.
is 2 deg C
The maximum surface
temperature
over the range of WMAX, a difference of more
than 10% in the diurnal temperature change.
The surface
moisture also depends strongly on this parameter, varying by
nearly 2 g/kg over the WMAX values.
change drastically.
The PBL characteristics
The dryest WMAX run has a final inversion
depth of twice th-at for the wettest run.
significant and imply a sensitivity
WMAX.
These changes are
in the model
Fortunat-ely, the changes are similar for the two
parameters, and hence a range of values for
while holding the other constant.
runs,
the value for WMAX
similar to a value used
data.
to GWB and
The GWB is
past rainfall.
During
the case study model
is fixed at 0.50 cc
b4 Deardorff
then varied
over a
one can be used
H.O/cc soil)
(1977) to model Kansas
large range
depending on
These results are then examined in
the case
studies to determine which are physically realistic.
PAGE 50
D.
GWO Test
The results
soil moisture
tests.
of the test
(GWO),
involving the
can be anticipated
The value of GWO makes a very
model runs.
surface
from the GWB,.WMAX
large difference
These results are shown on Table 2.3.
that the variation from standard values
-50%
initial
(gWO=70%,25%),
Notice
in GWO runs
just as for the WMAX tests.
in the
is +25%,
This allows
simple comparison of the magnitudes of the changes
in model
results.
The flux variations are similar to the WMAX runs.
nearly doubles
The change
almost 4 K,
from wet t-
dry,
while the LH drops by 30%.
in maximum surface temperature for the GWO runs
The surface conditions are 2
to 4 times mor-e sensitive to the GWO values than
WMAX parameters.
The height of the
is more than 3 times the height
initial
Pielke
the PBL development
(1981)
found
for the wettest run.
in the model.
that their model was
which
in
In
the case studies, a
The
parameter in
McCumber and
sensitive to soil
moisture, and Cooper et al. (1982) also found
important.
to the GWB or
inversion for the dryest
soil moist-ure is clearly a crucial
controlling
is
and the surface moisture changes more than 8 g/kg
(change from-GWo = 25% to- 70%).
run
The SH
soil moisture
series of runs are made
use a wide range of GWO values.
The range is determined
conjunction with the previous day's rainfall, but
so many
PAGE 51
other unknowns can
precipitation,
range can
2. 12:
affect the surface soil
plant cover,
be specified
soil drainage, etc. ) that only a
before the actual runs are made.
is
compared with observations
O'Neill,Nebraska PBL field study,
(1977).
modeling by Barnard
0.80 cc
shown
H O/cc soil,
and
about 8%
the
detailed soil
run uses
the
GWB = 45%, and GWO = 65%.
The sounding
in Fig. 2.8.
incident radiation
2. 14 and 2. 15.
(Fig.2.2) is
(as alrea-dy noted above):
integration.
flux
The soil heat
by
throughout most of the
is
just below on the
plotted
shown, to give some idea of the scatter
first four hours of
radiation to the
the run.
judging
Indeed,
soil flux.
difference in net
flux.
The reason for
The
soil heat flux
and
one near
actually
1200 LST.
in the data.
the model partitions
2. 15),
figures for SH and LH (Figs.2. 14 and
of the
over predicted
Notice
The observations which are plotted have error bars
same axes.
heat
during
VS= 1000 cm/s, WMAX =
The fluxes are shown in Figs.2. 2,
the net
taken
then wi-th
The O'Neill
TBAR=296.5 K,
following parameters:
all
(timing of
Comparison Runs
The model
is
moisture
radiation is
this model
If the soil
is
too much
from the similar
it appears that
given to the soil
behavior
has two peaks)
For the
is not
clear.
one near 800 LST
initially very
dry
at
PAGE 52
the
surface,
the earlier peak
moisture the
two peaks merge
disappears.
For very
into one near 1100 LST.
be evidence that the moisture dependent soil
affecting the
time
much
slower
the observed
(Fig. 2. 14).
the
observed flux,
The
latent
the
observed value
very
heat flux
and
flux
is as big as
is also much
in
falls
50% of
bar value.
different from
it
is within the
the LH was not well
so the observations are
imprecise.
fluxes is annoying,
important fo-r this thesis as
is
moisture, which
of the nearest error
Evidently,
anomalous behavior of the
capacity
the model
in the afternoon, although
large err-or bars.
observed,
The difference
(Fig. 2. 15)
This may
has difficulties late
flux falls, and
and 25%
bulk
in the morning.
The sensible heat flux modeling
the day when
heat
behavior--depending on the soil
itself changes rapidly with
high
the effects that
This
but
is not as
these flux
errors have on the PBL characteristics themselves.
As
the graph
effects are
morning
lower
C.
in
(Fig.2. 16) these
The soil
temperature is
the model
than the
observations, and
The difference
The effects on surface
run
temperature shows
slight.
temperature.
(Fig. 2. 17).
C
of soil
A 2
(700 LST),
for the rest
layer
is
oF the run
until
large, at most 2 deg
evident very
difference drops
1800 LST.
the
peaks at a
temperature are even
deg C difference
but this
is not
higher during
to less
smaller
early
in the
than 0. 5 deg
At this
time the
PAGE 53
observations
show cooling, which
Wetzel
(1978) attributes to
large scale advection, not PBL dynamics.
temperature
is very well
modeled.
The height of the inversion
is similarly
well handled.
within
of the observations until
100 m
the O'Neill atmosphere no
So the surface
Fig.2. 18 shows that the model
1400 LST.
longer has a well
is
After this,
defined PBL, so no
data is available after this time.
Barnard Comparison
The model reproduces -the fluxes at O'Neill
well,
with certain exceptions.
temperature,
generally
It also models the soil
surface temperature- and PBL height with
accuracy.
To test the model
a detailed
soil moisture modeling study.
an 80 layer model to
further,
is run with
Barnard
to soil moisture.
Fig. 2. 19.
For the present model,
The
sounding used
data from
(1977) uses
investigate climatic sensitivity
atmos.phere
TBAR = 310.0 K,
it
some
of the
is shown
in
the parameters used are:
VS = 1000 cm/s, WMAX = 0.50 cc. H,O/cc soil,
GWB = 50%, and GWO = 6%.
The model
results are
shown on
Figs. 2.20 and 2.21 along with Barnard's results for comparison
for two
times.
Notice that the values
compare very well,
well-mixed
and
the results of the modeling of the
layer agree quite well
excellent considering that
form.
of surface moisture
too.
These results are
the present model
is so simple in
PAGE 54
2.13:
Model
Comparison Summary
The model
reproducing
during a
above
is simple but effective
the PBL characteristics as
typical
inversion,
most
outlined
spring or summer
parameters for
instability and
on other model
temperature.
The model
results of PDL field
studies
detailed modeling work
It could
the model,
surface moisture are
(such
be argued
on soil
such as
These PBL
moisture, and weakly
surface wind
speed
as Barnard,
detail
reproduce the
1977).
should
different soil
multiple soil
layers and vegetation.
for not doing
so
be included
types,
such
unavilable and
would
require sweeping assumptions rendering the detailed
balance
complicated models
Thirdig,
work
details are,
moisture.
the model
Barnard's
by McCumber and Pielke
in
They
fact,
less
physics
(1981)
of soil
the
to duplicate more
1977 model)
important than
tested a model
itself
does seem able to model
sufficiently well
(e.g.
in
The most telling reason
is that data to initialize
parameterizations is generally
surface energy
or 50 cm
(such as O'Neill, Nebraska) and
that more
Secondly,
the
of convective
can successfully
incorporating perhaps
meaningless.
time
The height of the
inhibition to convection.
parameters
change with
the assessment
characteristics depend strongly
soil
day.
surface temperature and
important
they
in
and real
data.
indicates that the
the surface soil
characteristics which
PAGE 55
included
several soil
types, and discussed
moisture was much more
type.
Therefore,
influential
the model
form to analyze SESAME,
raw form and
numbers
for pressure level
instability
tabulated
(TS),
(PLI),
and
is used
for each
in the form of
layer temperature
for the fluxes than soil
in its
1979 case study
The modeling results
that the soil
present, simple
dat.a.
case are presented
"sensitivity values".
of the top
PBL moisture
of the PBL
(GS),
in both
The raw
(PH),
surface
convective
convective inhibition
(NA) are
for each run in each case.
The sensitivity values are referred
and tabulated at
to in the discussion
the end of the tables.
defined by referring to the model
These values are
run without any modifying
physical
factor-s.
example,
suppose the plain run for 50-50 soil
This run
parameters yielded
are added to
growth
the models
is called the
of the PBL of
the growth
"plain" run.
For
moisture
159 mb.
is reduced by
When clouds
10 mb.
The
sensitivity value for PBL growth' for the addition of clouds is
then:
Sensitivity =
(Growth with
=
growth
TS,
GS, and
10 mb
clouds - Plain growth)/
/ 159 mb =
PLI are handled
Plain
"b%.
similarly
The convective
PAGE 56
inhibition cannot be handled in this manner, since the initial
NA is undefined, and henc.e the change in NA is undefined as
well.
So, the sensitivity values for the convective
inhibition are defined by simple comparison of the values at
the end of the model runs with those of the plain runs.
examples
For
if the plain run discussed above had an NA of 20 and
the run with clouds had a NA of 30, the sensitivity value
would be:
Sensitivity = (NA of run with clouds = 10/20 = 50%.
Plain NA)/ Plain NA
PAGE 57
Table 2. 1:
SCHEDULE
OF MODEL CALCULATIONS
After initialization, the model runs as follows
for each time step.
New values for variables are
labeled New (variable).
Old values are left plain
(variable).
1.
Radiation Routine
A.
New IR flux = f(T i ,p
qL ) -sounding data plus
geographical location, time, day, etc.
B.
New Incident flux = f(Tj pq
same extra items as for IR flux
) --
sounding
data plus
Results in AT at each model level due to net radiative
flux divergence and the net radiation, NR, absorbed in
the soil surface.
The AT is not applied to the
sounding until the end of the cycle.
2.
Soil Heat -Flux
A.
New GS = f(TG, TGpastCone time step before], soil
characteristics such as heat capacity, thermal
conductivit. = fEground wetness] )
Results
3.
A.
New SH = f(T-G,GW, TSGSVS,
B.
New LH-= f(u*
and constants VKZ
GW,VS)
in new values for SH,LH.
PDL Characteristics
A.
New h = f(New SH, New LH,
B.
NewA& = f(Ae, Y
Results
5.
for GS.
Surface Sensible Heat Flux and Latent Heat Flux
Results
4.
in a new-value
in
,
New SH,
new values for h,
TS,
h,
Y
New LH,
)
h)
&.
Ground Variables
A.
New TO = f(T ,
New NR,
New SH,
New LH,
TBAR
(a
)
PAGE 58
constant), and soil
B.
New GW = f(GW,
and c )
characteristics)
Jew LH,
and constants WMAX:
Results in new values for TG,
6.
c = f(GW),
GW.
P3BL Variables
A.
New TS = f(New h, Newb, Y
B.
New GS = f(GS, qEinversion layer],
)
h, New h, New LH)
Results in new values for TS,GS.
7..
Change Sounding
A.
B.
C.
D.
Change
top of PBL --
New T,p
f=(New h),q
If current inversion is filled, find new stable layer
(next layer above) and recompute
.
Fill
in surface
(TS,GS) and ground
(TG,GW) variables.
Compute imposed changes above PBL -- change Tq
appropriately.
If necessary, changeT" too.
Results in new sounding at end of time step.
8.
Output Variables
PAGE 59
TABLE 2. 2:
MODEL RESULTS FOR SENSITIVITY TESTS
TBAR TEST
TBAR
SH
LH
OS
292. 5
294. 5
296. 5*
298. 5
300. 5
260
265
270
273
278
583
593
602
614
623
209
193
176
159
142
TO
TS
OS
36. 4
36. 7
37. i
37. 4
37. 8
32. 8
33. 0
33. 2
33. 4
33. 6
14. 4
14. 3
14. 3
14. 2
14. 3
833
828
822
816
807o
1108
1161
1233
1292
1388
PH(m)
PH(mb) PH(m)
VS TEST
VS
SH
LH
GS
200
600
1000*
1200
1500
224
262
270
269
270
540
582
602
610
618
296
201
176
168
160
TG
TS
GS
PH(mb)
44. 8
39. 0
37. 1
36. 5
36. 1
32. 0
33. 0
33. 2
33. 2
33. 3
15. 2
14. 5
14. 3
14. 2
14. 3
851
830
822
820
817
928
1149
1233
1247
1281
Fluxes are in mcal/sq cm min.
SH = sensible heat flux, LH =
latent heat flx, OS = soil heat flux.
TO = ground
temperature (deg C), TS = surface layer temperature (deg C),
QS = PBL mixing ratio (g/kg), PH = level of top of PBL.
TBAR = 50 cm soil temperature (K), VS = PBL wind speed
(cm/sec), GWB = bulk soil moisture (% of saturation), GWO
surface soil moisture (% of saturation), WMAX = saturation
soil moisture (cc H20/cc soil).
* denotes standard run.
PAGE 60
TABLE 2. 3:
MODEL RESULTS FOR SENSITIVITY TESTS
(CONT.)
GWB TEST
GWB
SH
LH
GS
30%
50%
70%
90%*
374
330
292
270
464
511
558
602
226
203
188
176
TO
TS
GS.
41. 7
40. 2
34. 8
34. 3
33. 8
33. 2
12. 9
13. 3
13. 7
14. 3
38. 6
37. 1
PH(mb) PH(m)
730
763
797
822
2254
1871
1498
1233
WMAX TEST
WMAX
SH
LH
GS
0.40
0.60
0. 80*
1.00
426
303
270
261
455
552
602
633
256
191
176
169
TG
TS
GS
PH(mb) PH(m)
42. 7
39. 1
37. 1
36. 0
34. 8
33. 9
33. 2
32. 8
13. 3
13. 7
14. 3
15.0
724
792
822
834
2321
1552
1233
1098
GWO TEST
GWO
SH
LH
25%
35%
45%
55%*
70%
405
359
311
270
223
399' 234
461
216
529" 194
602
176
146
716
See
Table 2. 2
GOS
TO
42.
41.
39.
37.
33.
for caption.
TS
5
2
3
1
4
35. 1
34. 8
34. 1
33. 1
31. 3
QS
10.
11.
12.
14.
18.
PH(mb) PH(m)
4
5
7
3
7
703
737
779
822
861
2569
2172
1691
1233
818
PAGE 61
5 -fa 6k
talcp
PrB
JIG;S
/
&Wi'
s6;l k~11,v-=BAI
- --- - -
Fig. 2. 1
JT
11
.
Schematic diagram of model atmosphere. Hypothetical
temperatures and dewpoints are shown by upright
lines, while model layers are delineated by
horizontal lines. Variables are defined in the text.
Also shown is flux diagram showing positive direction
of fluxes at the surface.
PAGE 62
s eYAnos bE
x-- x PREEawrrA
1000 -
'
NR
900
800 700600 500 -
400 300 -
/
200-
/
100
GS
/
6
Fig. 2. 2
/
7 8
9 10 11 12 13 14 15 16 17 18
LST
Comparison of model net radiation (NR) and soil (OS)
fluxes (meal/ sq cm min ) with observations and
modelling by Wetzel(1978) for O'Neill day number 2.
b
...
....
W6
o00
SMo
,,
10o0
+
0
Pqaef+tAdI1IOL4
3
,.c/dLy
Fig. 2. 3
Comparison of model IR cooling rates (solid lines)
with calculations of Rogers and Walshaw (1966)
(dashed lines).
PAGE 63
?l~c)
c5
I'
'i
56)
w 9
*--.c--u 0~ok =
o,- oc
O
sse
IOL
EcA
Lv
I
~
/
S
~1
I'
Fig. 2. 4
2
Same as Fig. 2.3 for comparison with Brooks(1950)
Elsasser (1942) (dashed
(soLid line connecting *),
line), and ECMWF (1979) (dash-dotted line).
2P(*
0o
2.00
too
'60
700
oO
ClO14J
LA1O~t
b.
$
CLa
8cd
CI'J
Q)/2
4,
O=Z~
___
_
T.
S
Fig. 2. 5
;
0
c\"'
--
Y;--
-/
-
3
fc, - rsf
-Y
-5
-4
7
-8
Model IR cooling rates (only) for c loud layer between
880 mb and 820 mb, of 50% coverage (dott ed line),
100% coverage (solid line), and no cloud (dash-dotted
line).
-9
cIb
-to
9 CT 6
elt3
aw e4'
~tw ~LrA)
&LTOe
.
CO~di4La
l&;ti~j
'y%5
CL cUwAs±r"&
Fig. 2.6
?EUJAB
1A
01,L p-
S-
67 1)
b
Clriv'n
t
-
C
IkCL
,-tu.As'so,6-
(OV'ornbb
Pvtss
Schematic diagram illustrating process of finding new
PBL potential temperature from new inversion
calculations. Horizontal axis is potential
temperature.
)J
BLA
\C
fl .
ope
.'4
eoi0
L
dr
O
obfl
Fig. 2. 7
Schematic diagram showing process of initialization
of S from initial sounding. Vertical axis is height.
Horizontal axis on left side is temperature, and on
right side is potential temperature.
-
PAGE 67
600
700
800
-900
1000;
-40
-30
-20
0
-10
10
20
30
40
TEMP
Fig. 2. 8
Sounding plotted on a pseudoadiabatic diagram from
Line connecting dots
O'Neill, Nebraska at 1200 GMT.
connecting * for
line
and
C)
(deg
for- temperature
deup-aint (deg C).
RoO
q5O
L
_
IL
6
Fig. 2. 9
L
7
1
i.
it
~~~I-~~--
-- I
t
II
1
--------
.
--
8 9 10 11 12 13 14 15 16 1-7 18
Time variation of pressure level of PBL top
standard model run.
(PH)
for
j
PAGE 68
FLUX
1000
900
800.
700
600
500
460
300
200
100
0
6
Fig. 2. 10
7
8 9
10 11 12 13 14 15 16 17 18
LST
Net radiation (NR) and sensible heat (SH) fluxes
(mcal/-sq cm min ) for standard model run.
P4-0
I1-0
Fig. 2. 11
A
10
.1
8
9
6
7
8
9 10111213 14 15 16 1"7 18
12
I
14
I
I
7
11
I
I
13
6
15
16
17
1
Time variation of PBL mixing ratio (GS) for standard
model run.
PAGE 69
FLUX
1000
900
800700
600
LH
500
400
300
GS
200
100
0
8 9 10 1.1 12 13 14 15 16 1"7 18
LST
Fig. 2. 12
Same as Fig. 2.10 for latent heat (LH) and soil heat
(GS) fluxes.
40F
TG
36
34
32
TS
Xx
30
28
26
y,
24
22
20
18
6 7
Fig. 2. 13
8 9 10 11 12 13 14 15 16 17 18
LST
Time variation of ground temperature (TG) and surface
layer temperature (TS) for standard model run.
PAGE 70
1000
900
800700
600
500
400
300
200
100
0
7
6
Fig. 2. 14-
8
9 10 11 12 13 14 15
LST
17 18
16
Same as Fig.2.2 for sensible heat flux (SH).
FLUX
1000-
900
800>
LH
700. -
fAEobL.
600
500 -
400:
400
200 .
6
Fig. 2. 15
7
8
9
10 11 12 13 14 15 16 17 18
LST
Same as Fig. 2.2 for latent heat flux (LH).
PAGE 71
TG
4038
36
,4
3432
I
/
/
28
I
I
26.
/
/
24
I
/
/
22
20
18-
/
/
--
--- 4
I " |
6 7
Fig. 2. 1.6
1%
I
30
~-----
L
- I
I
I
-
"
I
--
I
l
I
|
8 9 10 11 12 13 14 15 16 17 18
LST
Comparison of model output (solid line) and observed
(da-shed line) ground temperature (TG) for
O'Neill sounding.
TS
40
38
36
34
32
3Q
28
I
-
II
I
26
24
22
20
18
.
i
.
-
. a.
-•
F I
.
I
.
-
I
%'
.
_-
# - ' TI
?
:
-
_.
..
.
, .
.
.
..
I
.
.
..
"
f
.
._
I
6 7 8 9 10 11 12 13 14 15 16 17 18
LST
Fig. 2. 17
Same as Fig. 2. 16 for surface layer temperature
(TS).
PAGE 72
tso
4500
1000
'Soo-
I
6
Fig. 2. 18
I
--
SI
7
. -
i
I
8
-I
i
I
I
I
I
I
I
-- L
i
I.
i
.
I
9 10 11 12 13 14 15 16 17 18
Same as Fig.2.16 for growth of PBL (HOI in meters).
P
600,
700;
800
900
100oo
40
TEMP
Fig. 2. 19
Same as Fig. 2.8 for
( 1977).
initial
sounding
from Barnard
-( k,
3
I
!
I
I
t
I
SI
7
0
1
1
3
/
,5S
7
I
L
3
1600 kLr
L
70o k-
J , 1-4-2--- P
Mo-l
F'ig. 2. 20
Comparison of Barnard's (1977) model output (dashed
line) with present model (solid line) for PBL
moisture at 0700 LST.
Fig. 2. 21
Same as Fig. 2. 21 for 1000 LST.
PAGE 74
CASE STUDY I:
3. 1:
Introduction
The period
12 GMT April
characterized by weak
Fig. 3. 1.
Most of the
GMT, while one
day on
3.2:
constant in
Some of- the
outbreak
that the
During
in many different areas at
convection was ongoing
at
12
occurred in western Kansas very
19 April.
After a
this last outbreak
brief review of the
is studied
series of 500 mb analyses
cyclonic vorticity
late
synoptic
in detail.
Nebraska by
associated
surface trough
t-hough
the
just east
forcing
which
had
Nebraska, the surface trough
20 April
12 GMT moved
(Fig. 3. 5).
into western
Fig.3.4 shows the
stretched north/south across
of the Rocky
aloft
shows
(CVA) from a short wave
OO GMT 20 April.
Kansas and
the country
(Figs. 3.2 and 3. 3)
advection
Idaho west of the region at
GMT,
intensity.
Synoptic Analysis
The
in
throughout much
large-scale features were nearly
convection occurred
different times.
conditions,
scale activity
The SESAME region .is defined on
stationary, and were nearly
this period
19 to 00 GMT April 20 was
synoptic
of the SESAME region.
in the
19 APRIL
moved
Mountains at 12 GMT.
Even
over western Kansas
barely moved
or deepened
The SESAME region
and
by
expe-rienced
00
PAGE 75
undisturbed
this
during
3. 3:
southerly
flow at the surface most
of the 12 hours
period.
Mesoscale Analysis
in
The convection broke out almost instantaneously
between
and 22 GMT
21
in
western
this outbreak.
Notice
Kansas
that the
began earlier near Amarillo,. Texas
located
line is
It
the north
in Kansas was not forced by
and
clear from
storm.
these
Also,
On
GMT
line clearly
originated by
the surface analyses,
(Fig.3. 10) as a
temperature
16 GMT
in western
convection
(Fig. 3. 11),
in
the
the convection to
outflow
from the
Fig. 3.6 ,
as 100 km north
of
not
an
simultaneous.
the
"his
independent mechanism.
a warm tongue
at
was evident
12
large area of 298-300 K potential
Kansas and the Texas Panhandle.
tongue of warm air became
Bu
The squall
if this had been redevelopment,
outbreak would have been sequential
squall
convection
only 20 minutes after
Fig.3.7 was
storm.
(AMA).
films that
convection was beginning as far away
the southern
(OCK) at four
along the 20-25 degree
azimuth.
southern
is
the northeast
to
3. 9
large echo to the
south was a hailstorm which was associated with
which
line
Figs. 3. 6 through
Kansas.
show the radar film taken at Garden City,
times during
a
warmer
the cool
Panhandle
and better
outflow from
had distorted
defined
This
with
the active
he
shape
o
t'e
time.
PAGE 76
tongue.
However,
mostly clear
to the north
(Fig.3. 12) and
(Fig. 3. 13) a strong
in western Kansas,
the BLH continued.
By 21 GMT
temperature gradient existed
eastern edge of the intense warm
convergence was evident in
tongue.
skies were
on the
Notice that
the surface winds across this
gradient.
By 22 GMT (Fig.3. 14) the
notice
that the outbreak
convection had erupted,
was apparently
on the warm side of
the gradient, near
the axis of the warm tongue,
et al. (1958) found
in
convergence
surface wind
in the
changes were clearlq
3.4:
11
(Fig.3. 15)
showed a wet
This was a region of weak
field.
The thermodynamic
are examined here.
the
to a
the
(DDC) sounding
the surface topped by an
large, dry
the bottom of the
beginning to heat and
An isentropic
Kansas
layer near
gave way
14 GMT (Fig.3. 16),
adiabatic
layer.
layer near the
inversion was
By
surface
lower and sharper.
trajectory at 309 K showed 30 mb oF subsidence
inversion air during this period.
(Fig. 3. 17),
had
large and
GMT the Dodge City,
inversion which
for
study.
just as Darkow
Sounding.s
At
was
their
but
very
little had
produced a deep,
bu the inversion.
changed
well-mixed
This trend
By
except
boundary
17 GMT
that the strong DLH
layer,
conrtinued through
still capped
the next three
PAGE 77
hours.
and
20 OMT the
By
likely
contributed
The PBL no
dry
earlier
longer
to
hailstorm was
the
shown
structure
was well-mixed
just to
in
q,
although
Goodland.. Kansas
(Fig. 3. 19)
a
(OLD) soundings.
showed a
strong
During
Fig. 3. 18.
it
still
was
inversion,
with a
the next three hours
while the
Surface
7
11
GMT OLD sounding
large dry adiabatic
strong
sharpened
layer near the ground began to heat.
a huge,
layer wit#hout an inversion,
advection was calculated
at ea-ch:
Tour.
layer aloft.
subsidence, as
the inversion and
(Fig.3.20), the GLD sounding had
boundary
The
seen in the
shallow moist layer near the ground capped
trajectory analysis,
V
south,
adiabatic.
Another example of the BLH changes was
by
in
the
by
shown by
lowered
it,
By 20 GMT
dry adiabatic
due mostly to BLH.
computing
Tfere was very
little
V'-V
and
temperature
advection, but the moisture
drop at GLD had an advective
contribution.
there was dry advection at GLD.
On Fig. 3. 11.
Modelling was used to confirm this explanation of PBL
development.
The
initial
The 20 GMT model result is
sounding was the
parameters were 207
and 50%
soil moisture respectively.
of
the drying actually
very
large PBL growth.
shown on Fig. 3.21.
11 GMT OLD sounding,
surface
soil
The model
observed
The model
moisture and
did
although
showed
and the soil
bulk
not reproduce all
it did
duplicate the
a dewpoint of about
7
PAGE 78
deg C , while the 20 GMT observation was about 3 deg
advection must have been responsible for
C.
Dry
the rest of the
dr y ing.
Notice that
this sounding was
moist convection.
The sounding at Concordia, Kansas
on the cool 'side of the gradient.
showed a deep moist
adiabatic
GMT
layer above
layer aloft, with no
(not shown) CNK had
the dry adiabatic
upward and
too dry to support any
At
(Fig.3:22) it
the surface and
inversion
a subsidence
layer.
11 GMT
During
some downward vertical
By
14
inversion below
six hours, some
motion could
trajectories but all of modest amounts
a dry
in between.
induced
the next
(CNK) was
be seen on
(less than 10 mb
in
three hours).
The 20 G~T sound i-rng
bourndary latyer emerging
well-mixed
saturated
layer above.
This
incident radiation through
boundary
layer.
hypothesis.
reasonable numbers
under
the still nearly
sounding never got
enough
the clouds to sufficiently heat the
Modeling of
The soil
small
(Fig. 3.23) showed a very
this sounding confirmed
parameters used were 50% and
considering that the rainfall
50%,
of the
previous day was only between .02 inches and a trace
The model
was cloudy
run produced
all day)
and
saturation at the top of
the cloud
near CNK.
the PBL
(it
the PBL moisture stayed very high,
with a surface dewpoint of
16-17 deg C.
The
end
result for
PAGE 79
CNK was
a cool,
strongly
capped-PBL,
and no convection
(see
Fig. 3. 24).
None of the
soundings were
convection actually broke
out.
between GLD and CNK, and north
surface potential
Fig. 3.25 for
was in
in
The squall
of DDC.
the period
12 to 22 GMT.
of strong
heating
experienced very
developed
field of change of
is shown on
The area of
outbreak
(not the strongest)
the area of moderate drop
convective region
line
The
temperature and dewpoint
the region
was also
in the region where the
in dewpoint.
The
little advection.
Calculations were made from hourly analyses at each
outbreak
area:
Hill City, Kansas
appear on Table 3. 1.
2 K
to the change
in potential
this case.
activity was weak,
reports
temperature
was weak
unlikely.
As
day.
shown, synoptic-scale
setting were examined
the region,
so that symmetric
The most likely
As
already noted,
surface analyses
The results
from
trigger was an
frontal
The station
for gravity wave
found.
Wind shear
instability was
inland
across the remarkable temperature gradient which
the
of the
11 K.
be eliminated
but none of any consequence was
in
of
and remained to the north.
of altimeter
activity,
(HLC) and DDC.
end
The advection contributed between 0 and
Many possible triggers could
consideration in
but
sea breeze,
formed
during
some convergence was seen on the
(see Fig. 3. 13),
although
it was weak
and did
PAGE 80
not seem to change very much with
(or any
as the
clear
sea breeze)
time.
circulation should
temperature gradient sharpens.
source for
the convergence.
temperature gradient would,
require an
inland
increase
in
The creation of
through
sea breeze
intensity
There was no
thermal
other
the
wind arguments,
increase in the vertical shear.
breeze circulation would have satisfied
3. 5:
An
An
inland
sea
this requirement.
Hybrid Modelling
To test
the impo.rtance--of various
creation of convective instability and
a hybrid
factors affecting the
convective
sounding representing the outbreak area was
constructed
from nearby
data.
This construction was made for
11 GMT.
Surface analysis was used for the
and
isentro-pic and
both
data.
inhibition,
isobaric
Isobaric analyses were
surface conditions,
analyses for the
upper air
and 700 mb.
done at 500 mb
Most
soundings from the eastern side of the network showed a nearly
dry adiabatic
313 and 311
layer
yielded
Analyses were also
above 500 mb
were
between 311 and 313 K.
So analyses
the top and bottom of this
for
layer.
done at theta = 309, 305 and 301 K.
Data
interpolated from the three nearest
soundings:
CNK, DDC, and GLD.
(APRHYB) is
shown
.'his
It was used to
in Fig.3.26.
significance of various factors,
important aspect of the hybrid
11 GMT hybrid
as discussed
sounding was
sounding
investigate the
below.
The most
the prominent
PAGE 81
inversion at 750 mb,
thick, almost dru
which represented
adiabatic
layer.
This
be a barrier to growth o2 the PBL
above was responsible for the
instability.
The other
was that the lower
with cloud.
the bottom of the
and
inversion proved
the adiabatic
large potential
interesting aspect
layer
for convective
of
this sounding
part of the sounding was very wet,
This cloud
layer was
to
significant
filled
in the time
evolution of the PBL.
The bulk
and
soil
moisture was fixed
error to some extent,
over the past month.
small,
hence
be near
as well
as
This rainfall
the expected bulk
for the model
by
by considering
trial
rainfall
was not unusually
large or
soil moisture was expected
50% and no higher than 80X
integrations were perform-d with
at
the most extreme.
both values, with
expectation that the 50% bulk moisture would
realistic results umiiss the soil
proved
to
The
the
give more
to have very poor
drainage.
The
way
initial surface
but using
The map
soil moisture was
the previous
for Kansas is shown
region around GLD was
dry.
to HLC was generall- much
to more than .50
soil
day's
moisture
inches
in this
set
in a similar
(April 18) rainfall
in Fig.3. 27.
The area
Notice
in some places.
second area was
that the
stretching north
wetter, with amounts
pattern.
from DDC
over . 10 inches
Clearly,
the
initial
higher than that near
PAGE 82
GLD.
The rainfall
very dry
in an absolute
statements,
numbers
sense,
that the
for the
initial
therefore used
The model
but beyond
in
was run first without any clouds or changes
the PBL to establish th-e
importance..of
initial
Several aspects of the model
Table 3. 2,
were
the various
soil
surface
behavior, shown on
in this series of runs, there
interesting.
direct relationship between the final height
surface temperature-, and
the PBL was
top
A range of
the modelling.
moisture.
top,
these general
soil moisture.
combinations of bulk moisture and
was a
soil would not be
it was difficult to be more precise about what
to use
values was
above
also implied
(higher surface temperature),
grew and
understood.
surface moisture.
the
dryqer the PBL was.
The r-ise o-f
the air.
The warmer
the higher the PBL
This could
be simply
the PBL top depended upon the
sensible heat f-lu-x from the ground,
for heating
of the PBL
which
was also responsible
So warmer surface temperatures were
correlated with deep-er PBLs
other things staying
the same.
The PBL top rose by entraining air from above the, inversion.
into
the PBL.
The initial hybrid sounding was dryer with
increasing height from the
atmosphere.
surface to the top
of the
So as the PBL grew, dryer air was entrained and
hence the PBL dried
The deepest any
out.
of
the PBLs grew was 751 mb;
none of the
PAGE 83
runs
could
penetrate the
Fig. 3. 28 shows,
the rate of rise tended
almost to zero.
heat flux and
inversion at 750 mb.
This was
to drop
due to both the
the entrainment
Indeed as
off by 21 GMT,
decrease in
of stable a'ir
sensible
from the
inversion.
The model
top
predicted saturation or oversaturation at
of the PBL. after 16-17 GMT in all
runs, which
could
but
the dryest of the
have happened for two reasons.
model assumes a perfectly well-mixed PBL, with
ratio all the way
to the t-p.
especially with deep PBLs,
in
the model runs,
Often,
First,
in reality, and
q decreases with
height.
the saturation occurred with
A second possible reason for the
the PBL top ro-~e too quickly
However,
shallow PiLs,
at relatively
This would,
in these
saturation was that
cool
so that the saturation mixing ratio aloft was
available PBL moisture.
the
constant mixing
and was persistent enough to remain an inconsistency
runs.
the
temperatures,
too
in reality,
low for the
lead
to cloud
formation at the top of the PBL, which was presumed not to
happen in these runs.
in
judging
Hence,
the
inconsistency was
the realism of the results.
The PBL behavior was understandable
the various ground wetness parameters.
initial
the
important
soil
surface moisture,
drguest soil
50% bulk
had the warmest, dryest,
in relationship
The 20-50 run,
soil
to
(20%
moisture) with
and deepest PBL..
The
PAGE 84
wetter soils,
and
50-50 and
50-80 had successively
shallower PBLs.
The
instability
measured by PLI
values behaved
relationship to the ground wetness as well.
50-80, had
had
the
lowest.
Generally,
effect overcame
the PLI
essentially controlled
rose in response to both
in these runs,
by
the soil
and
The drying
the PLI was
For these runs,
moisture.
the negative area calculation followed
the same pattern as the
The most stable run also had the most negative
values.
area,
dryest run, 20-50,
higher moisture values.
the heating
in direct
The coolest run,
the highest PLI value, while the
warmer temperatures and
PLI
cooler, wetter,
by more than 20%.
Since no
changes were allowed
to
occur above the PBL during these three runs, the only ways the
negative area cou-ld
a
increase were for the PBL parcel
"colder" psqtrdo-adiabat,
leaving more otravel
through.
the
(The parcels, when lifted
equivalent
So by
for the parcel
to
for the PLI
line of constant equivalent potential
"colder"' pseudo-adiabat we mean a smaller
potential temperature.)
pseudo-adiabats
for the wetter soil
shallower PBLs,
compared
slightly more instability
shallower,
to be shallower,
the PBL
stable air above
calculation, follow a
temperature.
or for the PBL
to follow
In
these runs, the
runs more than
to the 20-50 run.
than the
giving a slightly
"warmer"
offset the
The 50-80 run had
50-50 run,
but was
larger negative area.
PAGE 85
When clouds
model,
(as
without any
shown on Table 3.3) were added
other changes,
but not in
still
in relation to the soil
clouds were present.
nearly as high
qualitative terms.
The PBL grew as
by 21 GMT as
Guantitatively,
are shown
in Table 3.2.
10%.
almost exactly the
surface and
before and reached
to
soil
6%, and the
moisture, adding
surface temperature
the 50-50 run without clouds had
same surface temperature and PBL depth as
The difference was only 0. 1 deg C
the 20-50 run with clouds.
at the
just as when no
The results for cloud runs
For the same
Notice that
moisture,
the PBL
clouds behaved similarly
soil.
clouds cut the PBL growth by
rise by
That is,
in the cloudless runs.
the runs with
cloudless runs with wetter
in
the results changed
particulars,
heated
to the
I
mb
in depth.
The PBL m-i-stur-es for the runs with clouds were, however,
only marginally bigger than for the same runs without clouds.
The drop
in GS was reduced
by
7%
for the cloudy runs.
instance, the 20-50 cloudy run had a PBL moisture of
while the cloudless 20-50 run showed. 10.0 g/kg.
explanation for this behavior
cloudless runs was only a
used when comparing
which
This
10. 1 g/kg
One
is that the entrainment for the
little greater than
(PBL depth 7-13 mb greater).
For
the cloud runs
But the same argument could be
the 20-50 and 50-50 cloudless runs in
the moisture difference was much more than 0. 1 g/kg.
suggests that the ground wetness which
is the .most
PAGE 86
important factor
in determining
major controlling
The initial
the
influence
behavior
subsequent moisture
moisture.
was low,
lower
the
latent heat
of
layers.
raising
of
period,
surface moisture changed
increased
soil
dropped
heat flux
the cloudy
less quickly.)
the P1L was growing
little.
the cloudless case,
than the
slowly
so the
When the clouds broke
moisture available.
were slightly wetter
the latent
the 50-50 runs,
50-50 run's surface soil moisture
the cycle began as with
the surface soil
soil moisture rose due to flux from
(In the case
During this cloudy
the
the runs with clouds determined
behavior by
the surface
was
on the PBL moisture.
When the clouds were present,
and
flux,
up,
but with an
Hence the cloudy
cloudless ones for
cases
the same soil
moisture.
The PLI
The dryer
showed
run was
the
the greatest
runs all
in
beh-ave d
a
similar way
least
unstable,
instability.
The
increase
when the clouds were added.
wetter and
while the wettest run
Notice,
showed less instability
counterparts.
as for the plain runs.
in
instability
the
was reduced by
cloudy
the additional
10%
runs were
heating
cloudless runs was enough to give bigger values
instabilitu.
that the
than their cloudless
Although
had shallower PDLs,
however,
of
in the
convective
PAGE 87
The inhibition change
between the cloudless and
runs was variable, depending on soil moisture.
soil,
adding clouds more than
soil was only greater
wettest
doubled
by
showed more than a 25% increase
reasons for the
had
increased
a smaller PLI and
increased
the
PBL had
There were two
cloudy 50-50 run
a shallower PBL, a:
the two runs.
the smallest
Both
of these
Between the cloudy runs
for the cloudless
and the
inhibition.
about the same
clouds reduced
grew too
still
cloudy 50-50 run
The
last point concerning these runs
showed satura-tion by
The
e followed a "colder"
theor
the neg.ative area.
most unstable run had
Although
The
stable run had the most necative area,
The most
The
the negative area.
the behavior was the same as
themselves,
For the dryest
inhibition.
inhibition.
difference of nine mb between
runs.
in
Also, the cloudy
pseudo-adiabat.
effects
11%.
cloudy
is that all of them
time as
the growth of
in the plain runs.
the PBL, the PBL
quickly for the amount of heating,
the top
and
became saturated
or oversaturated at the time when the clouds
were expected to
dissipate.
It was somewhat surprising that
the cloudy runs were
The
clouds
similar in
some respects to the cloudless runs.
which were
added were meant to simulate the actual behavior
the
outbreak region during
clouds
broke up
completely
the day.
by
As noted
16-18 GMTI,
already,
of
the low
and this behavior was
PAGE 88
incorporated
in
amounts were
set by referring
them to
the surface
The clouds were
and
the modeling with
thinner
at
clouds.
The actual cloud
to the soundings
and relating
observations of clouds when available.
thick
in the
14 GMT.
As
initial
sounding
the PBL heated,
(see Table 3.3)
the cloud
layers
which were entrained
into the PBL were assumed to dissipate,
although
itself could remain
the PBL top
This
the model
time variation
of cloudiness was a crucial
part of
behavior.
initial
energy
The
clouds affected
budget of the PBL for the whole day
major portion
the day.
of the heating
Additionally,
heating, only
halved
clouds present.
parameters
it,
and 3.31.
rise
in
slightly,
during
not
the PBL did
since the
the middle of
shut off all
heat even with
the
cloudy and
Notice that
the effect
present for the whole day.
cloudy for the whole
cloudless runs are
the net radiation
than half that for
of the PBL top was
shown
for
cut by
almost 50%.
only
The data
day.
The
in Figs. 3.30
the cloudy run
the cloudless run,
surface temperature was
case compared with
clouds.
run was made to document
for CNK, which was
results for
height
so
place
clouds did
for the 50-50 run with
clouds could have when
was less
the
took
only
the
Fig.3.29 shows the time variation of relevant
A related model
used was
cloudy.
and
that the
Similarly,
5.2 deg C for
the
the cloudy
8.2 deg C for the cloudless case.
PAGE 89
The next physical
changes imposed at
sounding,
the
the
the major
and
of subsidence at the
A
the DDC
and
changes
and appeared to be thle
result
in the middle
soundings at
11
and
of the atmosphere.
14 GMT are
illustrate the change.
50 mb at both
of
plotted
The
the 310 and 307 K
subsidence
isentropes,
also occurred at CNK and GLD between these two times.
similar
change occurred
This
using the
in
14 GMT,
at LBF between
effect was included
in a
The actual
changes
series of.integrations
The
results appear
incorporated in
sounding were +3
deg C at 775 mb and
with appropriate
drying of the moisture profile.
the PBL top
was 10-20 mb
the runs made without the
evolution
part of
is
surface
with
+2
the
deg
about
Notice
1530 GMT,
identical to the plain run
temperatures grew a
little
the hybrid
C at 800 mb along
By
21
GMT,
changes, compared with
inversion changes.
shown on Fig. 3. 33.
the run until
changes was
lower
A
14 and. 17 GMT.
same soil parameters as before.
Table 3.2.
In
at and
bottom (and perhaps throughout all)
together on Fig.3.32 to
was between 20 and
hybrid
DDC,GLDCNK and North
The most significant
dry-adiabatic layer
portion of
initial
was
significant changes occurred
inversion.
occurred between 11
large
the model
In the
specifically
Nebraska (LBF),
just below the
the
inversion.
in
inversion was most prominent at 750mb.
SESAME soundings,
Platte,
factor included
that
The
time
.for the early
the run with
in Fig.3.28.
faster,
by
inversion
The
about 4%.
The
PAGE 90
height of the PBL top began to be different by
and
the difference was
significant by 21
The biggest change appeared
By
1530 GMT.
for
the end
of the
the run with changes
run.
The
drying
in
about 1630 GMI,
GMT, almost 20 mb.
the PBL moisture,
integration
had dropped
33%
starting at
the surface mois
since the other
re
more than the plain
imposed above the PBL contributed
PBL through entrainment,
tu
to a dryer
parameters were
unchanged.
The PLIs varied with moisture
before.
the PLI
The differences
as
expected.
in PBL moisture had a big
The 20-50 run with
much more stable than its
growth
run,
of almost 30%.
but was
and ground wetness as
inversion changes was.
plain counterpart,
inversion changes.
between 80% and
showed
the PBL and
"cooler"
the
lessened convective
had
run.
The
difference
large change
the stability
above
Notice that the 50-50 run
than either the 20-50 run
effect of
between
This
increased
pseudo-adiabat which went with the
The 20-50 run had a
overcome the
increase in
instability.
less negative area
not
the
other aspect of
The values had
180% over the plain runs.
the effects of both
for the 50-80
10%.
The negative area changed more than any
the runs with
a reduction in
The difference was less
still more than
effect on
15 mb higher
a much
inversion which
"colder"
the 50-50 and
or the 50-80
could.
pseudo-adiabat.
50-80 runs was only
6 mb,
PAGE 91
but the
"warmer" pseudo-adiabat
for the 50-80 run could not
overcome the effect of a s.hallower PBL.
The other change which was apparent
that the saturation was much reduced.
inversion changes, none occurred.
happened
for two hours midway
Fig. 3. 33 for the 50-50 run).
to
the enhanced
In
the 20-50 run with
In the other two,
through
the
The change
integration
it only.
(see
in PBL moisture due
entrainment was enough to almost balance these
runs with respect to this
In addition
in these runs was
saturation condition.
to the changes at the
occurred above the
inversion
level
inversion,
changes
between 14 and
17 GMT.
Observations at DDC and GLD were used to determine appropriate
changes
to use for the m-odel
similar at DDC and GLD, but
GLD
sounding.
had differences in
the middle-atmosphere cooled
from 14 to
both
At DDC,- cooling
17 GMT.
These changes were
during
3.4 and
The two
sets of changes are
3.5, and were applied
11
only occurred
17 GMT, and was not as pronounced as at GLD.
the GLD pattern.
details.
At
to 14 GMT and
from 14 to
LBF also showed
shown in Tables
in separate runs to the hybrid
sounding.
The results
shown
of the runs with
in Table 3.6.
changes were
The most
the DDC
changes added
on are
important part of the DDC
those which occurred
between
14 and
17 GlMT near
PAGE 92
775 mb and
800 mb.
When DDC showed cooling
accompanied
by moistening as well.
entrainment
in the PBL was lessened,
So
there,
the effect
and
the amount of drying was
the plain runs.
more rise
by
1% to
Surface
The PLI
similar
changes only.
up
10% to
in. moisture
The
convective
computed
no
than
this series
the
of runs,
and
and near the values
This change in
with
the 50-80 run.
the cooling above
the
the DDC run.
For
the 20-50 run was the most stable and had
due entirely
50-80 runs exhibited the
of the previous runs,
However,
less than half that
"warmer" pseudo-adiabat of
least negative area,
The 50-50
for the 50-50 run.
inhibition was
the negative area was due to both
the
for the 50-80 run
the PLI
that
than those with
f-or -the plain model without clouds.
inversion and
5%
longer overcame the heating
inversion changes,
for the run wi.th
from
in moisture
15% higher
Curiously,
with DDC changes was now lower
difference.
to 30%
the PBL growth was reduced
s.trongly to this change
responded
The difference
by 10%
inversion
inversion change only runs.
to the
with the DDC PLIs ending
inversion
reduced
only
temperatures were about the same,
than the plain runs and
10%,
of
the PBL moistures in
the DDC runs were- much higher than those with
changes.
it was
following
the
to the higher PBL tops.
same behavior as most
trend
in the PLI values.
this series the 50-50 run had a higher PLI
The extra moisture also gave a
of saturation at the top
of
the PBL in
than
longer period
the 50-50 and
50-80
PAOE 93
runs.
With DDC
hours
instead
of two
The model
to
changes
this conditi.bn was
(see Fig. 3. 34).
was next run with
the inversion changes.
the same as
results
the GLD changes
except the GLD changes
GMT and
continued
for the GLD runs are
shown
till
in Table 3.6.
development was that the 20-50 run
growth.
In fact,
the PBL heated
inversion at 775 mb,
layer aloft.
moisture
enough
and reached
up
low, due to the entrainment
did not
775 mb
inversion.
almost
duplicates of the
The most
showed
so much
to bypass
into the dry adiabatic
very
of very
heat up as much and
As a result,
The
this run
in
The surface temperature was
The wetter runs
began with
17 GMT.
striking
the
in addition
The sense of the GLD changes was
the DDC ones,
cooling aloft at 11
present for three
high, and
dry
air aloft.
could not
the 50-50 and
fill the
50-80 runs were
50-50 and 50-80 runs with
inversion
changes only.
This
duplication was seen
50-50 runs were
Figs. 3. 33 and
identical,
3.35).
saturated at 21 GMT, and
just as
negative
50-50 and
even in
The 20-50 run
than the wetter soil runs.
Both the
in the PLI
values
too.
time sequence
(see
had much
less
The
instabilit
The 20-50 run's PBL top was
there was no neative area
50-80 runs were
their counterparts with
oversaturated
left.
for
two hours
inversion changes onll.
area for the 50-50 run was 20%
lower than
just
The
the plain
PAGE 94
50-50 without clouds.
stability above the
inversion.
value were identical
with
This low value reflected the
Both
low
the PBL depth and the PLI
for the 50-50 GLD run and the 50-50 run
inversion changes.
On Fig. 3. 36 the sounding for the hybrid 50-50 run with
GLD changes at 21 GMT is plotted.
Notice that the layer from
775 mb to 750 mb was superadiabatic.
having
imposed
changes at 750 mb and
layer cooled,
change at 775 mb.
inversion between
so that
much
indirect
17 GMT.
it made sense to include a
14 and 17 GMT.
of the
The PBL temperature at
to have enabled this
inversion itself.
ev-idence- to support
This was
the whole
Additionally, the GLD PBL grew past its
GMT at GLD was not high enough
without cooling
from
not having any
In reality,
corresponding changes at 775 mb.
adiabatic
This resulted
incorporated
to occur
Hence, there was
cooling at 775 mb
in a
17
from 14 to
later run.
The next series of runs added clouds to the DDC and GLD
changes,
to combine
these effects.
For the DDC runs, -shown on
Table 3.6, the addition of clouds made a small but
difference
slightly
grew 6%
in the development
cooler and wetter,
to
16%
and
by
10%
The PBLs were
slightly shallower.
The PBLs
less than the plain runs, but the surface
temperatures were nearly the
less,
of the PBL.
important
to 30% depending
same.
The PBL moisture
on soil moisture.
dropped
The 50-50 and
PAGE 95
50-80 runs developed more slowly
saturated for two hours
time, and were only
instead of three.
to the reduced heating.
At
cloudy 50-50 run was 20 mb
run. and
with
This effect was due
1600 GMT, the PBL top
for the
lower than for the cloudless 50-50
the surface temperature more than 0.5 C cooler (see
Figs. 3.34 and 3.37).
Just as for the plain runs,
the cloudy runs all showed
less instability than the cloudless runs, although
differences were small.
Notice by
comparison of Figs. 3.34 and
3.37 that the time evolution of the PLI
fell
behind the
the
for the cloudy case
cloudless DDC 50-50 run at about 16 GMT and
never quite ca-ught
up.
The negative area was about 40%
higher
than the comparable plain runs.
The GLD run with
cloudless counterparts
totally
different
series.
runs
,
differen.t from its
than the DDC runs.
The 20-50 run was
but this run was an anomally
in this
Close e-xamination of the output showed that
(cloudless and
actually
clouds was more
cloudy 20-50 with GLD changes) were
similar, but separated by
time.
The values of
temperature, moisture and PBL depth at 21 GMT with
virtually
Even
clouds were
identical to those of the cloudless run at 20 GMT.
the PLIs were almost the same,
cloudiness delayed the heating cycle
dvelopment
the two
back one hour.
4.0 and 4. 1.
by
The
enough to set the PBL
PAGE 96
The 50-50 and
at 21 GMT
50-80 cloudy GLD runs were less
than the 20-50 run.
dropped 20% more than
the 25% to
changes.
the
110%
increase for
almost
development was almost
C different.
similar PLI
about 2
by
values.
10 mb.
However,
the same, and
5%, so that
lower for the
the instability
the
final values only
just wet enough
0.1
to have
The cloudy PBLs reached saturation for
but delayed by
just as the cloudless runs did.
The saturation was, however,
the presence of clouds in
inconsistency.
imposed
imposed changes were warmer than the
The cooler PBLs were
hours,
one hour.
grew less,
The heights of the PBL tops were
cases by
compared with
the cloudless runs with
temperatures
cloudless runs with
cloudy
cloudy PBL's moisture
the plain runs'moisture,
The surface
plain runs.
The
different
the model,
nearly consistent with
so this was not an
Th-e negative area for
the
50-50 cloudy GLD run
was
higher tha-n t-he GLD clear runs, but much
for
the 50-50 p-ain
GLD
clear runs had
run.
The 50-50 GLD cloudy run
similar PLI values,
was responsible for the
smaller
than that
and 50-50
so the shallower PBL
larger inhibition
(compare Figs.3.35
and 3.38).
The final
series of runs
clouds and
inversion changes,
remove the
superadiabatic
used the hybrid
plus GLD changes modified to
layer above the
modification used was a cooling of
addition to the GLD changes
sounding with
shown
inversion.
1. 5 C at 775 mb,
in Table 3. 5.
The
The
in
results
PAGE 97
of the runs are shown in Table 3.6.
especially with
the
the 50-80 run.
50-80 run was
The
height of the PBL top for
only three mb higher than for the comparable
run without modification.
same and
The effects were small,
The surface temperature
the PBL moisture only slightly
differences were greater for
20.-50 run was not greatly
reduced.
was the
The
the dryer runs, but even the
different.
The 20-50 modified run's
PBL was 19 mb deeper, 0. 1 C warmer, and 0.4 g/kg dryer.
The PLI
values for the modified runs were all
than those of the unmodi-f-ied runs,
the 50-80 run.
though only
The modification had only
these parameters due to the presence of
layer
between 775 mb and 800 mb.
smaller
slightly so for
a slight effect on
less stable air
in the
This difference was only
realized by the model when the PBL top reached high. enough
entrain thisair.
So, the
the runs, and especially
change was only noticeable
negative area for the 50-50 run was
clear,
from the plain run.
the
to the change..
The
lowest of any run and
The reason for this was
since the inversion strength was much reduced
modified run, and the
late in
late in the 50-80 run.
The negative area responded strongly
lower by 41%
to
instability was only
slightly
in the
lower.
Figs. 3.39 and 3.40 show the 21 GMT soundings for the modified
and
unmodified runs, with
the negative area indicated
on them.
PAGE 98
3.6:
Summary
The final
since
series was the most realistic set of runs,
it included clouds,
initial
changes above the PBL including
superadiabatic
layer.
the period,
have been too far
The pattern
the reduction of the
Since the winds above the PBL were from
the southwest throughout
DDC would
inversion changes, and the
south
any advective changes at
to reach
the outbreak
area.
of subsidence affected the CNK,GLDsand DDC
soundings at the
same time, and
implying movement from the
subsequently
moved to LBF,
southwest as well.
Hence the GLD
changes were more representative of the hybrid area than the
DDC
changes.
GLD modified
The surfac-e temperature and moisture from the
runs were comparable
analyses as well.'
meter
to the observations from the
The model surface
observa-tiCn temperature;
layer was superadiabatic.
since
The model
temperature was not a
the 5 mb deep
output did
observation tempera-ture calculation which
gave,
surface
include an
for the GLD
50-50 modified run, a surface potential temperature
The dewpoint in the model run was 13. 5 deg C.
observations for
C.
little cold
the analysis
in the model,
showed, 0-2 K
outbreak region
advection.
of 307 K.
The actual
the region of the outbreak were 309 K and
The dewpoints were very close.
The
10
14
temperature was a
but as the advection calculation
in
of the observed rise in the
could be attributed
to horizontal
This would raise the model
temperature
temperature
to 309 K,
PAGE 99
the observed value.
Some statements
factors involved
and
can be made concerning the various
in the development of convective instability
negative area
in this case.
temperature rise in
implied this, and
the model
results confirmed
in determining
affected the PLI values.
the observed
The observations
it.
On Table
for the various parameters are
As the modeling showed,
important role
most of
the PBL was due to BLH.
3.7 the sensitivity values
shown.
Clearly,
the ground wetness played an
the PBL moisture, which
strongly
The wetter runs, although cooler,
were often more-unstable.
However, the wetter runs had larger
negative areas, since the heights of the PBL tops were lower.
The PB3L depth was l ower when clouds were present and for
the
inversion only
changes.
The changes aloft also
contributed to shrallower PBLs except
inversion was filled
in the case when
(20-50 run with GLD changes).
the
Aside from
that one run, the changes in growth were not large either, at
most
17% for
the modified GLD runs.
Soil
bulk moisture variations were similarly
much
input
surface moisture and
ineffective in having
on PBL growth.
Surface
presence of
surface soil
layer temperature was
clouds reduced
even less sensitive.
the rise by
moisture reduced
10%, and the
the rise by
The
increased
11%, but the rest of
PAGE 100
the variables had
less influence.
The PBL moisture values
were much more sensitive to all of the
soil moisture and
(both at
clouds.
The
factors except the bulk
imposed changes above the PBL
and above the inversion) contributed to
drying while the increased
soil
increased
surface moisture reduced the
drying significantly.
The convective instability
in most
of the runs.
When drying
values was reduced, often
by a
inversion changes series).
moisture had
convective
a large
varied with the soil moisture
increased,
significant amount
Similarly,
increased
positive effect on the
instability.
growth of PLI
The biggest
instability was already present
(e.g.the
soil
surface
growth of
part of the convective
in the
12 GMT sounding, with a
PLI of 3.0.
The inhibit-ion to convection was the most variable
parameter. and
soil
sensitive to all
moisture.
In some cases,
of the factors, even the bulk
the negative area responded
different ways to the changes depending on
characteristics.
For
increase generally
cloudy runs,
gave an
increase
however, the dryest
negative area.
inhibition,
instance, the soil
as
changes aloft
In general,
did the
individual
surface moisture
in negative area.
soil run
had the
the presence of clouds
imposed
(DDC or OLD runs)
in
For the
largest
increased
changes at the inversion.
tended to strongly
The
decrease
PAGE i01
convective inhibition, with
We can realistically
factor best by
run which
the runs with
in this manner.
Schematically,
decreasing
that factor
in the
This allows the
in the
the three case
(The other factors in each
the other case
case
studies.)
the 50-50 model run with
changes can be summarized as
follows in
order of importance.
PLI = initial
caond-itions + BLH +
changes + clouds + bulk
NA = initial
soil
surface moisture -
soil moisture
surface moisture -
conditions - BLH + soil
imposed change-s + clouds + bulk
soil moisture The
effects of
clouds are reversed when considered in this manner.
presence of
clouds
physical
imposed changes and clouds are
the results for
and GLD imposed
of each
of factors to be present
are not included in
imposed
the change in
To allow comparison between
studies, only
clouds
the effects
includes all of the factors.
comparison.
study
compare
considering
non-linear combinations
examined
some exceptions.
clouds had led
to lower PLI,
to the run with imposed
The presence of morning
The
while the addition of
changes gave more instability.
cloudiness exerted
a strongly
non-linear effect on the growth of convective
instability,
PAGE
It was
curious and
intriguing
that
most realistic results also had the
(for
This was not conclusive,
the
the run which gave the
smallest negative area,
the 50-50 soil moisture) although
unstable.
102
it was not the most
it was suggestive that
but
out where there was substantial
convection broke
convective ins tability and where the inhibition was the
weakest.
convective
The
and analyzed at
18 and 21 GMT.
Onig the new echoes
and 3.42.
plotted
instability and
on each
figure.
inhibition were
shown on Figs. 3.41
These are
see that
the
in an area where NA was a minimum and
convection
broke out
convective
instability a relative maximum.
not analyzed
time are
from the next map
On Fig.3.41, we
calculated
but the correlation
in detail
This outbreak
in
was
this instance is
illustrative of the relation between convective
outbreaks, NA
and PLI.
Fig. 3.42 shows the outbreak which
convection clearly
detail.
The
moderate
instability,
Unfortunately,
are
in
in
in a region of
but also a minimum of NA.
both of these
figures, the available data
too sparse to allow precision
However, coupled with the model
run
erupted
has been analyzed
in these conclusions.
results for the
"simulation"
(50-50 run with modified GLD changes and morning clouds)
the pattern
of outbreak
is well
estab'lished
in this case.
The
PAGE
convection began where and when moderate
instability coincided
to start the
levels of convective
,.ith low enough values of
allow the available forcing
convection.
103
inhibition to
(surface convergence on Fig.3. 13)
PAGE 104
Table 3. 1:
ADVECTION CALCULATION FOR APRIL CASE
DDC
HLC
Time
Adv.
Obs.
12-13
13-14
14-15
15-16
16-17
17-18
18-19
19-20
20-21
0
O0
<0
<O
<0
(0
-CO
0
0.
Total
0
<C
2.
2.
1.
1.
-0.
9. a
Adv.
<-0. 2
0
0
0. 1
0. 4
0. 7
0. 5
0. 4
0. 2
Obs.
unknown
1.0
2. 4
1.7
2. 4
1. 1
C-
1. 7
0
2. 3
11. 5
Advection calculated
at Dodge
City,
Kansas
(DDC)
and
Hill
City,
Kansas
(HLC)
from
surface analuses.
Changes are for
surface potential temperature in degrees K.
Time is GMT.
PAGE 105
Table 3.2:MODEL
RUN
RESULTS AT
PH
21 GMT,
19 APRIL
QS
PLI
NA
Cond
Plain
20-50 run
50-50 run
50-80 run
751
759
764
26. 0
25. 0
24. 5
10.0
10.9
11.1
5. 3
6.0
6. 1
20. 56
27. 39
28. 94
S+
S+
758
768
777
25. 1
24. 1
23. 5
10. 1
11.0
11.32
5. 0
5.7
5.8
42. 79
34.66
32. 16
S
S+
S+
26. 5
9. 1
10. 2
10.7
4. 4
5.3
5.7
57.67
50.08
60. 82
Morning Clouds
20-50 run
50-50 run
50-80 run
Inversion Changes Only
20-50 run
50-50 run
50-80 run
762
777
733
25. 4
24.8
21 GMT values -of :
pressure level of inversion (PH),
surface
temperature
(TS),
surface
moisture
(OS),
PBL lifted index
(PLI), negative-area- calculation (NA), and
condition
at
PBL
S
=
S
=
nearly
saturated,
=
unsaturated,1
top:
blank
saturated, S- = oversatura-ted.
PAGE 106
Table 3. 3:
CLOUDS
IMPOSED IN MODEL RUNS FOR APRIL CASE
14-17 GMT
11-14 GMT
P(mb)
800
825
850
875
900
80 %
amounts
Cloud
cloudc over.
are
0%
80 %
80 %
80 %
80 %
90 %
80 %
SO %.
expressed
as
percentages
of
complete
PAGE 107
Table
3. 4: IMPOSED CHANGES
(INCLUDES INVERSION
FROM DODGE CITY. KANSAS
CHANGES) FPR 19 APRIL
11-14 GMT
P(mrb)
T(deg
14-17 GlOMT
C)
T(deg
C)
0
525
550
575
600
625
650
675
700
725
750
775
800
825
C
0
0
0
(DDC )
O(gikg)
O
0
0
0
0
0
0
+0. 6
+0. 6
+3. 0
+0.
+0.
+0.
+0.
0
-0. 9
-0. 9
-0. 6
+2. 0
0
-3.
-8.
O0
+0. 9
+0. 9
+1. 2
+1. 5
+1.8
00.
0
-0. 3
0
0
Changes taken fprom Dodge City, Kansas (DDC)
soundings used
in
some
model
runs.
Levels not mentioned or.times not covered
had zero changes.
PAGE 108
Table 3. 5: IIMPOSED CHANGES FRk(M GOODLAND,
KANSAS
(INCLUDES INVERSION CHANGES) FOR 19 APRIL
P(mb)
500
525
550
575
11-14
T(deg C)
T(deg
14-17 OMT
Q(g/kg)
C)
-1..
-0. 6
-i.0
-0.0. 6
625
650
675
700
725
750
775
800
825
GM"
Q(g/kg)
.1 2
-1.2
-0. 8
-0.
+0.
0
0
-1.
0
-3.
+2.0
0
0
0
-0.
-1.
0Oi
-8.
0
0
Changes taken from Goodland, Kansas (GLD) soundings imposed on
Levels not mentioned or times not covered
some
model
runs.
had zero changes.
PAGE 109
Table 3.6:
RUN
MODEL RESULTS AT 21
GMT,
19 APRIL
PLI
NA
9. 4
10. 7
10.9
4.9
6.0
5.9
16. 15
19.42
31.83
7.8
10.2
10.6
2.9
5.3
5. 5
O
22. 12
24.63
PH
Con d
Inversion Changes + DDC Changes
20-50 run
50-50 run
50-80 run
752
773
780
26. 7
25. 4
24. 8
Inversion Changes + GLD Changes
20-50 run
50-50 run
50-80 run
694
778
793
27. 0
25. 3
24. 7
Inversion Changes + DDC Changes + Morning Clouds
20-50 run
50-50 run
50-80 run
760
777
789
26. 3
25. 1
24. 1
9. 5
10. 8
11.0
4.8
5.9
5. 7
29. 3
31. 16
39. 59
Inversion Changes + GLD Changes + Morning Clouds
20-50 run
50-50 run
50-80 run
776
786
792
256
24. 5
24. 0
9. 4
10. 6
10. 8
4.4
5.4
5.4
10. 84
24. 60
27. 33
Inversion Changes + Modified GLD Changes + Morning Clouds
20-50 run
50-50 run
50-80 run
This
table is
757
780
789
in
25. 7
24. 6
24.0
9.0
10. 4
10. 7
3.9
5.2
5.3
the same format as Table 3.2.
10.02
16.09
18.67
PAGE 110
Table
3. 7:
Physical Factor
SENSITIVITY VALUES FOR
MODEL RUNS
21
GMT , 19 APRIL
PH
PLI
NA
-6
-10
+11 to
+108
Morning Clouds
Inversion Changes
-10
Inversion Changes + DDC
-1 to
-10
Inversion Changes + GLD
-12 to
+36
Inversion Changes + DDC
-6 to
-16
Inversion Changes + GLD
-17
+33
Changes
+5
+10
+30
+83 to
+180
to
Changes
+3 to
+25 to
+11
+110
Changes + Morning
+3 to
+5 to
-4
+25
Changes + Morning
-5
+20
-6
-19 to
-77
Clouds
-11
+10
-21
to
-15 to
-100
+40
Clouds
-24
-6 to
-42
Inversion Changes + Modified GLD Changes + Morning Clouds
-4 to
-4
+32
-32
-35 to
- -6
-51
Soil surface mo-istur-e
-7
-I!
-31
+23 to
-22 to
+77
+61
Bulk soil moisture
-4
-15
+3 to
-8 to
-3
+21
Sensitivity measured as percentage change of a given
variable
compared
with maximum amount o- change in that variable after
application of physical parameter.
Variables
as
defined
on
TABLE
3.2.
Application
of soil surface moisture defined as
increase from 20% to 50% of saturation.
Bulk
soil
moisture
application defined as increase from 50% to 80%.
*
r44F
on
*
SEP
1
F
vO-1
Fig. 3. 1
Map of SESAME region showing sounding stations in
April, and two surface observation stations mentioned
later in the text.
I-
-70
Fig. 3.2
Synoptic-scale 500 mb analysis for 12 GMT, 19 April.
Heights in solid lines (dm) and vorticity in dashed
lines (x 10**-5 sec*-1).
h0
44
Same as Fig.3.2
Fig.3.3
Fig3 3
Sae
s
MT,
for
ig3.2fo
#01
0
20
g
.07 April.
GT,20Apil
TO
Fig. 3. 4
Synoptic-scale surface analysis for 12 GMT, 19 April.
Sea level pressure in solid lines (mb) with leading 5
or 10 dropped and 1000 to 500 mb thickness in dashed
lines (dm).
Fig. 3. 5
Same as Fig. 3.4 for 00 GMT, 20 April.
72 2 2
20
/"
154*1,
10
0:
0'
0:
0:
O:
O
GCK
2102
Fig. 3. 6
Photograph of low-elevation angle display from radar i
screen at Garden City, Kansas, 2102 GMT, 19 April.
Range rings are 20 nm apart.
22 24 2
20
.ga
6
• . .'. 85
-S
O:
0:
0:
O0
0.
GCK
2122
IuS
Fig. 3. 7
Same as Fig.3.6 for 2122 GMT.
G 5
PAGE 117
24
20\
-1 1 1to4
GCK
2140
•q I -
Fig. 3. 8
-8
Same as Fig. 3.6 for 2140 GMT.
24
5
.15
0
000 0
0
00 G0 0
0
0
GCK
2200
1-01 S9
Fig. 3. 9
Same as Fig. 3.6 for 2200 GMT.
PAGE
Fig. 3. 10
118
Mesoscale surface analysis for 12 GMT, 19 April.
Solid lines are surface potential temperatures (K)
Winds are
and dashed lines are dewpoints (deg C).
Sky
plotted in the convential manner (knots).
condition is clear (open circle), scattered (single
bar), broken (double bar), overcast (filled circle);
Cloud type is plotted
and obscured (x in circle).
(if available) as is current weather according to
conventional synoptic code. Radar echoes are
cross-hatched irregularly shaped areas.
aqq
PAGE
119
/oc
Fig. 3. 11
Same as Fig.3.10 for
16 GMT.
a
a
a
3.'.
'-a
-j
*0
'0
-I
0
3..b
0'~
0
'-a
-j
0
-b
*0
.GI
-S
~0
a
at
0
ID
at
d.
g~4
I-'
a'
at
U'
3~8
U'
3.8.
'U
Ca)
'Sm.
a
0
B
00
r.l
-ZJ
§-.6
14
w,j!
c0i
U-
0
0
PAGE 12i
K'
I
/
0-
r-J,/
z
Fig. 3. 13
;3oo
Same as Fig.3.10 for 21 GMT.
PAGE
ABA
;gs
26 (D2
\
122
/
/290
J
2q2
4:
~---O a
0
-
j
Fig.3.14
Same as Fig. 3. 10 for 22 GMT.
PAGE
123
600
~IC~
++
4t.
700
800
-30
-40
Fig. 3. 15
-20
+
+
+
+
900
1000
+
20
-10
30
40
TEMP
Sounding plotted on a pseudoadiabatic diagram from
Dodge City, Kansas for 1115 GMT, 19 April. Solid
line connecting dots for temperature (deg C), solid
line connecting * for dewpoint (deg C), dash-dotted
line for 303 K isentrope and dotted line showing
moist adiabat for mean PBL parcel (or selected parcel
if PBL is not well-defined). Dewpoints colder than
-40 C are plotted at -40 C.
P
+
+
+
+
+
+--+\
.
+
+
4-.
+
4-
+
+
600
+
+
700
+
+
-80soo
+
+.
900
+
+
1000
-40
TEMP
Fig. 3. 16
Same as Fig. 3. 15 for 1415 GMT.
K.
Isentrope is for 313
PAGE
124
P
600
700
800
900
1000
i
-40
TEMP
Fig. 3. 17
Same as Fig. 3. 16 for
1715 GMT.
P
600
700
800
900
1000
-40
-30
-20
-10
0
10
20
30
TEMP
Fig. 3. 18
Same as Fig. 3. 16 for 2015 GMT.
40
PAGE
125
P.
600
700
800
900
looc
-40
-30
Fig. 3. 19
-20
-10
0
TEMP
10
20
30
40
Same as Fig. 3. 16 for Goodland, Kansas at 1124 GMT.
70(
800
900
1000
40
-40
TFMP
Fig. 3. 20
Same as Fig.3. 19 for 2007 GMT.
PAGE 126
P
+
+
+
+
+
+
+
+
+
+:K
+
+
+
+
+
+
+
+
+
+
+
+
S+
+
+
I
I .
-30
-20
-10
LI
-40
+
I
+
+
700
+
+
800
,I
0
600
-
900
+
1000
,
20
30
40
TEMP
Fig. 3. 21
Same as Fig. 3. 16 for model output at 20 GMT from GLD
initial sounding. No moist adiabat is plotted, and
data above- 600 mb is not shown.
P
-40
Fig. 3. 22
+ +
+
+
+
+
+
.t
+
+
+
+
+
+
+
+
+ ,
+
+
+
+
+
+
+
-20
-10
0
TEMP
10
20
30
-60
70C
+0
0
900
1001
40
Same as Fig.3. 16 for Concordia, Kansas at 1108 GMT.
PAGE 127
P
600
700
800
900
1000.
-40
40
TEMP
Fig-. 3. 23-
Same as Fig. 3. 22.
or 2008 GMT.
P
600
700
800
900
1000.
)
-40
TEMP
Fig. 3. 24
Same as Fig.3.16 for model output at 20 GMT from
No moist adiabat is plotted.
CNK initial sounding.
PAGE
128
0
)
0/
/
0
0-
-
-
-8
I
0o
IJ
/
/
o
/
I
//
/HLC
GLD
-
CNK
0
P
0o
0
/
0
0
- 0
o
GCK
o
/
DPC/
o \
4,t"
I
0
I
S\
j
I I
Fig. 3.25
I
I
Change in surface potential temperature and dewpoint
between 12 and 22 GMT, 19 April.
Solid lines for
potential temperature (K) and dashed lines for
dewpoint (deg C).
PAGE 129
P
600
700
800
900
1000
4O
-40
TEMP
Fig. 3.26
Same as Fig.3. 16 for initial hybrid sounding, 11
GMT.
I-
)5 TH MERIDIAN
"IME
.;-
L.%
..
4 IFrancis
I
90TH MERIDIAWI
* TIME ZONE .
ZONE
Ic
rIed
. cM
.
l.
i AlmaI
l sonville
W
I
-
Ie
-
,uIe
Sain
l
C...
-*----£--
4lVne
--
o
G.r
ioPpr
Jly
O
5
owl~ad
Gr
-1
tt
Fi
3
-I
0
a'
27
lac
-l
-
9
Corri
*
o
CedarIlluf
Map
o
"
e
l
t ,Ia,
1p1. D
a
1
e
IC tr,
A
0r
- til
0be
I
0..
,,
El,,,..--
in Kansas
rainfall
to n
on
ollos
1
....
o
1I$
%
dc
o
1
.
r
18 April.
.
1
0 1
.
CWn
10 and
E
2a
c4r
.
rpt
S<
I E
It
.i.
DContours
(T),
trace
.
.
Dold-iPn.
,
.:
o__,o_._oo
C ,o
Ale VN
.nss
?m
*
sh-
Imp
r
1 ..
IsEot
do
o,_I
0ioot
<
ooaO
aa
Crea Lois
I0
inches
s.
tartso
Edmett
NW "0i -.
<P
I.Nr
It-"
f it*
•f
AO
Cl
Ite
,.
l
o
-e
©L-0"
Ie~d
Jefterll
4
s
o0I..n.
o 1lll
7
th
ml are0y
,n 2n...
If
or amounts as
Sdran
O
.
-
L trr
Or
rn
l
P
No
a..
.l
l
0
.
4*
ro
Il
CliC-l
Ciiton
WNa
a
r)T cit
660 0ths
L ke
P
............. LrL
pI
.W
ilh
icoln
lei t
81!-dUuat
I
r
Po
21
P *.'_7
el
r
Ih
Wqt
Ref
Uf
CS1
NNW
-
illv
._..IO S-
O111.8
...
..
L IlPawn....Ci
-rr
00
0
"
-
-
tsm
l
0L
CLodeW
"I--
I SS
,.F..l
Ilbroat
ol.lrbur
sl. o.
su
D 1sW.
areo
. 50
eAl
vo)
PAGE i31
TS NR
25 900.
QS
24 80023 700,
22 600.
12
21 50020 400
11
19 300
10
18 200
17 100'
16 , C
9
PH
700
750
800
850
900
Fig. 3. 28
Time variation of model output for APRHYB sounding 19
April, 50-50 soil parameters# with no extra factors
modelled (Plain). NR is net radiation into the
surface (mcal/sq cm rin), TS is temperature at top of
surface layer (deg C), OS is PBL mixing ratio (g/kg),
PLI is convective instability (deg C), PH is pressure
at top of PBL (mb).
Condition at top of PBL is
indicated above time axis:
blank = unsaturated, S =
nearly saturated, S = saturated S+ = oversaturated.
PAGE
TS NR25 900-
132
QS
24 80023 70022 600-
12
21. 500
20 400
11
19 300
10
18 200
17 100
16
9
,C
PLI
PH
700
750
800
850
900
Fig. 3. 29
Same as Fig.3.28 for run with morning clouds
imposed.
PAGE 133
TS
QS
25
24
23
22
12
21.
20
11
19
10
18
17
16
9
PH
700
750
800
850
900
Fig. 3. 30
Same as Fig. 3.28 for CNK sounding,
parameters.
70-80 soil
PAGE
TS NR
25 900-
134
QS
24 80023 70022 600
12
21. 500,
20 400
11
19 300
10
18 200
9
.17 100
16- , C
PLI
6.0
PH
TIME
700
5.0
750
4.0
800
3.0
Ph
850
900
Fig. 3. 31
Same as Fig.3.30 for run with all day cloudiness
imposed.
PAGE
DDC
135
11000-
14000- -0
600
700
800
900
1000
-40
Fig. 3. 32
0
TEMP
0
Portton of sounding data from Dodge City, Kansas
plotted on a pseudoadiabatic diagram, 11 and 14 GMT,
19 .A4ril. Solid line connecting dots is 11 GMT
temperatures, and dashed line connecting circled dots
is i4 GMT temperatures, where they are different from
the 11 GMT data.
Dash-dotted line is 313 K
isentrope.
PAGE 136
TS NR
25 900
QS
24 800.
- 12
23 700,
22 600
21 500
20 400
11
19 300
10
18 200
17 100
16 , (
9
2021
PH
700
750
800
850
900
Fig. 3. 33
Same as Fig. 3.28 for model run with inversion
changes imposed.
PAGE
TS NR
25 900-
137
QS
24 80023 700-
12
22 600,
21 50020 400
11
19 300
10
18 200
17 10016 , C
PLI
9
PH
700
750
800
850
900
Fig. 3. 34
Same as Fig. 3.28 for model run with DDC and
inversion changes imposed.
PAGE
TS NR
25 900.
138
QS
24 80023 70022 600
12
21 500
20 400
11
19 300
10
18 200
17 100
16
9
C(
PLI
PH
700
750
800
850
900
Fig. 3. 35
Same as Fig. 3. 28 for model run with GLD and
inversion changes imposed.
PAGE 139
P
600
700
-800
900
1000
-40
Fig. 3. 36
0
TEMP
40
Same as Fig. 3.16
bor
model output from APRHYB
initi-al sounding at 21 GMT with GLD and inversion
thanges imposed.
PAGE 140
TS
NR
25 900-
QS
24 80023 700-
12
22 60021
500-
11
20 40019 300-
10
18 20017 10016
0
9
PH
700
750
800
850
900
Fig. 3. 37
Same as Fig.3.28 for model run with DDC and
inversion changes and clouds imposed.
PAGE 141
TS
NR
QS
25 900
24 80023 700'
22 600,
21 500
12
bs
-11
20 400
19 300
-10
18 200
17 100
16
- (
9
2021
PH
700
750
800
850
900
Fig. 3.38
Same as Fig. 3.28 for model run with GLD and
inversion changes and clouds imposed
PAGE 142
P
600
700
800
900
1000
-40
-30
Fig. 3. 39
-20
-10
0
TEMP
10
20
30
40
Same as Fig. 3.36 for run with GLD and inversion
changes and clouds imposed.
Negative area is
cross-ha tched.
P
600
700
800
900
1000
-40
-30
Fig. 3. 40
-20
-10
0
TEMP
10
20
30
Same as Fig.3.39 for run with modified GLD and
inversion changes and clouds imposed.
40
PAGE 143
Fig. 3.41
Mesoscale analysis of convective
and
S.i
convective inhibi tion
d lines
are PLI (deg
(NA)
C) and
instabilitu
(PLI)
or 17 GMT:
19 April.
dashed lines are NA
S(mn;+2/s*.2
Cross hatching is -or new radar echoes
appearing between 18 and 2-1 OMT.
Numbers in boxes
are point vles
of PLI,
circled
numbers for NA.
points.
circles denote da-t
Filled
PAGE
144
O0
0
.
\00
f
O
LBF
*o
350 -
NK
oo
0
O
DDC
0
o
0
40
Po
0
100
0
Fig. 3. 42
Same as Fig. 3. 4 1
crm 21 to 23 -MT
or 20 GMT,
and new radar
echoes
PAGE 145
9 MAY
CASE STUDY .II:
4. 1:
Introduction
On this day,
afternoon in
border.
and
two convective
tornadoes as illustrated
July and Turner
details of
formed
(1980).
on this day
became severe and
on Fig. 4.1 w hich
is
A brief discussion of
taken from
the
synoptic
The character of the
soundings pertinent to this case are .included in
Finallu,
caused hail
foliows, after iw-hich the mesoscale
the case are described.
discussion.
in the
the Texas panhandle region near the Oklahoma
The convective activity
organization
outbreaks
the mesoscale
the modelling results for these two
outbreaks are presented.
4. 2:
Synoptic_ Analysis
The synoptic
importance
On Fig. 4.2
to the convection.
maximum could
be seen near the base
500 mb at
12 GMT, 9 May.
vorticity
advection
(CVA)
Notice
which
-northeast into New Mexico.
this area of CVA moved north
near
in this case
organization
By
proved
a strong vorticity
wave trough at
of the long
the region
stretched
00 GMF
n
to be of
of
cyclonic
from Mexico
10 May,
(Fig. 4.3
)
to the New Me:.ico/Texas border
the Texas Panhandle, suggesting
upper-level
support for
At the
surface, at
12 OMi'T,
cyclogenesis
in that
regionn.
9 May
PAGE 146
(Fig. 4. 4)
of low pressure
a v-shaped region
stretched
northeast across the Texas panhandle, and reached
Geostrophic
into Mexico.
most of Texas and
trough
had moved
winds were
slightly
south-southwesterly
By 00 GMT May
Oklahoma.
east.
10
and had
moved
northwest corner of the Texas panhandle by
these synoptic-scale
synoptic-scale
forcing could
Panhandle during the
geostrophic
laterT
features
have
time.
implied that
Notice
or southe.asterly
10 May
The
the Texas
that the
across most of
(Fig. 4. 5).
Mesoscale Analysis
Using
network,
700 mb
the data available from the special radiosonde
shown in Fig. 4.6, analyses were made at 500 mb and
for each
of the
GMT, and 20 GMT.
four nominal
The 500mb
shown on Figs. 4. 7 and 4.8.
from El
Paso, Texas
winds changed
slightly
were
this
to the
been important in
Texas and Oklahoma even at 00 GMT,
4.3:
this
hours of 9 May.
wind was south erly
(Fig.4. 5)
over
Also, a cold front located
northwest of Texas had sharpened
presence of
southeast
in
analyses for
A band
southwesterl
Mississippi river.
11
11
between
GMT,
14 (MT,
17
GMT and 20 GMT are
of 50-70 knot winds curved
(ELP) to Omaha, Nebraska
little in direction
speed
times
(OMA).
or location, and
the two times.
Winds
These
increased
in the network
from the mountains in the west to the
PAGE 147
At
500mb,
approaching
and
4. 10
ratio
of
the
trough
expected cooling
occurred
ields
show the
primarily
of change
12 t o 20 GMT an-d
from
Nebraska
(LBF),
New Mexico
sharply
fell
to
(ABc)
outbreak
in
a band
Dodge City,
temperature
both
We
Fis.
and
4. 9
mixing
The magnitudes
periods,
but the
from North Platte,
Kansas
after 20 GMT.
thermodynamic changes
in
during
the
after 20 GMTl.
from 20 to 23 GlT.
moisture change were small
temperatures
associated with
(DDC) to Albequerque,
can conclude that
at 500mb were small
prior to
the
the
of convection.
At 700mb
(Figs. 4. 11 and
4. 12),
from the
same direction although
dramatic
change
to western
in mois ture
Illinois,
from Amaril lo,
Texas
lower in
gradient
speed.
essentially
There was a
from the Texas Panhandle
by 20 GMT the moisture difference
so that
(A MA)
winds were
the
to Oklahoma City,
to 6 i1 g/kg.
The
Oklahoma
panhandle
(OKC)
region
grew from 2 .7
g/kg
experienced a
strong
The fields
change of temperature and mixing ratio at 700mb
are
of
in crease
in moisture through
i
shown o n Figs. 4. i3 and 4. 14.
to two
degr ees C
occurr ed
Notice
in a band
gradient no ted above
by 20 GMT.
that cooling
along with
moisture in creases whic h were responsible
These
this period.
the
for the
changes
of
large
increase in
were
likely
resujlt of v ertical moti on associated with a frontal
c irc ul1 at ion
k irnematic
a: low leve Is.
vertical
motion
Ogura
tor
this
et al. (192)
case,
and
calculated
their
one
results
the
PAGE 148
show upward motion along this
At the
band region.
surface, the situation also showed change during
this period.
are shown on Fig. 4.15.
12 GMT conditions
The
to southeasterly flow over Texas and
There was southerly
southwesterly
Oklahoma becoming
in western Texas and the
from
front stretching
There was a cold
Mid-Panhandle region.
the Panhandle northeast through Kansas, associated with a wind
shift and a
pool
lay ahead
(LBB).
drop
of the front centered
17-18 deg C
and
warming
15 GMT (not shown),
(TCC).
strong rises
By
tongue,
in
lost some of its
18. GiMT
with
(Fig. 4. 16),
central
warming covered a
to Oklahoma, and
showed evidence of
to the north
(+ 3 K at
temperature, as much as 7 K at
(CVS), and
Consequently, the
panhandle
(CDS).
The slopes of the Rocky Mountains
Dalhart, Texas [DHT]).
Clovis, New Mexico
reported at
or Childress, Texas
the warm pool
Texas
moist, with
or fog
low cloud
(up almost 2 K at LBB) and
showed very
near Lubbock,
front was very
stations such as AMA, LBB,
By
A warm
in dewpoint and temperature.*
The warm sid-e of the
dewpoints of
many
sharp
4 K at Tucumcari, New Mexico
temperature gradient in the
strength.
the warm pool
warming of
about
had
K
become a warm
from 15 GMT.
large region from the mountains all
north
to K6nsas.
The
front retreated
This
the way
to the
PAGE 149
Oklahoma panhandle,
the Texas
located
while the
winds
the
on
panhandle became easterly.
slopes region of
A wind
shift line was
near the Oklahoms/Texas Panhandle border (northerly
winds at Gage, Oklahoma [GAG] and Canadian, Texas ECANJ),
the temperature gradient was
gradient was not
available data,
the mountains
GMT value.
sharp, at
with a
(CVS).
At 21 GMT
drying
gradcual
(Fig. 4. 17) convection began at
GAG.
of the up,-ind
period
warming
of
temperatures
and
and
to 21
drying
the convection
warming
of the tongue.
northeast through
along
The
the
central
little
side of the wind
GMT.
were
in
broke out
drying.
in
in
resulting
ranging
drying can be
12 GMT
and
continued,
intense to the north
(southerly)
to 9. 5 g/kg respectivel,
of warming
shift
temperature of 306 to 310 K, mi.xing ratio of
potential
potential
some
source air for the convection must have had
characteristics
13. 0
the wind
To the southwest heating
Oklahoma/Kansas- border, curving
shift:
its 12
still near
reporting stratocumulus.
temperature gradient was still
The
from COS to
now mostly cirrus with only
changes of 4 K/3 hours in the center
Kansas.
The moisture
to west
east
Moisture at CDS was
still
line near CAN and
with
north.
least within the resolution of the
Cloud cover was
Oklahoma stations
farther
but
equivalent
from 338 to 342 K.
seen on Fig. 4. 18 for
i'lotice
that
the western
a region
tthe major
si-de
of the
of relativel
The pattern
the
entire
centers of
region,
moderate
and
PAGE 150
By 23 GMT, Fig. 4. 19,
southwest
the
of the
second
April
first.
second convective outbreak began
the
The radar
outbreak was separate
pictures (Fig.4.20) that the
and
in this
triggered outbreak
line of echoes between points A
would begin.
convective outbreak
second
of 316 to 313 K and mixing
likely
The
the way
growing PBL, and dry
observations:
second
The
outbreak,
occurring
least +2
located at a clear wind shift
convection, but
forc ing.
quantities
The moisture
inflow to
inflow to the
and 23 GMT.
the second
line.
12 and 23 GMT
zero for the
K for the
between 21
the first outbreak,
of a region of weak
From the
for moisture and
of -1. 5 g/kg b.etween
The temperature advection was
Unlike
dry
were estimated on each analysis.
first outbreak, and at
The
of available moisture into the
temperature advection were estimated.
at LBB.
This
to 346 K.
entrainment from above the PBL.
advection was on the order
for this
ratios of 8. 5 to 11. 5 g/kg.
particularly at LBB.. the values
and V' VT
a gust
temperatures
had potential
lower for a combination of reasons:
advection, vertical mixi-ng
V- V
inflow air
temperatures of 341
equivalent potential
moisture was
the
sequence of
B appeared simultaneousily, not sequentially
front
gave
just as in the
from the first
Notice
in chapter 3.
19 case
film from AMA shows that
outbreak was not
There was
some
indication
convergence on either side of the
the data
does not
indicate strong dynamical
PAGE 151
4. 4:
Soundings
For the 21 GMT outbreak) GAG and CAN at 20 oMT were th,
closest
sounding locations
However
both
stations had
to the actual
outbreak
been experiencing northerly
flow, so these soundings were not representative
air.
Shamrock, Texas
(SHM) was apparently
although
parcels would
the wind
shift line from SHM.
GMiIT,
11
which
(Fig.4.21).
This was capped
became almost dry adiabatic
by a
inflow
inflow air,
have r'equired about 3 hours
At SHM at
surface
of the
in the
showed a moist layer near the ground extending
100 mb
area.
to reach
the sounding
upwards about
dry stable
above 700 mb.
layer
The PLI for
SHM at 11 GMT was 4.5.
By
17 GMT
developed
(Fig.4.22), an approximately well-mixed PBL had
this PBL 25 mb
higher, became
from 800 to 700 mb.
buoyancy,
Parcels . lifted
in the lowest 60 mb.
saturated and
After a brief region
the parcels were buoyant past
from the top of
positively buoyant
of
small
500 mb,
equivalent potential temperature of about 346 K..
higher equivalent potential
surface observations at 21
km before
over 100
wUarming
and
potential
being
temperature
GMT, but this
lifted
drying took place,
temperature.
warm tongue showed
at 21
than
lowering
This was a
to travel
during which
time
the equivalent
Soundings closer to the
less moisture and
with an
inferred from the
air had
GMT,
negative
center of the
deeper PBL development.
PAGE
CDS at
17 GMT (not shown)
depth, with
slightly
At 23 GMT,
had a well-mixed
PBL of over
lower values of mixing ratio
the nearest
152
100 mb
in the PBL.
soundings to the second convective
outbreak were AMA and CDS. neither of which were actually
upwind.
warming
The AMA sounding
and
drying were
the result of BLH and
layer
up
was downwind of
pronounced at AMA during
mixing.
There was a very
to almost 600 mb at AMA by 23
The CDS sounding
layers aloft.
A
forcing
-MT
the
deep
day,
The
both
dry mixed
(Fig. 4. 23).
at 23 GMT was noisy, with superadiabatic
of the CDS 20 GMT sounding
comparison
(Fig. 4. 24) with the 23 GMT sounding
dynamical
the convection.
(not shown) suggested
had modified the vertical
structure.
structure at 20 GMT showed an almost perfectly
that
The
dry adiabatic
PBL from the surface to 725mb.
This PBL was shallower by 50
mb
the 23 GMT CDS sounding was
on
the 23 GMT sounding.
not useful
for the analysis.
characterized
about 342 K,
moisture
data
by
a surface
although
in
The 20 GMT sounding was
equivalent potential
temperature of
the value was uncertain due to missing
(see Fig. 4. 24).
This variability
difficulty
So,
in
analyzing
structure accounted
this case.
Ogura
for the
et al. (1982),
after
studying this case to determine what
trigger mechanisms were
present,
conclusions.
could make
only
c:onjctura.
There
was
a
PAGE 153
noticeable
of
lack
sufficient
storms
oF
developed..."
analyses, and
"The lack
the critical area where the
the reason
the upper air
that their
307
data
emerged
examination, of the
generallg
be characterized
313, 317, and
and 313 K usually
319 K.
contained a strong
sometimes
layer was
GMT.
from
The overall
by
layer
pattern
very
slowly eastward with
region.
time at an average
over the region
in the
layer.
The depth
on Fig. 4.26 at 20
depth through a
each
side.
speed
that
height of
temperature
of 317 K).
low)
stCong
inhibition of
well
il
ustrated
b
Where the
and moved
of about
this
rather
varied mostly
the bottom of the
This
10
lauer of air
from the Mexican Plateau
form of a finite pool
of the pool
the
adiabatic
The pattern on Figs. 4.2 5 and 4.26 suggests
air was
of the
between 307
throughout the period,
Ogura et al. (1902) determined
been advected
to
of 4
size of the latter
GMT and
small depth
seemed quasi-steady
Snots.
The
The
the heights
inversion, while
pattern was one of large
strip and
isentropic
soundings.
The
considerable depth.
plotted on Fig. 4. 25 at ii
north/south
conclusions
"..
From 31i7 to 319 K represented a nearly dry
layer of
had
data over
subsequent
could
isentropes:
is
their analysis.
in
speculative"
One aspect of
soundings
icity
rawinsonde
are admittedly
layer
speci
pool
pool
than a continuous
by the variation
(i.e.at
was deep
convection existed.
the O lahoma Cijt,
that this
Oklah oma
potential
(and therefore
This point was
(OC) 2) O IMT
PAGE 154 •
sounding
shown
the pool
was over OKC.
the
on Fig.4.27.
bottom of
100 mb,
the nearly
The low level
large depth
could
pool
as a result of the
adiabatic
bottom of
have been
the movement of the
parcels following
this
of
even
buoyant
low level of
the
layer from 317 to 319 K.
the pool and
the
coincident
the result of subsidence.
pool was
(500 to 700 mb),
the deepest part
on Fig.4.27 were negatively
dry
of the
this time
Notice that PBL
342 K pseudoadiabat
for more than
At
independent of
the winds
Since
in the
pattern of movement and depth
have been the result of a wave passing through
could
the middle
troposphere.
During the time period
11
to 20 GMT, the eastward
shift
of this feature left the Texas and Oklahoma Panhandle regions
with 317 to 319 K
depth.
This
convection,
lagers
change was
less
the thinner
higher bottoms of
tWe layers.
least by
lid.
the possibility
frontal
effect.
below to
surface convergence
and movement aloft of
the
et al.(1982) also found evidence for
of symmetric
inland sea-breeze
are examined
OgLra
of
data analysis, seemed to have
its associated vertical motion,
317 to 319 K
initial
layers were associated with
been the result of PBL heating,
with
of their
important to the outbreak
since
The convection, at
than half
instability
and perhaps the
The modelling results for this
quantify the
thermodynamic factors
case
PAGE 155
including
4. 5:
GHMT
BLH,
clouds and
21
GMT Modelling
The
modelling was
corresponding
regions
soundings,
a hybrid
sounding
soundings were averaged
isentropic
analyzed,
two
two times,
(MAYHYB) was
and
and were used to
21
GMT and
outbreaks.
represented
interpolation.
surfaces,
the PBL.
convective
were not well
isentropic analysis and
above
analyzed at
to the
convective
the 4
changes
by any
created
Since the
of the
using
The AMA and CDS 11 GMT
these data were plotted.
Then
theta = 307,313,317,319 K were
interpolate the pressure
levels for
those 4
points for comp-arison with the averaged data.
profile
below the 307 K
a moist adiabatic
23
isentropic
lapse rate,
level was adjusted
since the corresponding
The
to give
layers
at CDS and AMA as well as other nearby soundings were also
moist adiabatic.
The profiles in the rest of
the hybrid sounding needed
this
process was
shown
adjustment.
sounding
used
in
The result of
for the modelling
in Fig. 4. 28.
The area was
days,
give
the hybrid
no such
the layers
so low soil
without rainfall
surface moistures
the most realistic results.
measurements
Center
during
(GWO) were expected
The
lower moisture
to
two
to
few pan evaporation
(not shown) available from
sugqested
the previous
the National Climatic
the west,
but we can
infer
PAGE
that the
soil surface was
fairly
west of Arkansas and Missouri.
dry
over
156
the entire region
The previous month's rainfall
(not shown) was near normal.. suggesting moderate values of
bulk
soil moisture
soil
surface moisture were varied
saturation and
70%.
(GWB).
With
these aids.
the values for
between 5% and 30% of
the bulk moisture was varied between 30% and
One additional value for OWO was tried
value, namely a 50% value
(GWO = 50%,
for the 70% GWB
GWB = 70%).F
to test the
effect of a more moist soil, even though -this was unrealistic
in
this case.
The model
was first run plain
(P),
without clouds or
changes above -the PBL to determine sensitivity to soil
moisture parameters alone.
an
easily understood
way.
cooler, and we-tter PBLs as
the PBL top
PBL moisture
(PH-)-,
area
Th-e wetter soil runs had shallower,
evidenced
surface layer
(GS) behavior
was high-e-r for
sensitivity
the
All PBL characteristics behaved in
the wetter
soils,
femperature
indicating
The sensitivity
For OWO from 5 to 30%
(TS),
and the
instability
greater
The negative
in a complicated manner.
surface moisture runs had
moisture runs had
The
to the heating.
(NA) varied with moisture, but
higher bulk
the pressure level of
(see Table 4. 1).
to the moisture than
The higher soil
by
lower NA, while the
higher NA.
parameters explained what was happening.
the
sensitivity
in GS was 20%.
0
157
PAGE
However,
the G-1WB range,
for
TS and
PH sensitivity
•airly
small
depended
for .both
(5-6%)..
only
allowed
drop
in PBL height and
eff
for the NA,
wetter
for
soil
bulk
the GWB values was
although
only
only 8%,
while
The
run.
remained
for GWO it
is
shown
The
for
the
small
the GWB,
to overcome
was an
(as in
the
the other
increase in
the increase
NA for
in moisture
of PLI
to OWB was
was 28.
on Fig. 4. 29.
through
temperature
quickly rose until
15 GMT.
21 GMT.
Notice that
the PBL
As Table 4. 1 shows,
only those with low soil
continuously during
rose
The PLI
top
dropped
After this time,
briefly
this
surface
the day,
then
little change
in
occurred.
The
addition
Clinton
overcome
But
The sensitivity
while the moisture dropped.
fog
and
sufficient to increase the PLI,
slightly.
unsaturted
moisture.
of
QS sensitivity
to
Notice that
true for all runs,
instability
similar
The
time variation of various parameters for one plain
5-70 P,
wlas not
High
the result
11%.
of NA must have
sufficient
values.
only
"W4B lwere
sensitiity
temperature.
and
was
and
additional moisture
increase was not
ects
GW
S. values.
GWO)
moisture
the
so
on the
the
0S sensitivity
first modification made to
of clouds.
and stratus
Sherman
AFB
The surface
cloud
the plain runs was
observations showed
in the vicinity of CDS,
Oklahoma
(CSM)
through
the
evidence
AMA, and
15 GMT.
The
PAGE 158
cloud
amounts used
amounts were based
and
the relative
these
in these runs are shown
clouds was tested
in
in the soundings.
of PBL tops averaged about 9%
greater
losses were 9%
than the soil
smaller.
than
cloud cover
lower.
clear runs.
less sensitive to
to soil moisture.
The change
less unstable, with PLI
in NA was significant however.
The NA for the .cloudy cases were roughly
the
and the
less,
moisture sensitivities discussed above
The resultant PBLs were slightly
growth 6%
The change
These sensitivities were
TS, but the PBL moisture was
for PH and
to lower PBLs
for all soil
the temperatures rose 11%
Similarly
conditions.
moisture
clouds contributed
and higher moisture values.-
lower temperature
in growth
The importance of
(C) of runs.
this series
As Table 4. 1 shows, the
with
on the surface observations,
subjectively
humidity
The
in Table4.2.
The lower PBL tops and
35% higher than for
lower surface
temperatures were- sufficient to overcome the slightly higher
OS values to give
"cooler"
pseudoadiabats for parcels
combined with
cloudy runs.
The cooler pseudoadiabats
greater depth
in the atmosphere for parcels to travel,
significantly
higher NA.
The time variation
Fig. 4.30.
of the cloudy 5-70 run
Notice the response
in the
the
is shown
of the net radiation
gave
in
to the
PAGE 159
removal
of clouds at
17 GMT.
The
morTe quickly afterwards, although
the plain run.
but was still
temperature began
it remained
to rise
lower than in
The moisture began to drop faster at 17 GMT,
higher at 23 GMT than
that the diminution of PLI
mirrored
in the plain run.
the moisture drop at
GMT.
The height of the PBL rose more slowly
ended
up
leveling
off at a
slighlI
Notice
lower
17
under cloud and
level
than in
the
plain run.
As already noted, during this period many changes
occurred above the PBL.
a
series of model
development.
derived
in
These changes were incorporated
runs to test their effect on mixed
The changes used
in the model runs
the following manner.
averaged at each
(H) were
2 deg K.
in
terms of vertical
These changes were
time to give a profile of changes at
deg K for the h ybrid
the
layer
The changes between each
sounding at CDS and AMA were determined
motion of theta surfaces at each
sounding.
each 2
These were then plotted
hybri.d sounding and the changes
The changes in moisture were
into
for
in temperature measured.
taken directly from the average
of moisture changes for the AMA and CDS soundings between each
time.
actual,
Clearly,
this method was an approximation
but unknown changes which did occur.
variabilitu
in
Since there was
the changes over the SESAME region,
linear variation
scale.
to the
However,
between AMA and
CDS was
crude on a
assuming a
small
the important mesoscale changes related
to
PAGE 160
the movement of the pool
stations.
The resulting profiles of change in
moisture are shown
atmosphere
lower
The
oscillated with
tendencies
Theu
cooled aloft with
moisture varied
time.
during the
period.
in Table 4.3.
generally
levels.
pattern
of Mexican air were similar at both
show that the
time, and warmed at
by a
large amount and
Notice that a moistening over drying
first period gave way
of nearly
to opposite
the same magnitude during the next
This wave-like
behavior is more indirect evidence for
a wave disturbance above the PBL which could
responsible
have been
for the Mexican pool movement.
The results from these runs
4. 1.
(H series)
are shown
The PBLs were all deeper, with about 10%
for all
soil
runs, but
runs.
(2%
less growth)
a large amount
(almost a 20% larger drop).
5-70 runs (P and H)
and
is- clear- that the H runs showed a
during
the
relatively
14-17 GMT period,
low levels.
out
moistening
time sequence
when strong
cooling
lessened
faster growth.
took place early
run was drying
in
The cooling
above the PBL and allowed
drying
larger growth
and the moisture
Comparing the
it
in Table
The temperatures were lower than the plain
by a smal.l amount
values were- doan b-y
4.31),
temperature and
(Figs.4.29
jump
took place at
the stability
The accelerated
since the change imposed
low in the atmosphere in
in the 775 to 825 mb
further diverging until after
in depth
on the H
the first period.
The
layers kept the two runs from
16 GMT when the PBL. in the H run
PAGE
grew past the moistening
layers and
began
entraining
161
the dryer
air above 750 mb.
The accelerated growth
responsible
for the slightly
the growth of the PBL and
use
smaller
stability,
expense
of
heat flux
more growth
the heating
(i. e. for similar soil
than the
if
Additionally,
the PBL, the
temperature of
entrainment added
the PBL
With
in the PBL will .occur at the
moisture contents).
plain runs
in
Both
from the ground.
the sensible heat flux
parameters.
the
cooler PBL temperatures.
the rise in temperature
the available sensible
cooler
due to smaller stability was
for the same
since
So
is the same
the H runs were
soil moisture
the stability was
entrained air was
less heat to the
less above
less,
so that
growing PBL, which
then
stayed cooler.
These diff-erences, especially
to a
contributed
instability.
from
In
its initial
large drop
the lower OS values,
in the development of convective
three of the H runs,
value of 4.0.
The
the PLI
sensitivity value was 58%.
The difference between the plain run and
shown on Figs. 4.29 and 4.31
positively
to the changes,
moisture cases contributed
higher PBLs and
the
"cooler"
lessened
the H run
dropping by
as
The high soil
52%.
to this
began;
The NA responded
at about 15 GMT.
strongly
actually dropped
percentage.
stability aloft overcame
pseudoadiabat effect as shown by
The
completely
the PLI changes,
PAGE 162
-and gave all
of the H runs less
higher PBLs also contributed
all
of
the runs were
inhibition
to saturating
to convection.
the PBL tops, as
oversaturated at 21 GMT.
Finally, a series of runs was performed
both morning
Since many
only
the
(C)
clouds and
the
change at all
the clouds
(H series),
effects was not
The change in
to 6%
compared
with the plain 5-30 run.
lower
The
for the combined run by
The 30-30 HC run ended
less drying).
understand.
Both
separately gave
strange for
height of
up
with the same GS as the 30-30 plain.
The TS changes were the
the clouds and
the imposed
lower PBL temperatures,
the PBL was not
combination tendencu was
as
(as
simplest to
changes
so it would
the combination to show anything
produced a slightly
namely
by 8%.
three runs all were wetter than the plain runs,
much as 6%
12%
The PBL moisture obeyed no clear trend.
the 5-30 run, the net effect was greater drying,
The other
less
the drgest soil run, 5-30 HC showed no net
from the plain run.
runs
changes
towards lower values (up
temperature rises were all
very
imposed
The results appear on Table 4.1.
although
(HC series).
had opposite signs for
the runs with
PBL height was generally
growth),
incorporating
changes aloft
expected .outcome of their combined
predictable.
For
imposed
of the tendencies
runs and
The
have been
else.
simple to understand,
The
since the H
larger effect on the model.. but the
in the
towards lower PBL tops.
direction of the C runs,
Evidently, the
loss
of
PAGE 163
incident radiation was
sufficient to
than -the reduced stability
the
could make up
time variation of PH on Fig. 4. 32,
plain run
a 40 mb head
After the clouds were removed,
rose faster
than
the
by 21
deficit
but could
entrained
during
the
By the time
thinly and
the
the
the
the rapid
clouds were
the OS value was higher than
the plain run,
concentration
When
the PBL grew to the
the moisture was
spread more
(mixing ratio) was
less.
wetter runs, the PBL never grew quite deep enough
effect
For
run was dryer due to
17-21 GMT period,
total moisture content.
level as
not quite make up
But the PBL was now shallower, and actually
the plain run.
same
over the HC run.
to the PH values.
same P-values, the H
from the combination run.
less
that of the !5-70
GIMT.
period of -the HC run.
removed
of
the combination run's PBL top
the plain run,
5-30 run, with the
growth
A comparison
the clouds were present
start
'The GS variations were related
dryer air
for.
with
that while
(Fig.4.29), shows
the plain run built up
had
limit the PBL growth more
to take place, and
for
the wetter soil runs had
For
the
this
slightly
higher GS values.
The instability
than in
differences
rising
values were all
the plain runs.
(PLI
dropped
for the 5-30 run)
The
dryer
from its
smaller for
runs showed
original
while the wetter
the HC runs
the biggest
value
instead of
runs showed
onl
PAGE 164
small
differences
run).
This
(10%
behavior was clearly a result
As a group,
moisture.
increase in PLI
less
and
individually,
for the 50--70
of the higher PBL
the addition of
significantly higher
clouds to the H runs produced
instability.
The NA for.all
plain runs,
by
an average
of more than 60%.
than
less conve.ctive instability
followed
other
to lower NA.
to higher PBLs.
due
to
their plain
4.34 the soundings
result of
shows
diagram.
that the
On Figs. 4. 33
runs are
of these
A comparison
difference
in NA was a
the HC run.
in Fig. 4. 32 showed the very
saturation
flat
of the PBL top at and after 21
convective instability
entire heating cycle
(.3S.
left for the
the different structure above the PBL in
behavior of PLI and
The
lower or equal
for the 5-70 P and 5--70 HC
The time variation shown
GMT.
lower NA could not be
the
imposed above the PBL.
plotted on a pseudo.adiabatic
so
did not
only reason
The
and
higher NA if all
since the HC runs were all
counterparts.
two figures clearly
gives
This obviously
Similarly,
lower. NA was the changes
and
The HC runs had
plain runs,
the
cooler pseudoadiabats, which
aspects are unchang-ed.
contribute
lower than for the
of the HC runs was
changed
little through
despite the clouds and
the
changes in TS and
PAG(E 1 '5
4.6:
Summary for 21 GMT Modeiling
The responses
of the various parameters are
Table 4.4 for the 21 GMT runs.
insensitive
changes
to all
'The PH values wre-re
of the effects, although
(H) made a 10%
summarized
the
in
latively
imposed
increase in the growth of the PBL.
The
surface temperatures responded to the presence of cloud and
this effect was
changes aloft
both
enhanced
(HC).
slightly when coupled
Surface soil moisture and
to the
imposed changes
had a profound effect on the PBL moisture, 0S.
The
presence of clouds made a noticeable difference too,
sensitivity was only
The convective
changes above
area
negative
one half that
and
but the
imposed
changes..
instability responded most to the imposed
the PBL, and also to soil
p-roved
surface moisture.
to be sensitive to nearly all
factors, particularly the imposed
imposed
of GWO and
imposed
changes seeme-d
changes aloPt..
to be dominant
the NA calculation, reducing both
The
of the
In fact,
the
in both the PLI values
instability and NA.
The
soil moisture parameters were the only positive contributors
to convective
The
instabil.ity besides the basic
other factors, clouds and
changes aloft, tended
the PLI.
The NA tended to drop,
increased
soil surface moisture and
addition of clouds and an
increased
inhibition.
heating
on the other hand,
imposed
increase in bulk
changes.
itself.
to reduce
for both
.Both the
soil moisture gave
PAGE 166
The
case.
non-lineariaty
As
in
the April
opposite effects when
of the problem becomes acute
case,
imposed
the presence of clouds had
considered alone and when examined
connection with the other
clouds and
in this
factors.
in
When the 5-70 run with
changes aloft is
compared with the same run
without clouds, the results are as follows:
the addition of
clouds yields hig-her instability and
lower inhibition.
the 5-70 run,
importance, the relation
in
decreasing order of
For
is:
PLI
clouds +
=
bulk
NA =
-
soil
initial
initial
moisture -
imposed
conditions -
BLH -
surface moisture - bulk
The net effect
lower
soil
conditions + BLH + soil
for the
soil
changes
imposed changes -
clouds
moisture
sum of the factors
convective ins-tab-.ility
surface moisture +
exchanged
(HC runs) was a
for a drastically
reduced NA.
The comparison with observations was reasonably
when using
shown
the 5-70 HC run.
in Fig.4.35.
after 15 GMT.
high.
top
This comparison of TS and QS is
The model was slightly too hot.and
This suggests that
At Cheyenne, Oklahoma
was near 800 mb,
close
dry
the PBL depth was likely
(CHE) and SHM at 20 GMT,
compared with 725 mb
too
the PBL
for the model
at 20
PAGE
However, the
GMT.
the region
represent
clouds used
in
hybrid sounding was designed
in
farther
the model
in
southwest, and
were not as thick
the CHE and SHM soundings.
The
167
originally
to
the
particular,
long
or
lasting as
17 GMT SHM sounding
(Fig. 4.22) showed relative humidity >75% over a 100 mb depth
some
above the PBL implying
(Fig. 4. 36)
showed RH . 75%
AMA at 17 GMlT
cloudiness.
everywhere.
These observations
gradients of cloudiness and hence PBL heating across
implied
the data void
region where the convection broke out.
passage of time, this
translated into gradients of PBL depth
The model PBL
surface temperature differences.
as well as
With the
depth at 21 GMT was therefore closer to the actual PBL depth
in
the
Notice finally
outbreak area.
began at the wind shift
20 GMT.showed very
instability
line,
low NA
the convection
the CHE sounding at
even though
(2.86),. and high
convective
(6.0 ).
These results su-ggested that
determining
that
the first
the critical
convective outbreak's timing
location was the existence of the wind shift
vertical motion was sufficient to remove
present and release the convective
the
factor
factors modeled were important
and
line.
This
the inhibition
instability.
Of
course,
in determining what
inhibition was present for the vertical motion to remove,
the model
outbreak
results
but
implied that the NA was greater at the
area than to the east near CHE.
Similarly,
the CHE
PAGE 168
region
showed higher
instability
than the region of outbreak.
Fig. 4.37 shows the analyzed fields of convective
instabilitu
18 GMT, with the echoes at 21
inhibition at
and
Notice
GMT superimposed.
occurred west of the region of
PLI.
This
figure agrees well
basis of the model
might be
results.
4.7:
with
lower NA than
showed
the outbreak region.
in
in an area of low
the conclusions made on the
The model
there clearly was
The surface wind
overcome this inhibition
convergence was necessary to
to release the
lowest NA and
from the analysis, but
interpolated
inhibition still
of convection
that the outbreak
in order
instability.
23 GMT Modelling
The modelling performed for the 23 GMT outbreak was
largely an
was the
extension of the 21
simulating dry advection in the PBL.
observationsi. 1.5 g/kg
simulated.
sounding
of the drop
-As noted
that of
in the
in mixing ratio at LBB was
The temperature advection was not
This omission was made
much more intimately
for
initial
One additional effect was addedi
same.
due to advection.
The
changes aloft and clouds amounts
same and the imposed
were also the
GMT runs.
involved
since the temperature
in the model
parameterizations
the PBL characteristics than the moisture.
temperature advection would
have required
is
To
include
significant model
PAGE 169
changes,
and this was deemed beyond
The expected
effects
discussed below.
the 21 GMT runs!
discussed
in
were estimated,
Many
of
so only
the model
the present scope of work.
however, and are
responses were similar to
those which were different will be.
detail here.
The plain runs were used
soil moisture.
The results
to determine the sensitivity
for PH, TS, QS,
and PLI
essentially
the same as for
Table 4.5.
The NA reaction was quite different.
of
the NA to the different
5-30run had
additional
less NA
the 21 GMT runs,
soil
the NA.
more NA than the 30-70 run,
Similarly,
suggesting
the
conflicting
for NA,
inhibition to
to particular
be seen that th-is trend
combinations of
continued
Clouds were added as
the
results were similar
The PBL growth
as for
Again, the
drop
10%
and cooler
this ambiguity
parameters.
GMT runs
surface
It will
(Table 4.2) and
(see Table 4.5).
temperature
The PLI
the 21 GMT runs, resulting
lower
The
throughout the 23 GMT runs.
the 21
less.
results.
convection can be very
in all categories
was 8% less,
and PDL moisture
exactly
in
that
opposite tendency.
discussed also showed
sensitive
The
the 5-70 run had
21 GMT HC runs previously
that the
shown on
The response
implying
However,
the bulk moisture showed
implying
were
moistures was variable.
than the 30-30 run,
GWO increased
and are
to
rise 8%
less
values behaved
in 6%
less growth.
PBLs were less unstable even
PAGE 170
though surface moisture
increased.
runs by more than 40% over
since all
were
of the
in the
inputs,
direction of
the plain runs.
increasing NA.
saturated
(see Table 4.5).
The
One
lower PBL top)
other
at the PBL top
change was
by 23 GMT,
by 21 GMT.
When changes above the PBL were
results were again similar
in these
This was expected,
(cooler pseudoadiabat,
that 3 of the runs were saturated
whereas only 2 were
The NA was higher
imposed
(Table 4.3),
the
in most respects to the 21 GMT runs
instability
growth was reduced
less for
the 2 extra hours of integration, down only
47% from the plain
run compared
The NA change was
with 58% for the 21 GMT runs.
the most variable, ranging
soil
to a
decrease of
NA was very
from an
100% for wet soil.
for dry
In these runs,
the
sen-sitive to the exact pseudoadiabat which
characterized th-e instability.
unstable enough
The wettest soil
run was just
for PBL parcels to be able to miss the
inversion- at 650 mb when
zero NA.
increase of 18%
lifted moist adiabatically,
yeilding
Notice an Table 4.5, however, that all of the runs
were saturated at the PBL top,
too quickly
suggesting that the PBL rose
for the available PBL moisture, and
GMT structure could not be realized
The HC runs had both
clouds and
characteristics except the
dependent on
soil
moisture.
that this 23
in the real atmosphere.
imposed changes.
All PBL
surface layer temperature were
The range
of variation for PBL
PAGE 171
growth was 4%
more growth to 3%
less growth,
The surface temperature was held back
imposed
changes, resulting
in a
The moisture change, varied
The instability
+/-
was generally
10%
5%
by
both
values than the dry
duplicated
depending on soil moisture.
lower,
although
its behavior in
almost no NA.
Higher
the H runs.
just as with
The dry
unrealistic
the wet soil runs had
the H runs,
all of the
tops, a
structure.
The simulation of the dry advection produced
drying
in PLI
soil HC runs
PBLs in the HC series were oversaturated at their
physically
soil
The NA variation essentially
inhibition, and
However,
the amount of
to their plain counterparts
soil runs.
had marginally higher
clouds and
smaller temperature rise.
change also depended on the soil moisture.
moisture runs were closer
a small effect.
by a factor of 30%
(see Table 4.6).
greater
This did not
affect either the PBL depth or the surface temperatures.
change
in QS was not significant
energy
balance, hence the flux distribution was the same,
giving
the same PBL depth and
moisture reduced the growth
less
runs.
in affecting the surface
temperature.
lower PBL
for the dryer
on NA was also important,
order of 65% to the NA of the plain runs.
also kept all
The
in PLI by almost 60%, yielding
instability than the initial values
The effect
The
adding
soil
on the
The reduction
of the runs from saturating at
the PBL top.
in aS
PAGE 172
When dry
changes
advection was
added to clouds and
(GCH) the results were not surprising
The
temperature and PBL depth were
and
the PLI
The
behavior
moisture dependent.
plain run, while
run remained
pattern
run
in
characterized
4. 8:
this
the PLI
and
the
was physically
However,
(Fig. 4. 38),
to 23 .MT.
the
This
The
so it was the only
unrealistic.
the modelling results -for 23 GMT.
i;.n
the NA calculations for most of the runs,
variation in the HC run.
The growth
of the PBL
surface lay.e.r temperatures were relatively
of the physical
depended mostly
surface, and
factors
on the imposed
soil
included.
unaffected
Surface moisture
changes aloft, dry advection at
surface moisture.
instability was affected
except
prior
the
important dif+-erences from the 21 GMT results were
and
the
just
of
for 23 GMT M-odelling
the variability
by all
5-70 run shows
until
All
the PBL top.
saturation much earlier,
Table 4. 7 summarizes
The most
over the
the 30-30 and 30-70 runs as well.
series which
Summary
had no NA at all.
or oversaturated at
subsaturated
50-70 run reached
by 75%
The 5-30 run had a NA 67% higher than the
time variation of the
the
lower,
the NA calculations was soil
the 50-70 run
runs were saturated
as
of
(see Table 4.6).
unchanged from the OH runs,
values were consistently
plain runs.
imposed
significantly
The convective
by all
the presence of clouds, with only
soil
of the factors
surface moisture
PAGE 173
contributing
depend
positively
in a very
imposed
complicated way
changes.
The NA seemed to
on soil
moisture and the
It responded more consistently
presence of clouds
adding
to the amount.
(adding to NA)
and dry
advection
(also
inhibition).
When the effects of the various factors
in
to the
isolation,
but
discussed above
in connection with
change, just as
each
are examined, not
other,
the results
in the previous cases.
particular,
the role of clouds reverses, and their
contributed
to
Symbolically,
run with
in order
clouds and
PLI =
clouds -
less inhibition and more
of decreasing
imposed
importance for the 30-70
changes:
changes + bulk soil
NA = initial- conditions -
BLH -
surface moisture - clouds - bulk
Determining which run was
temperatures as
observations.
(QCH) heated
imposed changes -
None
soil
of the runs with all
to reach
the low end
This was actually
reasonable since the
moisture
closest to a simulation of the
enough
high as 313 K,
surface moisture +
soil moisture
inflow air was somewhat difficult.
of the effects
presence
instability.
initial condi-tions + BLH + soil
imposed
In
potential
of the implied
fortunate and
physically
inflow air reached 313-316 K
only with
PAGE 174
the
help of warm advection which was not
model.
simulated
in the
An estimate of the-ef^ect of 2 K of warm advection
between 21 and
23 GMT was made by
considering the mechanisms
in the model.
The sensible heat flux
from the ground depends on the
vertical gradient of potential
atmosphere and
the ground.
temperature between the
if warmer air
location, the immediate result would
heat flux.
However,
balance, letting
heat the
soil.
this would
is advected
be a drop
lead
in sensible
change the surface energy
some of the available net
This would
over a
solar radiation
to a higher soil
surface
temperature, and an increase
in the sensible heat flux.
the first
be no change
order e-fect would
Thus,
in the sensible heat
flux.
The growth
and the
of the PBL depends on
stability above the PBL.
the effective stability
the PBL would
since
b-e reduced.
If the PBL is warmer,
lead
The PBL would then
available sensible heat
the entrained
Feedback
air up
in the
effect would tend
level
to
the
the
low turbulence of
of the PBL turbulence.
limit the
the
of the PBL
stability, because
flux must bring
to the
grow faster,
still be the same and
However, the faster growth
to an increase
then
inversion or stable layer above
the sensible heat flux would
stability reduced.
would
of an
the sensible heat flux
increased
This
growth rate.
PAGE 175
To include the advection
the
of temperature
in a crude way,
opposing tendencies were assumed to balance, and the PBL
temperature was simply raised 2 K without any
height.
The soundings for the QCH 5-70 run and the modified
GCH 5-70 run
(with the
on Figs.4.39 and
and
temperature advection effect) are
4.40.
the NA reduced
by
The PLI was increased
50%.
It
is safe
in all
to 3.9,
the
of the runs (see Table
suitably modified
temperature advection.
Due to the
available
it was difficult
surface data,
run from the OCH group as being
imprecision
especially
However) since there was no
integration which
most
likely
for the surface
in the
to pick a particular
close to being a
indication of a
strong convergence line at the surface, it was
the
shown
to conclude that the OCH integrations were
reasonable simulations when
simulation.
from 3.1
The effect also removed
overrsaturation at the PBL top
4.6).
change in PBL
expected that
simulated reality would
yield a- nearl-y s-aturated PBL top,
with low NA and relatively
high convective in-sta-bility.
50-70 run was best in this
sense, except that it became
before the
The
supersaturated at
surf-ace advection acted.
Both the 30-30 run and
the 30-70 run gave reasonable results
tops) with an equivalent potential
discussed
may
well
major second outbreak
set
(nearly saturated PBL
temperature in the range
in the observations and very low NA.
explain the
small
19 GMT, long
This result
radar echo at 23 GMT between the
and CDS
(see Fig.4. 19).
More than one
of PBL characteristics had become unstable and needed only
PAGE 176
a small
amount of forcing
this case,
the location and
directlu related
or no
to become convective storms.
dynamic
release the
timing
of the outbreak were
to the modeled parameters.
forcing available to remove
instability.
There was little
inhibition and
The convection occurred when and
where the convective inhibition was reduced to near
convective
instability and
values were obtained
large, one
GMT).
small)
of minimum
the analysis of
The
from 21 GMT soundings and the echoes
are new echoes at
instability
zero.
inhibition shown on Fig.4.41.
Notice that the convection
moderate
by
conclusion is reinforced
This
In
(14.0)
inhibition.
the next map
time
(23
broke out in a region
of
and was contained within the area
Just as
in
the April
case, the
observations show th-e pattern which the modelling results
predicted.
(one
PAGE 177
Table
Run
Type
4. 1:
MODEL RESULTS AT 21
PH
GMT,
9 MAY
GS
PL I
10. 2
10.8
10. 5
11. 2
12. 1
4. 7
5. 2
4. 8
5. 4
6. 0
31. 17
20. 23
42. 38
29. 57
22. 71
10. 5
11.0
10. 5
4.
47. 29
41. 67
53. 22
49. 14
37. 55
CO ND
Plain
5-30
30-30
5-70
30-70
50-70
693
702
705
716
728
30. 5
30. 0
29. 8
710
723.
29. 5
28.
28. 7
27. 9
28. 3
"-'SAT
"SAT
SdT
Early Clouds
5-30
30-30
5-70
30-70
50-70
734
750
?L. .i
11.5
12.5
7
5. 0
.
,
5.2
,. .
-
"SAT
+SAT
+C l I'l
AN
Imposed Changes
5-30
30-30
5-70
30-70
50-70
Clouds and
5-30
30-30
5-70
30-70
50 -70
667
674
676
694
716
30. -.
29. 7
29. 6
29. 0
28. 3
9. 3
10. 3
9.5
10. 3
11.5
3. 5
3. 8
3. 5
9. 9
10.8
10.6
11. 4
12. 3
3. 3
4. 16
4. 3
5. 0
5. 3
23.
6.
28.
14.
4.
45
08
42
41
60
+SAT
+SAT
+SAT
+'SAT
20. 40
6. 53
12. 83
8. 46
7. 44
+SAT
+SAT
'SAT
+SAT
+SAT
Imposed Changes
693
712
714
29. 2
28. 5
28. 4
728
737
27. 1
27.1
Run Type is soil surface moisture.-buk soil
moisture.
PH is
pressure level in mb of top of PBL.
TS, QS are temperature(deg
PLI is
C)
and moisture (gikg) at top or sur face iayer.
instability index (deg C).
NA is necative area above PBL
which a parcel must overcome to reach the level o, iree
convection (m**2/g **2).
PAGE 178
Table 4. 2:
CLOUDS
IN MODEL RUNS- FOR iMAY CASE
IMPOSED
Press ur e
1!
800
825
850
875
80 %
50 %
70 %
70 X
0 %
rmbb
mb
mb
mb
GMT
Percentage cloudiness used
levels not listed
had 0 %.
Table
Pressure
600
625
650
675
700
725
750
775
800
825
850
4. 3:
12-14
T
0.0
-0. 4
-0. 6
-0. 8
-1.0
-0. 6
0.0
0.0
0. 4
0.0
0.0
14 GMT
17, 20 Gi"IT
0 %
0%
o0
in
the early cloud
runs.
IMPOSED CHANGES ON MODEL RUNS FOR
GMT
G
0.0
0. 2
0. 4
0. 6
0. 4
0. 4
0. 0
•-1. 0
-3. 4
-1. 4
-1.6
14-i'7
T
0.0
0. 0
0. 0
-0.
-0. 9
-1. 2
-!
-
-0.
-0.
0.
0.
6
6
0
0
P r essure
MAY CASE
17-20 GMT
GlMT
G
0.0
0.0
0.0
-0. 9
-1.5
-1. 5
3.0
4. 5
1. 5
0.0
T
0.0
0.0
-0. 3
-1.2
0.0
0.9
1. 2
0.0
0. 0
0.0
0.0
0. 0
0. 0
2. 1
2.1
3.0
3.0
0. 9
0.0
0. 0
0. 0
0. 0
Imposed, changes in temperature (deg C) and moisture (g/kg)
between sounding times.
Imposed linearly in the model with
time.
Pressure levels
not listed
have zern changes at all
times.
Times not covered also have zero changes.
PAGE
Table 4. 4.
Changes
Early
in:
SENSITIVITY VALUES FOR 2i
MODEL RUNS
PH
@S
OMT,
i79
9 MAY
PLI
NA
-- 6
+37
clouds
-11
Imposed
- 9
*i
changes
+10
-52
Both clouds and
impose d changes
0
-1
t
+ 8
to - 6
to - 6
Soil
Bulk
surface moisture
- 6
soil
moisture
- 7
-
6
-62
-21
-11
-.39
+11
+22
Sensitivitu measured as percentage change of a given variable
compared with maximum amount of change in that variable after
application of physical parameter.
Variables as defined on
Table 4.1.
Application of soil surface moisture defined as
increase from 5% to 30% of saturation.
Bulk soil moisture
application defined as increase from 30% to 70%. on saturation.
PAGE
Table 4. 5:
Run
Type
MODEL RESULTS AT 23 GMT,
GS
PH
PLI
9 MAY
NA
CONDI TION
PI a in
5-30
30-30
5-70
30-70
5'0-70
Eariu
9. I
10. 4
711
30. 2
29. 4
687
697
701
716
733
10.9
11. 9
4.7
5. 2
4. 9
5. 4
6.2
12.
17.
23.
15.
10.
47
28
95
38
21
' SAT
+SAT
"SAT
+SAT
+SAT
30. 7
30. 1
29. 9
29. 1
28. 0
10. 0
10. 6
10. 4
11.2
12. 3
4. 6
5. 0
4. 7
5. 3
6. 1
30.
23.
32.
30.
20.
63
2441
47
30
31.
30.
30.
29.
29.
9. 1
9.6
9. 4
10. 1
11. 1
-
4. 1
3. 7
4.4
5.3
15. 12
10, 29
15. 17
5. 72
"::0.0
+SAT
+SAT
+SAT
3. 9
4.4
4. 2
5.0
6. 0
12. 67
4. 48
7. 27
0. 30
":O. 0
+SAT
+ AT
+SAT
+SAT
+SAT
Clouds
5-30
30-30
5-70
30-70
50-70
Imposed
"'SAT
"SAT
+SAT
Changes
5-30
30-30
5-70
30-70
50-70
Clouds and
655
661
663
670
686
Table
1
6
4
9
3
-7
+SAT
+SAT
Imposed Changes
5-30
30-30
5-70
30-70
50-70
See
31. 6
31. 1
671
681
685
4.1
665
672
676
697
717
30.
29.
29.
29.
23.
for
legend.
3
8
6
i
3
9. 6
10. 1
10. 9
12. 1
180
PAGE 181
Table 4.6:
Type
Sur:face
MODEL RESULTS AT 23 GMT, 9 MAY CONTINUED
oGS
PH
Moisture
5-30
30-30
5-70
30-70
50-70
PLI I
671
681
685
697
711
31.6
31. 1
30. 9
30. 2
29. 4
8. 7
9. 3
9.1
9. 8
10.8
664
672
674
695
718
30. 3
29. 8
29. 7
29. 1
28.2
8.6
9.3
9. 1
10.0
11.2
OCH Modified for Surface Temperature
5-30
30-30
5-70
30-70
50-70
See
Table 4.1
664672
674
695
718
for
CONDITION
Advection
Surface Moisture Advection, Clouds,
5-30
30-30
5-70
30-70
50-70
NA
32. 3
31.
3.1.
31.
30.
legend.
8
7
1
2
8.6
9. 3
9. 1
10.0
11.2
3. 4
3.9
3.6
4. 1
5. 0
and
39. 08
43. 85
55. 62
46. 42
33. 48
Imposed Changes
2.7
3. 1
3. 9
5.0
37.93
27.93
30. 86
14.01
0.50
Advection
3.6
4.2
3.9
4. 7
5.8
18. 45
7. 62
15. 43
1.31
0.21
(OCH)
#,SAT
+SAT
+SAT
+SAT
+SAT
PAGE 182
Table 4.7:
Changes
in:
SENSITIVITY VALUES FOR 23 GMT
MODEL RUNS
9 MAY
PH
TS
GS
- 8
- 8
-10
-
6
+42
+10
-
+21
-47
+18
to -I100
+ 5
-26
+2
to -100
PLI
NA
Factor
Early
Clouds
Imposed Changes
Clouds and
4
Imposed Changes
+ 4
-10
to - 3
Surface Moisture Advection
0
0
Clouds,
5
+30-
Imposed
Changes, and Surface
+27
+ 4
-10
to - 3
Surface Soil Moisture
-5
Bulk
to -
Soil Moisture
- 6
-59
+65
Moisture Advection
+67
-75
to -95
.-
5
-16
+30
-
7
-10
+14
+ 3
to -'59
to -44
See Table 4. 4 for legend.
PAGE 183
1URNAUOLS REPURiED SY -BSERVLRS
'
'
sux
*%0 5 5 7
TORNADOES REPORTED BY .RADAR
FUINEL CLOUDS REPORTED BY OBSERVERS
CV
555
HAIL WITH DIAMETER SIZE Il INCHES
0
AL
31RI
05z
STRONG STRAIGHT LINE WINDS
0
SEVERE THUNDERSTORm INDICATED BY RADAR
I
OE'N
0467
OW
0
SQUALL
COS
P.
SKF 0
SL
QSLN
O
04s4
ZONE
O.x
AL$
Cqu
0441
04S2
TAD
0460
p.
0
SAF
NO
0
ODHT
o18vo
OTUL
0a"
LVs
0
MKO
0
o357
CSOH
357 TIK
QHB"R
NLC
0
0 LTS
.OOM
0 PVW
02ROW66
0
FSI
DUA
0
LI67
02s?
0FWo
SCH
MWL
O
ABI
8
80
0026
YS
'GOP
DAL
GSW 0 0256
OFrw NBE
259
0 S'P
2;o
OWAF
265
0 BWo
OSJT
263
0262
O
Fig. 4
1
RFP
Severe
T
convective
AOCT
257
OGRK
OJT
eriod
eather;F- vents drring
10 ;,7.
ul.' 1s one rTe
out.-reaks during 24 hour
256
OTPL
OCLL
2OUS
0..au
204U3s
12 GMT, 9 May
ers to are op
period.
CTYR
Fig. 4. 2
Synoptic-scale 500 mb analysis for 12 GMT, 9 May.
Solid lines are heights (dm), and dashed lines are
vorticity (10**-5 1/s).
s 82,
Fig. 4. 3
Same as Fig. 4.2 for 00 GMT, 10 May.
20
\
Fig. 4. 4
Synoptic scale surface analysis for 12 GMT, 9 May.
Solid lines are sea level, pressure (mb), leading 9 or
10 digit(s) dropped and dashed lines for 1000 mb to
500 mb thickness (dm).
p
Fig. 4. 5
Same as Fig.4.4 for O0 GMT,
10 May.
o0
PAGE
188
r
-V-
r
-1
*469
*1553
*562
L
I":
*
53 21
456-- II
456.
,
i
t
-- I
----
* 349
_
,- '-
-I
I
*433
I
*451
__
r -L
*365
-r
I
*354
-- - *327
l
_.-'340U
*229\
260*
*265
%270
247*
/ e235
"
S*240
-I
'T 26
-232
"-
,
255
GAG
29
OSEL
36
CAN
22
0
SIM
.37
CDS
25
CHE
23
OHEN
31
SUD
OTVY 0 38
HNT
33
CSM 032
0 OUN
26 26 TV OCHK O 35
34
24
ADA
LTS
OFSI
EMC 020
21
28
27
HEA
0 30
Fig. 4.6
Locations of radiosonde launch sites for this case.
-"nJ
C,
%-t
4
\iV
vv
-ID,
Fig. 4. 7
500 mb analysis using special network of Fig.4.6 for
11 GMT, 9 May. Solid lines for temperatures (deg C),
and dashed lines for mixing ratio (g/kg). Winds
plotted conventionallu (knots).
-15
Its
/I
0
L00
UMmN
4VA
0
-C
*KC
-1
a,
-
SAP
~
/
a
^04F
7CT
-/0
vtr
,4
4I
W4
A
.r*~r
A
7
~
bM
u
of e0 B
7
0neCe
wup0E
00 3o oy
UIOJ (U0
s.u
I
ue
Paqsep P
seurl
t
su
Z
.c8
pl~
) stiu V~
~ asUP
.d
6 't -
Drflt-
LCi
se
1
a.
1'
4-14..
4. 0
o00
'
Fig. 4. 10
Same as Fig.4.9 for
L
C
Li
at
0
m=
Ct)
Fg. 4. 7 for 700
b.
UyA
b
3
TJ
.'77
k
O
is
3O'
S
12
Samee'as
4
--
9YCi
1o0.
/
"I,
aVA
.4'
o
5~-
L--
/
0
w.
I
qo
t
.*/
/
I
.
'I
i
(
-
a
I
g
8
4.
measFi.
or 70o
mb..
lA
iA
vey
+
6LPC
TiC
It
UU
I
cu.
9 . 4 *1 3
S
! 0
me a
8
1
2-s-s
.4)
0-1
i'
i.
4. 14
Rame as F
.
4 .
13 fPo
20 to 23 QMT.
PAGE
Fig.4.15
197
Mesoscale surface analysis for 12 GMTI 9 May. Solid
lines are surface potential temperatures (K) and
dashed lines are dewpoints (deg C). Winds are plotte
conventionally (knots). Sky condition is clear (open
circle), scattered (single bar), broken (double bar)
overcast (filled circle), and obscurred (x in
circle). Cloud type is plotted (if available) as is
current weather accordina to conventional synoptic
code. Radar echoes are cross-hatched irregularly
shaped areas. Surface cold front is depicted by a
line with triangles along it.
PACE
198
e~q!
_Q4-
jv
ae~S9~
'"4~Q
'LDr
I
Fig. 4. 16
Same as Fig.4. 15 for
18 GMT, 9 May.
PAGE :t99
Fig. 4. 17
Same as Fig. 4. 15 for 21 GMT, 9 May.
PAGE 200
OKC
0
O 0
0
0
-'
O
I
0
0
0
00
o00
I' /
0 o
Fig. 4. 18
Change in potential temperature and dewpoint from 12
GMT to 21 GMT, 9 May. Solid lines for potential
temperature change (K) and dashed lines for change in
dewpoint (deg C).
PAGE 201
Fig. 4. 19
Same as Fig. 4. 15 for 23 GMT, 9 May.
PAGE 202
LE
a Sa 9
0o0
@00
000
000
On .
30
-o
a
o
osa
00
OO
000
\
-
1Z
Fig.4.20
Photograph of low-elevation angle display from radar
screen at Amarillo, Texas at 2242,2247, and 2254 OMT#
9 May. Range rings are 20 nm apart.
0
pau.
'T
0,'
f
{o.
0
0
o00i
0000
6 o
PAGE 204
Pressure (mb)
400
500
700
1000
-40
-30
-20
-10
10
20
30
Temperature (deg C)
Fig.4.21
Sounding plotted on a pseudoadiabatic diagram from
Shamrock, Texas for 1143 GMT, 9 May. Solid line
connecting dots for temperature (deg C), solid line
connecting * for dewpoint (deg C), dash-dotted line
for 313 K isentrope, and dotted line showing moist
adiabat for mean PBL parcel (or selected parcel if
PBL is not well-defined) Dewpoints colder than -40 C
are plotted at -40 C.
4L
PAGE 205
Pressure (mb)
400
500
802
-40
--308
-20
-10
0
Temperature
Fig. 4.22
(deg C)
Same as Fig.4.21 for 1705 GMT, 9 May.
PAGE 206
Pressure (mb)
500
600
900
-40
-30
-20
-10
10
30 '
40
Temperature (deg C)
Fig. 4. 23
Same as Fig. 4.21 for Amarillos Texas 23 GMT, 9 May.
PAGE 207
Pressure (mb)
400
BOO
88
-40
-30
-20
-10
0
Temperature
Fig. 4.24
10
20
30
(deg C)
Same as Fig. 4.21 for Childress, Texas 2006 GMT, 9
May. Missing dewpoints between 825 and 725 mb are
plotted as -40 C.
48
PAGE 208
1oo
Fig.4.25
Depth of nearly dry adiabatic layer (mb) between 319
and 317 K isentropes, 11 GMT, 9 May.
PAGF 209
00
0
oo
0
0
0
oo
0.
0
0 0
90
00
00
0
0
O
0
00
Fig. 4.26
Same as Fig. 4.25 for 20 GMT, 9 May.
PAGE 210
Pressure (mb )
500
600
700
-40
-30
-20
-10
0
Temperature
Fig. 4. 27
10
28
30
(deg C)
Same as Fig. 4.21 for Oklahoma City, Oklahoma 20 GMT,
9 May.
PAGE 211
Pressure (mb)
I
U
500
600
-
•
700
""
-
-'
sess
a
ssi
n es
.
S.
-5
n
s
ssi
ns
s
ali
ns
a
sia
t
a
situ
as
alan
ass
taU
900
1000
-40
-30
-20
-10
0
10
Temper ature (deg
Fig.4.28
20
30
C)
Same as Fig. 4.21 for MAYHYB sounding at 11 CMT, 9
May.
40
PAGE 212
TS NR
QS
30
29
28900
27 800.
14.0
26 700
25 600
24 500
13.0
12.0
11.0
23 400
22 300
10.0
9.0
21. 200
20 100
8.0
19
7.0
0
PLI
6.0
L
I
I
L
I
I
I
I
I
I
I
I
- 1
I
I
I 1
12 13 14 15 16 1"7 18 19 2021 22 23
TIME
0
PH
650
700
5.0750
4.0
PH
800
3.0
850
2.01
Fig.4.29
Time variation of model output for MAYHYB sounding, 9
May, 5-70 soil parameters, with no extra factors
modelled (Plain). NR is net radiation into surface
(mcal/sq cm mih ), TS is temperature at top of surface
layer (deg C), GS is PBL mixing ratio (g/kg), PLI is
convective instability (deg C), PH is pressure at top
of PBL (mb). Condition at top of PBL is indicated
above time axis: blank = unsaturated. "S = nearly
saturated, S = saturated, S+ = oversaturated (P).
PA G
2:13
TS NR
QS
30
29
28 900
27 800.
14.0
26 700
13.0
25 600
12.0
24 500
11.0
23 400
10.0
9.0
22 300
21 200
20 100
19
0
PLI
6.05.0
8.0
7.0
I
'
PH
I
1
I
I
I
I
I
I
I
I
1
I
1
I
I
3
I
1
I
1
12 13 14 15 16 17 18 19 2021 22 23
TIME
650
- --- 700
91-7
750
4.0-800
3.0
850
2.0
Fig.4.30
Same as Fig. 4.29 for run with morning clouds imposed
(C).
PAGE r14
TS NR
QS
30
29
28 900
27800
26 700
25 600
14.0
13.0
12.0
24 500,
11.0
23 400.
22 300
10.0
21 200-
8.0
20 100
19
0
7.0
PLI
6.0
9.0
PH
650
-
700
5.0
750
4.0
800
3.0
850
2.0
Fig. 4. 31
Same as Fig. 4.29 for run with imposed changes above
PBL (H).
PAGE 2-i. 5
TS NR
QS
30
29
N
28 9002780026700
25 600
- 14.0
-13.0
-12.0
-
24 500-
11.0
10.0
23 400
22 300
//
21. 200
8.0
20 10019
0
PLI
6.0
5
I
9.0
IiI
Ii
7.0
sPH
11
1
650
12 13 14 15 16 17 18 19 20 21 22 23
TIME
700
p
5.0o~
1t
1
750
4.0°
800
3.0
850
2.0
Fig. 4. 32
Same as Fig. 4. 29 for run with both morning clouds and
imposed changes (HC).
PAGE 216
(mb)
Pressure
_
__
__
___
_
_C
__
500
608
888
908
1888
K
t i,,
It. 1.1111
1111 1111
IIRII
1 .,
nil
,l,,t
.....tilltlllit
II 11111
III11.111
1tI11.1L11I ILII, 1.,,1,i
Itlilit
-48
ll
-38
I
-20
-18
0
18
l,1111
38
Temperature (deg C)
Fig.4.33
Same as Fig. 4. 21 for model output a t 21 GMT from
MAYHYB initial soundings 5-70 soil parameters, and no
additional factors modelled (Plain) . Negative area is
cross-hatched (P).
PAGE 217
Pressure (mb)
700
800
1000
-30
-20
-18
10
20
Temperature (deg C)
Fig. 4.34
Same as Fig. 4.33 for model with both clouds and
imposed changes (HC).
PAGE 218
TS
QS
30
29
MDEL-
28.
'TS -
27
14.0
/e.r
26:
25
-J
13.0
12.0
11.0
_OS_,
)
24
23'
22
10.0
9.0
21.
20
8.0
7.0
19
'
I..
-I
II
I
I
I
..
]t..
e•
I
I
I
I
n=
I
iI
,
• ii
i
I
1
I-- -
I
12 13 14 15 16 17 18 19 20 21 22 23
TIME
Fig.4.35
Comparison between 5-70 HC model run and surface
observations taken from analyses. TS, OS as in
Fig. 4. 29.
PAGE 219
Pressure (mb)
__
I _
_ __
L
988
10800
ti ll
li tLI1IttIll
------~--------
,
--~-I------- T--- lIII
-t------- I~II
; -1~11111111111111111
--*------T I--;--- /!Illllllllllll(
- ------ --- - '--1~~111~1
-20
-10
0
10
Temperature (deg
Fig. 4.36
Same as Fig. 4. 21
Mau.
for Amarillo,
40
38
C)
Texas at 17 GMT,
9
PAGE 220
0
O O
Fig.
4.
37
Mesoscale analysis of convective instability
and convective inhibition (NA) for 17 GMT, 9
Solid lines are PLI (deg C) and dashed lines
NA (m*.2/s12
s).
Cross hatchina is new radar
Numbers in
appearing between 18 and 21 GMlT.
are point valuec o.F PLI, circled numbers for
(PLI)
May.
are
echoes
boxes
NA.
PAGE 221
TS NR
QS
30
29
28
27
14.0
26
25
24
13.0
12.0
11.0
23
22
10.0
9.0
21
20
8.0
7.0
19
PH
650
700
750
800
850
'i
a
as
r-hances,
i.
.r
.l,
aud
:rface
run
t
moistur
-
clouds: imp o;e d
advection (GCH).
PAGE 222
Pressure (mb)
400
58800
880
900
108B8
-40
Fig. 4.-"
-30
-20
Sa e e_s
imp ose :
( ,'CH)
-10
0
10
Temperature
(deg
48
C)
Fig. 4.33 for run at 23 GilT with clouds,
chang es,
and
sur-.ce
moisture
advection
PAGE 223
Pressure (mb)
I
I
-
't- -
---
-~
'
I
500
600
888
900
1I
;I
I)
II .L
trtl
.........----
-40
-30
I
1
1
1LL-III I II -r I..-I ) I I I I I II I I II)
-28
1
I I
-10
Temperature
Fig. 4. 40
11111111111,,,,I,,,,
,It,ttI
.--
,,,2,1
L
30
(deg
C)
Same as Fig. 4. 33 for run at 23 OMT with clouds,
imposed changes, sur face moisture advection, and
modified for surface .emperature advection (QCH
mnodiiied).
lt.T.
PAGE 22..
co
6
, SID
0
/ f~
TCC
OKC
S0
0
0
0
0
10
oO
0
Oo
0
Fi .4
.1
Sa
im
rT'onr
a;
2.1
Fiq. 4.37 fi.r
to
23 G T;.
20. (iT.
0
0
0
New radar
echoe-
are
PAGE 225
CASE STUDY
5.1:
with a large number
tornado
the previous two,
The
(e.g.wail
was
which
formed
were
this
funnels
sighted,
size
the SESAM E region.
just northeast of Oklahoma City,
Oklahoma
(OKC) around
18 GMT, and
The
synoptic
factors
modeling Tresults are
SYnoptic
At
High
The
5 00 mb,
shown and
to the southwest
From these cells.
and
soundings,
18 OMT before the convection, are
15 and
5.2:
formed
are discussed first,
is documented.
mesoscale analysis
more formed
line
not
Nevertheless,
cells formed
a squali
but
analysis did not
isolated
Eve-ntualiy
1. 5
was that the
case
any
19 GiMT.
although
was
indicate
by
in
No
substantial,
Morning
out uneIxpectedly.
convergence
were
Maximum hail
interesting a-spect of
convection broke
was not associated
severe hail.
clouds)
seen on radar.
thunderstorms
The
severe.
of tornadoes or
precursors
a mesocuc lone
cm.
Jurne
Introduction
This case, unlike
or
6
iII:
then the
available
discussed and finally
interpreted.
Analysis
June,
Piains area of
12 OiT
the United
(Fig. 5. i
States..
at
a
with
trough
an
lay
in
elongated
the
PAGE 226
vorticity
weak
maximum co-located with
or zero
had
The weak,
occurred.
all
of Texas and
spread-out vorticity
km.
for 12 GMT, 6 June
surface, the map
only weak
Oklahoma,
south-southwest geostrophic
Increased
north
to a low center
(west of the
There was still no
in Canada.
and
organization of the synoptic-scale
and
both at 500 mb
discontinuities
well-defined
500
the geostrophic
south-southwest, stronger than before.
still
the analyses,
(Fig. 5. 3),
over CO, as an extension of a trough which
well-defined thickness gradient nearby,
flow was
flow
flow over
At 00 GMT, 7 June (Fig. 5.4) a low developed
stretched
The
thickness gradient within
with no well--defined
CVA at 500 mb)
5.3:
border.
southwesterly over Oklahoma.
At the
showed
had become
localized with weak CVA extending northeast
more
from central Oklahoma to the Kansas/Missouri
still
Oklahoma at
a significant change
(Fig. 5.2),
By O00 GMT, 7 June
stronger and
was
CVA was
There was
eastern Kansas.
flow over nearly
southwesterly
this time.
axis.
throughout most of the SESAME region with the
of the weak CVA in
largest
the trough
in
the surface
the wind
suggesting
Mesoscale Analysis
field or
that
synoptic-scale
in this
features occurred
case, but
thermal
there were no
field
in any of
low-level convergence along
front was not
present.
a
PAGE 227
The anailyses of the mesoscale
were
fields of
performed at 500 mb and 700 mb at
available data consisted
in Oklahoma, shown
in Fig. 5. 5.
and
some cases
in
at
any
in
clearly
error.
15 GMT reported heights which
af the
(CHK) and Altus, Oklahoma
the
18 GMT.
The
15 stations
The data were noisy,
Fort Sill,
Oklahoma
(FSI)
were about 6 dm lower than
surrounding stations.
other stations. and
15 and
of soundings at
mostly
upper air data
Winds at Chickasha, Oklahoma
(LTS). were much
stronger than at the
were not associated with
any discernible
.height gradients.
Some conclusions could
the curvature of
maximum, noted
Oklahoma at
be drawn from these data.
the streamlines as a guide, the vorticity
on the synoptic-scale analyses remained west of
18 GMT.
Indeed, at
15 GMT there was more evidence
for a mesoscale vor-ticity maximum as
indicated
by the
curvature of the streamlines on Fig. 5.6 than at
Fig. 5.7.
Oklahoma
The wind at Gage, Oklahoma
(HEN) veered
the flow at
of Oklahoma
uniformly from the
5.7.
contribution
to
18 GMT,
southwest.
The
This cooling aloft represented a
convective
in the
temperatures
by comparing
instability
over
the
convection broke out near the edge of the strong
rather thaTn
leaving
(including the
dropped noticeably at 500 mb as can be seen
Figs. 5.6 and
18 GMT on
(GAG), and Hennessey,
at 500 mb between 15 and
18 GMT over most
central part)
Using
center of
region.
The
cooling aloft
the region of cooling.
PAGE 228
The 700 mb
field
same tendencies for
generally
temperatures at
Childress,
The convection began at
12 GMT, 6 June,
in
the
in the region from
cool air,
cooler air
(HBR),
in this air.
To the
the dewpoints were above
they were 3-5 C lower.
Weak radar
No stations reported
or even any visual
of cumulus clouds.
At 15 GMT
(Fig. 5. 11)
the cool
existence, with
slightly
winds generally
from the south
Oklahoma although
(SPS) skies
pool was
still
higher dewpoints of
stratocumulus clouds were still
potential
mostly from the south
northerly
thunderstorm activity however,
sightings
of
covered and
(CDS) northeast to Hobart, Oklahoma
echoes were located northeast of OKC.
any
the edge
the mesoscale surface
surface were weak,
southeast of this slightly
but
the coolest 500 mb
A. cool pool was located
Texas
increased
Notice the very moist conditions
the winds were calm or weak
20 C,
temperatures
that Oklahoma was mostly cloud
Winds at the
or southeast.
and
the
air.
On Fig. 5. 10 at
analysis showed
5.9) showed nearly
field,
same region as
18 GMT.
this moist 700 mb
and
the wind
over Oklahoma.
which were in the
foggy.
(Figs. 5.8 and
or southeast.
in
18-19 C,
and
Stratus or
covering most of
the
state of
just over the border at Wichita Fails, Texas
had cleared and BLH had
temperature of
begun
5.7 K between
(increase
12 and
in
15 GMT).
PAGE.229
18 GMT
At
near OKC. and
(Fig. 5. 12),
towering
cumulus
clouds were
The
centered on LTS, although
it was weaker
cool
strong
the
area was still
pool
southeast near FSI and
over Oklahoma, and
Notice that no
south.
than the warmer region to
drger
especially SPS.
the precursors to
At 19 GMT, shown on Fig. 5. 13.
squall
still observable,
in Oklahoma at this time.
convergence line was observed
the cool
and
than at 15 GMT.
at the surface was occurring all
winds were generally moderate from the
at FSI1
observed
pool was
Ardmore, Oklahoma (ADM).
Heating
echoes had appeared
scattered weak
the heavy
line formed north of FSI and southwest of CHK.
This
broken line of radar echoes extended northeast past OKC, and
was visible from Tinker AFB.. Oklahoma
seeing a
line of cumulonimbus clouds east and southeast.
was reporting towering cumulus clouds.
convection had a potential
Warmer temperatures were north
on,
as seen on the radar sequence
north
of FSI
and
strong storms,
The
cooler and dryer.
(not shown),
the
to a cool
temperature
trough by
deficit to BLH.
Later
storms
developed
of which formed a mesocyclone after
cooal pool was reduced
lost much of its
Notice the
OKC, TIK, and SPS.
southwest of CHK eventually
one
to the
of HBR and Clinton Sherman AFB,
while LTS was still
(CSM),
inflow air
to 22 C.
the observations at FSI
uniformity of
The
FSI
temperature between 305 and 306 K,
the surface of 21
and a dewpoint at
Oklahoma
(TIK) which reported
into
21 OMT.
19 GMT, having
PAGE 230
Fi
and
.
14
shows
dewpoint between
temperature
than
10 K,
McAlester
K rise.
LTS, the
rise,
at
temperature.
the
least
than 11
this
less
period
in terms
large rise
about
in
in
in
i C.
This
large change
of surface equivalent
be addressed
detail in
from
less than 7. 5
The dewpoints
suggested a
in
potential
connection
soundings.
Advection was unimportant
had an anomalouslu
the
K.
in
Nelw
tongue extending northwest
during
This will
changes
Plains region of
cool center, had a verg
slightlg
temperature
Most of Oklahoma warmed
in addition to heating,
instabilit,
largest
to OKC and TIK showed
temperature, more
Oklahoma rose
The
the High
and Texas.
(MLC) up
in potential
19 G1MT.
and a significant
ADM and
potential
changes
i2 and
occurred along
Mexico., Colorado,
with
the net
high potential
gradient between SPS and
in this analysis.
convective
The
considered
in
of the Texas
temperature and
the Oklahoma
parcels arriving
outbreak region
advection was weak.
north
The
came
the
ignored
initial
of SPS, where
impact of advection will
connection with
SPS
dewpoint) so
border was
in the
from east
border.
be
soundings.
5. 4 Soundinns
The
special
shown on Fiq.
.
radiosondes were
at
a ni
18
launched
7 prior
to
from the sites
convection.
lo
12
PAGE 231
GMT soundings were
the regular OKC
launched from these
were
also plotted at
12 GMT.
and Dodge City,
sounding at
adiabatic
less stable than a
500 mb.
stable layer of 20 mb.
layer from 650 mb
to 600 mb,
layer from 600 to 500 mb.
showed
a variation of this
to 565 mb and a st-able
The UMN
capped by an additional
The plot
of the SEP
pattern with one
dry adiabatic
layer above the inversion
layer above 800 mb,
up
lagyer from 540 to 585 mb with
inversion
from 490 to 500 mb.
inversion or stable layer.
varied
from 500 mb to
no dry adiabatic
UMN sounding
All
at
sounding
of these
The height of
575 mb.
layers,
sharp
soundings had
this
to
an
stable layer
but the DDC sounding had
two and the
had one.
the surface.
had
It had a 30
The OMC and SEP soundings had
of these 12 GMT soundings had
or near
a simple,
layers form the surface up
moist adiabatic
to 535 mb.
and a second dry
adiabatic
All
sounding
inversion from 575
The DDC sounding was different in other respects.
nearly
At
12 GMT had a similar structure, but had a dry
stable
mb
(DDC)
(Fig.5.15)
from the surface to
lapse rate,
there was a small
Kansas
The OKC sounding
showed a very smooth structure, slightly
500 mb
as.a
Soundings from Stephenville.. Texas
(UMN),
Monett, Missouri
moist adiabatic
Additionally,
12 GMT sounding was available, as well
1312 GMT sounding at HEN.
(SEP),
sites.
saturated
Above
layer
the
high
surface,
Above
500 mb,
relative humidity
only
alJ
the UIN
o.
the
PAGE 232
soundings were
layer
except for OfKC which
quite dry,
just above 500 mb.
1312 GMT
showed
sounding.
way
features
Aside
temperature
The pict
to
similar
from a low level
It had an
to SEP, with only
Just as
very
had a sharp
upper
in the
just
above 600 mb
dry adiabatic
the
the
similar
layers at 860, 800
these
speaking
Using
GMT.
initial
air analyses, changes
of the
All
PIL from the
were
inversions
sounding
the
was clear that
examining
the
surface
layer
to 915 mb.
top
o:
changes
the
(Fig. 5. 15),
which
preceded
of verq
convection.
With only
it
is
the
and
depth
shaillow
it
had a
At EMIC,
were
in
15
superadiabatic
a well-mixed
PI3L above
values and
the values at
possible that
the coarse
surface
entreiy
FSI,
hence were useful
the sounding apparently
the FEL,
assuming
12 GMT
18
gone by
Oklahoma (EMC) and
of the convection, and
(Fig. 5. 17),
or
the PBL top.
Two soundings, Elmore City,
upwind
Generally
inversion between 870 and 885 nib at HEN
(Fig. 5. 16) was above
located
and
GAG
separating an
sharply reduced
smooth as OKC at
was as
except
free atmosphere.
as a guide,
the HEN sequence
in
occurred
15 GtMIT soundings
inversion between 800 and 900 mb,
well-mixed
apparently
the
the O\C 12 GMT*
of
inversion at 885 mb,
inversion
small
the soundings as well.
up
those
from HEN at
585 r.b.
and
GMT
the soundinQ
followed the 334 K moist adiabat all
nearly
to 400 mb.
of
had a moist
PAGE 233
resolution
the surface
above
An
the well--minied PBL's beginning
sounding missed
layer.
inversion capped the PDL between 915 and 875 mb.
Another
inversion was
between, the
located
between 600 and
lapse rate was conditionally
535 mb.
unstable, and the
relative humidity was over 70% for nearlu the whole
existence of clouds,
600 and
Fig. 5. 11
there should
and
630 mb,
suggested
overcast,
that cloud
which agreed
with
cover was
301 K,
mixing ratio of 16 g/kg)
sounding was 3.3.
to 895 mb,
and
Fig. 5. 18).
the 500 mb
still
had
The
this
indications.
(potential
When
temperature of
became saturated at 905 mb.
to 775 mb.
The PLI
for this
the well-mixed PBL extended
inversion was
up
(see
and
gone as well,
Clouds were
the surface observations, and
the relative
When
and
was 4. 9.
7 hour period.
almost
indicated the
lifted
870 mb
PLI
dropped
taken place, so the convective
during
5 tenths and
1. 5 C.
the sounding
at
between
from
reports
inversion had disappeared
temperature
saturated
20 mb.
18 GMT,
level
the PBL. top.
about
By
The upper
humidities in
were
buoyant up
the lower
indicated by
above
PBL parcels
clouds present
Surface
the sounding
adiabatically,
negatively
have been
near 915 mb.
lifted
and were
sounding:
relative humidity above 75% implies the
Assuming that
between
In
presence of clouds
adiabaticaiiY,
PBL parcels
were negativel y buoyant
Dy
19 G1T,
instability
Notice that
the
for
more
heating
had
increased
mixing
rat-io
onl
had
in
PAGE 234
the PIBL decreased
The
slightly
convective
northeast
of FSI.
suggesting. either
outbrek
triggering mechanism
15
GMT FBI
sounding
sounding showed
It
600 mb.
mb
temperature
lifted,
top),
for
By
mb,
and
surface
(Fig. 5. 19)
were negatively
than to
at 865 mb,
ratio
of
at 920 mb
layer at 650
(potential
14. 5 g/kg)
were
(probably the PBL
up
18 GMT, FSI had a well-mixed PBL extending
lower
inversion was
layer was delineated by
lapse rate,
600 mb,
the
The temperature
and a strong
FSI,
the 500 mb
adiabatically,
mixing ratio
negatively
at
gradient
(potential
of 14.5 g/kg)
became
buoyant for
50 mb.
in mixing
the upper
dropped
at EMC.
temperature
A
traversed a cloud
500 mb
PBL parcels
Fig. 5. 20).
level
When lifted
of 303. 5,
saturated at 850 mb,
sounding
between
only 0.5 C at
temperature
This
to 875
sounding, with a
Jwhich over-emphasized
inversion.
unlike
(see
gone
Aloft, the sonde apparently
565 mb and
The PLI
3. 6.
sounding 'as
the
The
another at
and
to about 800 mb.
buoyant up
The
the east.
defined PBL.
When PBL parcels
mixing
the
destabilize or that a
wet, with one cloud
near 885 mb.
to
just
out around FSI
a poorly
had
inversion
became saturated
superadiabatic
ratio.
to
later at FSI
acted
of 303. 5 1K
they
and
this
slower
also was fairly
and another
19 GMT was
convection broke
was
a strong
at
region
Later,
FSI
18 GMT at EMC.
between 15 and
was
clearly
and
more
PAGE 235
stable
than that
convective
iess
at
EMC at
inhibition.
18 GMT from
The FLI
than at EMC.
instabilitg
near 20 OMiT
place untJ.il
f1or
standpoint of
the
this
sounding
IMore heating
convection
when
of
was 3. 9,
this
began to
air
tool
spread
to the
FSI area.
As
happened
decreased
absence
surface
in moisture
latent heat
example
.luxes.
of this
it
should
have
in
ratio
the
difficult to re concile this
It is
of
analtjses.
change
in the surface
It
well
may
the
have been that
lower part of
the
that must have existed
to
The
18 GOT was a good
sounding at FSi at
(Fig, 5. 20).
advection will be
The question
give upward
of moisture
addressed again when the modelling results
discussed.
For
comparison,
the regions north
at
particularly
in Fig. 5. 21.
from the
and
Notice
L
up
interestinq
south
of
It
examine
convective
that the PBL apparenti
tothe
scattered
only
the
to
18 GMT sounding
The
base
sur.ace values
the
morning.
it is
18 GMT.
surface
case,
reporting
all
as
surface measurements were sampling
gradient
this
heating,
the apparent lack
the
the PBL mixing
sounding,
moist advection.
of
values from
are
the EMC
the PBL
during
decrease with
the
for
should
sky
have
of a s al
seem too
conditions,
lo',
in
area,
at CSM is
shCown
exennded
i00 mb
inversion.
In
since
and
soundings
had
had a superadicabatic
CSM was
been warming
surface
PAGE 236
layer
by
this time.
The PLI
much negative area remained
boundary
the
SPS at
17 GMT
inversion ca.pping
buoyant
had a
the PIL.
1. 7 and
Parcels from the
for almost
(Fig. 5. 22) had a PLI
significant amount of
The CSM sounding
sounding was
above the PIBL.
layer were negatively
south,
had a
for this
00 mrb.
of 4. 2,
To
and also
negative area to overcome.
deep but dry PBL, and
still
SPS was characterized by
shallower. more moist PBL, but also had
a stable
had an
a
layer aloft.
Although SPS had almost as much
convective
instability as EMC,
convection was held
inversion.
As already noted,
there were
inversions
(including SPS).
between
15 and
These
inversions changed
This
three hours.
During
Convection broke
little
in
15 GMT.
strength
15 GMT, and was steady for
this time,
cloud cover
out at those
soundings at
suggests that a mesoscale
into Oklahoma by
strength depending on
and/or
the
in almost all of the
18 GMT.
inversion moved
the next
back by
BLH took place) the
and perhaps soil
places where and
a convective trigger released
moisture.
when the BLH
the convective
instability.
The soundings showed that the convective
the
PBL increased
during the period
increase was modest,
The convective
considerable
and due
in part
12 0MT to
instability of
19 GIT,
to changes at
inhibition changed relatively more,
values at
15 GNMT to nearly
zero by
but the
500 mb.
from
13 GMTI at EMC.
PAGE 237
To
quantifu
the factors
was constructed
FSI,
and
time
evolution.
involved:
to represent the
the PBL model
a -hybrid sounding
inil.ow air
was run with
this
(JUNHYB)
EMC and
between
sounding
to give
the
These results follow.
5. 5 Hybrid Modelling
The hibrid sounding
Notice
that
for
it was quite
amounts
used
initially
values,
and are
12 GMT is shown
humid
were
at low
derived
levels.
shown on Table 5. i.
1 to
was weighted 3 to
sounding.
The position of
be upwind
of the middle of
The
at
changes were
Rainfall
from
to
and EMC.
(shown with
5 June.
an asterisk
Surface values
analyses.
The
June)
was plotted
(5
_,
on
ig
5. 2)
A tongue
parameters.
inch along
region of
on
J-
moisture
The
Table
5.2.
day
ore
This
area.
from the three soundings
of moderate precipitation extended northeast
fell at OKC on
chosen to
in
soil
with amounts reaching over
the OKC
shown
the previous
help set the
was
The
JUNHYB as
_
Fig. 5. 24 to
input to
sounding
from the surface
interpolated
and applied
(FSI,CHK,EMC)
profile.
the convective outbreak
12 GMT were interpolated
imposed
initial
the hybrid
humidity
12 OMT OKC and SEP
primarw
between CHK, FSI
was midway
location
ive
The cloud
from the relative
soundings were averaged to derive the
average
in F-ig. 5.23.
the
was
FTromi
SPS to MLC,
the axis.
hybrid
dry
the
Rain aiso
sounding
prev'ious da.
PAGE 238
On 4 June.. one day
stations
soil
in
earlie-, rain had only
Oklahoma, all
in the southeast cornar.
So
surface moisture could not have been verg high
region.. certainly no
moisture
for the
higher than 60%.
inflow air
had
The soil
the
for this
surface
to reflect a contribution from
since the trajectories
band of moderate rainfall,
the
fallen on four
from 12
@1IT to 19 GMT crossed this band.
The previous month, and
were moderately wet
to 60%.
(soil
So
first 3 days of June
values were used
for the model runs:
moisture) 10-30, 30-30,
soil
surface m6isture-bulk
soil moisture was
in the range of 30%
should have been
the following
30-60, 60-60 with
the
in Oklahoma, so the bulk
Likely va-lues
moderate.
indeed,
the expectation
that the middle two runs
woul.d be most rea-listic.
The modeli-n-g
results for the runs without clouds or
the PBL presented no surprises.
imposed changes above
figures appear
the deepest and
In
fact,
the
over
The
sounding.
the range of soil
lower for the
top
of
large difference
between the
that for 30-60 run.
instability.
10-30 run than for
the PBL varied roughly
almost 2 g/kg.
PL.I value
This difference
50 mb
the PBL moisture values
moistures, and
showed a wide variation as well,
(10-30) had
soil run
also the least
dryest PBL, and
the 19 GMT PLI w as
initial
The dryest
in Table 5.3.
The
There was a
for the 60-60 run and
in convective
instability
PAGE 239
due to
was
the much
higher PBL moisture content.
both
GS and PLI
true
for the surface temperature,
but
direction..
from the 30-60 run.
"Jumped"
the moisture
temperature to give higher
None
coupled
that all
oversaturated
by a
The
Fig. 5. 25.
in
Although
the GS dropped
that the model
oversaturated
lacked
clouds were added,
model behavior.
Table 5.3.
Generally:
and more moist.
That
is,
the PBL
the plain runs.
to
The anount
growth under
The surface
the clouds as well,
clouds.
of
and
The PBL moisture
an
effects.
occurred in the
in some form,
The results are
change
in
cloud was
grew 25%
less
higher,
of
shown on
shallower, cooler
PBL growth
was 33%.
1/3 less than that of
temperatures were verY
stayed
the PBL
unrealistic,
the PBLs were
of
time,
were thinner at the end
for details).
all
shown in
important physical
The clouds were present.
(see Table 5.1
the model
the entire time.
large changes
throughout the run, although they
is
with
quickly
runs were all
expected result, since they
When
throughout
time of the 30-60 run
to stay
This was
of the runs were
large amount, and
fast enough
the cooler
The runs were actually
behavior
implied
the run
oFfset
of .the runs showed any negative area.
at the PBL tops.
This
in the other
convective instability.
to the observation
top rose
This was also
increase easily
oversaturated
runs.
going
Notice how
sensitive
than without the
as expected.
The
PAGE 240
drying
the PBL was reduced
more than 25%.
with
run on Fig. 5. 25 that
the plainT
throughout
the run.
The
the di ferences were built
heating of the surface,
drying as
the moisture was mixed upwards.
magnitude
of the
changes was
instability!
even decreased slightly.
and
moistening effects
the
same growth
in
In all
PLI.
As
all
unrealistic
it
the cooling
yielding almost
none of the runs had any
were oversaturated throughout
included in
since clouds were
predicted by
the PBL growth
the model
yielding PBL tops much higher than
The sensitivity
but only
higher with
the saturation
the
by
level of the
unrealistic.
imposed
the opposite of the sensitivity
in
The results are shown
shows the 30-60 run.
However,
excessive,
of PBL characteristics to the
(Table 5.2) was nearly
to clouds.
the runs.
was still
So these runs were still
PBL parcels.
runs,
proved
In this case, oversaturation was not obviously
the run.
changes
balanced,
before,
19 GMT, and
negative area b-y
however,
of these runs,
were nearly
the clouds.
For the wettest run,
insensitive to the clouds.
relatively
and
However "the
drastically cut under
of convective
The growth
compared
behavior of the PBL. was the same
rising PBL top,
qualitatively:
Fig. 5. 26 shows
Notice, when
time evolution for the 30-60 run.
the
up
in
Table 5.3 and Fig. 5.27
The PBL growth was slower than the plain
about 7%.
changes, by
The
surface temperatures were
almost
10%,
again not a
large
PAGE 241
impact.
6%.
The PBL moisture
The PBL responses
resulting
in
from the run with
imposed
GMT, the
to the
closely,
imposed
changes
C.
After
changes at
750 mb and
growth, which
warmer and wetter.
15 GMT.
The
Entrainment at the PBL top
character of
Between
12 and
the
15
by more
occurred
in
inversion seen
inversion slowed the PBL
became noticeable at
the same sensible heat flux
diverge
the PBL top warmed
This created
This
to
began
15 GMT no further changes
15 GMT soundings.
but by only
imposed changes were minimal
the plain run
temperatures above the PBL.
on the
slower,
explains what happened.
layers between
than 1 deg
little
PBLs which were shallower,
When compared
the
dropped a
15 GMT.
Slower growth
for
gave higher PBL temperatures.
gave warmer values as well.
of these effects contributed to higher TS values.
moisture responded similarly,
having
less
Both
The PBL
room for the
spreading out of similar amounts of moisture, giving
less PBL
dry ing.
The combination of warmer and wetter PBLs
make a significant
difference
was enough
in the PLI values.
The
tendencies for temperature and moisture were both
direction
the
in
of higher
imposed
changes
instability.
aloft
growth of convective
The
resulted
in
instability.
a significant effect onl'y
small PBL
more
The
on the PLI vaues.
to
in
the
changes
under
than a 40% increase
imposed changes
The negative
had
PAGE 242
area was
zero
For
all
unrealistically oversaturated
When both
clouds and
.just as before.
imposed
changes aloft were added
the model,
the results were as
tendencies
for smaller PBLs combined
more than 40%
only
lower than the
reached above 820 mb,
surface
the clouds, which had
themselves..
35%
shown
lower than
This was
induced
or oversaturated,
and
had no
The
The
dryest soil
860 mb.
run
The
in the plain runs,
clearly
such a
due
to the-effect
large change by
tendencies combined
This resulted
to
to give PBL growth rates
the wettest only
The moisture value
less drying.
in Table 5. 3.
plain runs.
and
temperatures stayed
growing 20% more slowly.
of
and the runs were all
the runs again
in PBLs which were
negative area
to give
saturated
just as before.
The PLIs responded to the higher IOS values,
generally growing
more than the-plain runs by 40%
The time evolution
The figure
(Fig. 5.26),
of the
looks very much
instability rose
clouds and
imposed
clouds only run
this
like the run with
except for the behavior
convective
during
30-60 run appears on Fig. 5. 28.
time.
between 16 and
the PLI
The other
the
period
of heaviest
The
19 GilT
when both
as opposed to the
was appr'oximately
level
noteworthy aspect of Fig. 5. 28 was
that the run showed oversaturation
hours.
of the PLI.
changes aloft were added,
in which
clouds
cloud
only
during
the first three
After that,
the P1L
PAGE 243
growth was balanced
end
of the run.
This behavior was very realistic for
convective region.
the
the
first three hours and the
growth
two dryer soil runs, which
the clouds and
in these runs.
oversaturated for
changes were
the
This qualitative behavior was also present
for
Apparently
conditions until the
just.saturated
giving
last
imposed
The
several
were oversaturated
15 minutes of
unable to control
the run
changes balanced the PDL.
wettest soil
hours,
for
only
run,
however,
so that clouds and
was
imposed
it.
5. 6 Summary
The
Values
sensitivity
are
shown
summarized on Table 5.4.
values are
moisture
.for the soil
parameters
The ,bulk soil moisture (GWB) was generally nrot
importance for a-n-y
factor
the
of 2
of the parameters.
(30% to 60%)
surface temperature rise by
6%, and the PLI
The surface
although
growth
soil moisture
GWO
between the two
from 10% to
30% was
by
only
5%,
5%, the PBL moisture drop
by
(GWO) was more important,
its effect was less than the clouds or
The GWO varied
The two numbers for each
difference
growth
by 6%.
changes for some variables.
60%.
great
Changing GWB by a
the PBL
changed
of
as well.
from 10%
to 30%7 to
PBL characteristic reflect the
jumps.
less
imposed
Curiousil,
important
the tripling
than the
doubling
of
from
PAGE 244
30% to 60%.
increases,
The PBL growth was cut 5%
and the rise
to
the additional moisture.
budget.
strongly. to the
from
10%
to 30%.
to
following
changes,
PLI
imposed
changes
(+
added
it
for one of
to the
the runs.
clouds and
imposed
importance:
clouds + bulk
surface moisture +
soil moisture)
exerted too weak an
to show up
is possible
convective instability
conditions + soil
parantheses
this series of runs
for
even for the change
from the 30-60 run with
initial
The factors in
PLI
of
in decreasing order of
=
the PBL
responded
the other case studies,
in a conceptual manner
is
in
grew 25% to almost 75% more for the
summarize the sensitivity
various factors
The
The PLI
12% to 30%
a direct role
surface moisture,
As in
the two
understandable,
The convective instability
soil
increased GWO.
dropped
This was
since the surface soil moisture plays
moisture
by
in temperature was affected similarly.
"The PBL moisture was more sensitive, and
less for
11%
influence
on
in the non-linear comparison.
imposed changes only was the same when clouds were
for the 30-60 run.
moisture were
inferred
The effects of clouds and bulk
from the isolated
runs
("plain",
soil
clouds
only).
In
which
this case it was fairly
agreed with the
simple to
find a model run
suirface observations.
The
only runs
PAGE 245
which
were not
oversaturcted
10-30, 30-30 and
changes aloft.
clear PBL top
for most of
30-60 runs with
Of
these,
for any
only
period during
soundings.
excellent.
Hence,
changes simul.ated
No dynamic
start
The
comparable values taken from
frrcing
=e.emed to be
Strong
imposed
convective
inhibition at
changes
However,
involved as a trigger
the model
negative areas were
the 30-60 model
EMC at
had no
18 GIMT was close to
data, what the PBL structure really
output from the "simulation" run:
calculated hourly
but was
19 OMTI
starting
Notice
at 15 GMT.
that the negative
positive before that.
the amount of negative area was substantial,
even by
18 GMT a :parcel wLould
vertical velocity.i
buoyancy.
The
run with
clear from the coarse
These values are shown on Table 5. 5.
zero at
to
and no surface
("simulation" run)
19 GMT.
of the sounding
area went to
imposed
temperature gradients were not
this condition, but it was not
fact,
on Fig. 5. 29, was
the surface, wind shear was weak,
Using
not predict a
reality quite well.
clouds and
was.
imposed
The 30-60 values
comparison, shown
convergence was disco-vered.
resolution
the run.
the 30-60 run ,with clouds and
the convection.
observed at
both clouds and
the 30--60 run did
for TS and GS were compared with
the nearby
the time were the
so that
have required almost 3 m/is
to penetrate to
sounding for
In
15 GMT
the
level
(Fig. 5. 3
considerable negative area at that t8ie.
At
of positive
) shows
18
the
-MT (Fig. 5. 31)
PAGE 2"46
there
was
still
and
any
small
small
Fig. 5. 32,
parcels.
left,
a
region
werT'e
velocity
vertical
These conclusions are
Fig. 5.33
and
at
appeared
and were not
case and
.
This pattern
the April
case.
that
confirmed
fields
Notice
is
the
area was
of the PBIL with
as a trigger.
by
observations.
of convective
that
the
within
echoesC
same
as
instability
wich
the region
the areas of greatest
in
for PEL
o n egative
perturbations
19 OMT were contained
at
inhibition,
instabilit
18 GMT.
buoyancy
free to rise out
shows the analyzed
inhibition
negativE
OMT, shows
at
PBL par-cels
of
in
of minimum
convective
the May
23 GMT
PAGE 247
Table
5. 1:
CLOUDS
IMPOSED
if'N MODEL RLUNS FOR JUiNE CASE
15 GiT
12 GlMT
Pr e
18 giT
sstr e
800
825
850
875
900
925
o
0 %
X
10
30
50
PBL top
PBL. top
-25mb
0
0
0
POL
top 30
PBL topC 30C
30
70 %
%
%
7.
Xi/
80
Cloud amounts expTressed as a percentage o
complete cloud
approfm ate pressure
PBL tops are shown in their
cover.
depends on part.cular Trun.
level
s--a tual level
Table
5. 2:
IMPOSED CHANGES ON MODEL RULNS FOR JUNE CASE
12 -
Pressure
T(deg
G(g/kg)
0.
0.
0.
500
525
550
575
T(deg
0.
O.
+1.
-i.
625
650
675
700
725
75/
775
800
825
850
+0
-0.
+0.
+1.
+1
+i
+0.
0.
0.
O.
-U.
0
I
0.
+ .
+0
-,.
+0.
-0
+0.
lever
0.
O.
-.
+1 .
9 75
900
O.
+0.
+0
O.
O.
O.
0.
G(g/kg)
+0.
+0.
+ 1
19 GiMT
+2.
+C0.
-0.
-0.
C)
-0.
-0.
0.
+0.
O.
+0.
ou
PT-essure;
C)
15 -
15 GMT
•
nct
PT~uc
Ievl
ore nrtment
ha
v.
ca
.
,c
PAGE 249
Tabie
R un
5. 3:
PH(mb )
TS(
MlODEL RES3ULT-I
SC(g
C)
AT
19 GMT, 6 JUE
PLI(
kg)
C
NA
CONDITION
Plain
10-30
30-30
30-60
60-60
12.
13.
13.
i4
30.
30.
29.
28.
724
733
745
772
S+
S+
S+
S+
Cloud
10-30
30-30
30-60
60-60
Imposed
10-30
30-30
30-60
60-60
800
809
820
844
5+
S+
S+
Er+
Y.-
14
27.
26.
15.
Changes
31.
736
749
761
790
Clouds and
10-30
30-30
30-60
60-60
27.
30.
29.
12.
i3.
13.
14.
S+
13.
14.
14.
15.
S+
S+
S
S+
S+
S+
S+
Imposed Changes
818
829
841
359
28.
283.
27.
26.
Run ID is expressed in form:
Surface Soil Moisture/Bulk Soil
Moisture.
PH is
pressure level of PL__ top.
TS is temperature
at the top oF the surface laer.
GS is PL
moisture.
PLI is
convective instability.
NA is ne gative area (convective
instabilit).,
in mr**2/s*2 (energM/mass).
C ondition refers to
the saturation condition at the top of the PBL:
blankunsaturated,
oversaturated.
"'S
=
nearly
saturated
S
=
saturated,
S+
=
PAGE
5. 4:
Table
Change
in:
SENSITI:VITY YAL.UES
MOCDEL R UNS
OR 23 C4T
PH
6
2L49
UN
PLI
Fac tor
+10
Cl ouds
to
-08
Imposed Changes
+
Clouds and
Imposed
-41
Changes
-- 22
Soil Surface Moisture
10 to 30%
- 5
30 to 60%
-1i i
Bulk Soil Moisture
30 to 60%
- 5
9
- - li 5.
-
6
+44
-35
+40
-12
+25
+73
-- 56
- ,
+ 6
Values are percen-tage changes of normal variation of
parameters, expressed in relation to the maximum change in the
plain runs.
PH is the pressure level of the top of the PBL,
TS is the temperat:ure at the top ofi the surface layer,
S5 is
the PBL moisture value, and PLI is the convective instability.
Table 5. 5: NEGATIVE AREA FOR 30-60 RUN WITH
CLOUDS AND IMPOSED CHANGES
Time
15
GMT
16
17
18
19
GMT
OMT
GiMT
GiMT
Negative Area
(m**2/s**2)
17. 20
12. 96
4. 92
3. 78
0. 00
Negative area is related to needed updraft velocity to reach
tion
v = (2*Neqative
bjuogancu bu rei
level of positive
Area)*1
/2.
PAGE 250
55f,
Fig. 5. 1
Sy.noptic-scale 500 mb analysis for 12 GMT, 6 June.
Solid lines for heights (dm) and dashed lines for
varticity (10**-5 1/sec).
PAG
251
55s
5E7-
56e
Fig. 5. 2
Same as Fig. 5.1 for O0 GMT, 7 June.
PAGE 252
Fig. 5. 3
Synoptic-scale surface analysis for 12 GMT, 6 June.
Solid lines for sea level pressure (mb) with leading
9 or 10 digit(s) dropped, and dashed lines for 1000
to 500 mb thickness (dm).
PAiGE
253
./
Fig. 5. 4
Same as Fig. 5.3 for 00 GMT) 7 June.
PAGE 225
Fig. 5. 5
Sounding network for June 6-7 case.
PACE 255
*
I,
K0
MLC
0
CDS
0
0
Mesoscale 500 mb analysis for 15 GMT, 6 June.
Solid
lines for temperatures (deg C) and dashed lines for
mixing ratio (g/kg).
Wind plotted conventionally
(knots).
Fig. 5.6
0
ob MLC
7o
cqs
OO
-%14
Fig. 5. 7
Same as Fig. 5.6 for
18 GMT, 6 June.
PAGE
256
'
0
MLC
0
00
Fig. 5. 8
Same as Fig. 5.6 for 700 mb.
MLC
0
00
0V
Fig. 5. 9
Same as Fig. 5.8 for 18 GMT, 6 June.
PAGE 2,57
dS y"P
3
0
3DO
DDC
7r-
I'R
UMI
r~-
---S ~
'A
0*
%
EN
SEP
t
Fig. 5. 10
0
K
Solid
Mesoscale surface analysis for 12 GMT, 6 June.
and
(K)
temperature
lines are surface potential
Winds are
dashed lines are dewpoints (deg C).
Sky condition is
plotted conventionally (knots).
clear (open circle), scattered (single bar), broken
(double bar), overcast (filled circle), and obscurred
Cloud type is plotted (if available)
(x in circle).
as is current weather according to conventional
Radar echoes are cross-hatched
synoptic code.
irregularly shaped areas.
PAGE 258
Fig. 5. 11
Same as Fig. 5. 10 for 15 GMT, 6 June.
P AAGE 259
Fig. 5.12
Same as Fig. 5.10
for
18 GMT, 6 June.
P AGE
Fig.5. 13
Same as Fig. 5.10 for
2r0
19 GMT, 6 June.
PAGE 26.
-Th
UMN
MLC
EMC
0
0
ADM
oo
0
)o
/
(.P
r
fo
Fig. 5. 14
Change of potential temperature and dewpoint between
Solid lines are change in
12 GMT and 19 GMT, 6 June.
potential temperature (K), and dashed lines are
change in dewpoint (deg C).
PAGE 262
Pressure
(mrb)
400
500
600
700
800
900
1000
-40
-30
-20
-10
0
Temperature
Fig. 5. 15
20
(deg
30
C)
Sounding plotted on a pseudoadiabatic diagram from
Oklahoma City, Oklahoma for 12 GMT, 6 June.
Solid
line connecting dots for temperature (deg C), solid
line connecting * for dewpoint (deg C), dash-dotted
line for 313 K isentrope, and dotted line showing
moist adiabat for mean PBL parcel (or selected parcel
if PBL is not well-defined).
Dewpoints colder than
-40 C are plotted at -40 C.
PAGE 263
Pressure (mb)
__
_I
___
_
_~
_
_
888
I~ti;1lllllI
l
lit 5 lit---LI
- II----I it
-48
:
-3Z
1
I
If
~Ltil lit
iiit
Ill~lr
llllllllTllllll~l1ll1111111111
LItlllii- LYI--lII -- t II-- LL~UILLLLLLLL~
~ ii~111111111111
28
-20
38
Temperature (deg C)
Fig. 5. 16
Same as Fig. 5. 15 from Hennesey, Oklahoma for 1312 GMT
6 Jun-.
PAGE 264
Pressure (mb)
__ __
400
____
~ __
__*
__
_ _ ___
_____L
___
500
600
700
900
;Iltillltl~llllllttl111
III1I1I1I1II11IIIII1LIIIIIL1LI1IL
-- ~l--~'-~t~-"~~--- ~'----------~~LI~---C- --------
-40
-30
-20
-10
ILuntantIL
1-11
-- ~-------------C--
'---
10
Temperature (deg
Fig. 5. 17
20
C)
Same as Fig.5. 15 from Elmore City. Oklahoma for
GMT, 6 June.
-1
40
PAGE 265
Pressure (mb)
4008
-
-
--
I
I
---~I
688
*ol
98800
900
(Itt~ll(Llllillllll~1111111111
h-111
.l----- .....
11 -----------i , --.ti
.,I -----------/~I.----.J,,
... l.....L
---- -- -- ---- ---- -~~--~--~---- I- ---.i...
- -~------
-40
-38
-20
-10
Temperature
Fig. 5.18
38
(deg
C)
Same as Fig. 5.17 for 18 GMT, 6 June.
PAGE 266
Pressure
(mb)
400
500
600
-48
-33
-2
-10
0
18
Temperature (deg
Fig. 5. 19
20
48
C)
Same as Fig. 5. 15 from Fort Sill, Oklahoma for 15 GMT6 June.
PAGE 267
Pressure
(mb)
4008
-
---- -- - --
- ----
----
608
800
98800
"i
111 --!~-~1~"~'
IIf
- - l~'''~-~-
1ffi!1
II 11. 1
-''"Y
U'Y"
11 1.
111111 111"'"''
II'I
""
""U"~'
--
_
--. - .
.. . .......
. -.----. ' L- -h,,
,-i,,,. .,,
-,,
,!
, , I,,.,, -.,: .--,,".,
,_ ,
,
-48
-30
-20
t
-~
~
1:
:-1
.:7- 1-::
-10
Temperature
Fig. 5.20
-I. I
IILLIIILIII1(111~11:LI11111111;Y I
28
(deg
Same as Fig. 5.19 for 18 GMT,
C)
6 June.
11-.-
-1-1-1.
-1
..-
PAE 268
Pressure (mb)
400
I
--
~--II~--
--
- I~~----- --
-- - - ----
; -
600
800
IIll II L
t tlln
I:.I1
.
-u
~-----,,~l~----*rrrr~
---- n-- ---
-40
-30
-20
ill,
--L,,I
~ -- L .I.LIitt
r
----------
-10
Temperature
Fig. 5. 21
l
tltti
30
(deg
C)
Same as Fig. 5. 15 from Clinton Sherman AFB, Oklahoma
for 18 GMT, 6 June.
-i
40
PAGE 269
Pressure
(mb)
--- ,
--
--- I
I
_ _ ._
ee
9088
IIItII
1000
aminninh
-48
-38
th
iu
hnoubol
S-108
haimbniub
8
18
Temperature (deg
Fig.5.22
28
i
38
4
C)
Same as Fig.5.15 from Wichita Falls, Texas, 17 GMT,
June.
6
PAGE 270
4-
v
+
+
600
+
+
700
+
- 800
'+\
-900
-40
0
+
+
- 1000
20
30
40
TEMP
Fig. 5.23
Same as Fig. 5. 15 for JUNHYB, 12 GMT, 6 June.
..
9
OKLAHOMA
j,,..
,-
.
.5
*...
PACH90TH
"- 4........*
intU ME
r "I
p
-"
,
9'9*,.9
.
RAL
,_,
'!'
4.
.
..
.
T.
0
.44)
.J
,
.
T.
9-
j9.r -A
..
too k,
.
'
,.^ , for t
'drawn
.
......
10,
.
50,,.
1I.-r
ok
It .,
N
oo.
S A
W- 4.
to
2 k,
• ,
- -. o. I~
inches.
'
to
r
...
-
IV.
,
...
1
1
".
..
'9.4)9,
r " ". ,:,
-
I....
.
1..
- ...
/
ADS
's.I
. ......A op ,. _
SADP.L A.,ND -W
'81
.. ...
d
,,
3n.,
L
AV
I.A.,TkNE
.
9
Lr.'
_
...
.0,444,
I
.ITH
".FAA
)CEN
rf
........
ANTWEZONE
M
Io
;O
. '
Th FIT.. .
It.40
OO~j
:
PA.E
272
TS NR
30 100029 900 -
QS
-17.0
28 800-
-16.0
27 700 26 600
-15.0
25 500
24 400 23 300
- 14.0
22 200 -
-13.0
21
100
20
C 5+ k
PLI
s+
-t
- 12.0
I .
- 6.0 12 13 14 15 16 17 18
TIME
5.0 -
-,-
PH
9
750
P
800
4.0
- 850
3.0
- 900
- 950
Fig. 5. 25
Time variation of model output for JUNHYB sounding, 6
June, 30-60 soil parameters, with no clouds or
imposed changes aloft (Plain).
NR is net radiation
into the surface (mcal/sq cm rmni), TS is temperature
at top of surface layer (deg C), GS is PBL mixing
ratio (g/kg), PLI is convective instability (deg C)
and PH is pressure at top of PBL (mb).
Condition at
top of PBL is plotted above time axis:
blank =
unsaturated, "S = nearly saturated, S = saturated,
and S+ = oversaturated.
PAgE 273
TS NR
30 1000
29 900
QS
17.0
F
28 800-
16.0
27 70026
600
- 15.0
25 500
24 400 -
-14.0
23 300
22 200
21
13.0
100
20
-12.0
PLI
I I
(
I I
19
18
17
16
15
14
13
12
6.0 TIME-
PH
750
4.0
850
3.0
- 900
" 950
Fig. 5.26
Same as Fig. 5.25 for run with clouds.
PAGE 274
TS NR
30 100029 90028 800-
QS
-17.0
ls
-16.0
27 70026 600 -
15.0
25 500.
24 400 -
S
-14.0
23 300
22 200 -
-13.0
21 100
20
-12.0
PLI
I
I
I
6.0 , 12 13 14 15 16 17 18 197
TIME
5.0 -
PH
800
4.0-
- 850
3.0
- 900
950
Fig. 5. 27
Same as Fig. 5. 25 for run with imposed changes aloft.
PaGE 275
TS
NR
QS
30 1000 -
17.0
29 900
28 800.-
as
16.0
27 700-26 600 -
15.0
25 500
24 40023 300
-14.0
22 200 -
-/13.0
21
100 -
20
5
t
PLI
6.0
I
i
s
I.
5
i
I
12 13 14 15 16 17 18 19
TIME
- 12.0
I'
PH
750,
5.0
800
4.0 -
850
3.0
- 900
950
Fig. 5. 28
Same as Fig. 5.25 for run with clouds and imposed
changes aloft.
PAGE 275
QS
30
17.0
29
28
16.0
27
r/i-
26
15.0
25
24
14.0
23
13.0
22
H
21
20
12.0
I
!
I
*
i
S12 13 14 15 16 17 18 19
TIME
Fig. 5. 29
Comparison of TS and OS values from 30-60 model
run
with clouds and imposed changes aloft with values
taken from Elmore City. Fort Sill, and Chickasha,
Oklahoma soundings.
PA E. :77
SP
+
-40
-30
Fig. 5. 30
+
+
+
+
+
700
+
soo
00
-600
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
> 1000
-20
-10
0
TEMP
10
20
30
40
- 900
Same as Fig. 5. 15 for model output at 15 GMT from
JUNHYB initial sounding, 30-60 soil parameters, with
clouds and imposed changes aloft. Negative area is
cross-hatched.
PAGE 278
P
600
700
-800
900
1000
-40
-30
-20
-10
0
10
20
30
40
TEMP
Fig. 5. 31
Same as Fig. 5. 30 for
18 GMT.
P
600
700
800
1000
-40
-30
-20
Fig.5.32
-10
0
TEMP
10
20
30
Same as Fig. 5.30 for 19 GMT.
40
PAGE 279
GAG
G
...
C
cs
T
-Tl
Fig 5. 33
Mesoscale analysis of convective instabilitu (PLI) and
Solid
convective inhibition (NA) for 18 OMT, b June.
lines are PLI (deg C) and dashed lines are NA
Cross hatching is new radar echoes
(m.n*2/s**2).
appearing by 19 GMT.
PAGE 280
COINCLUSIONS
The three
cases from SESAME, 1979 were analyzed
in
detail
to determine where and when afternoon convection broke out.
The
surface and PBL characteristics were delineated as clearly
as the
The model
data allowed.
the results were tabulated and
case, and
summarized
used
Concordia, Kansas
in all
As
temperature
rose
enough
output and
and
bu entraining
Hence
and
The
case, at
The
(GLD).
observations was
the moisture
it
three
dryer.. warmer
dropped
the needed
dropped..
air
temperature
unless
m,isturo.
cases was
grew in depth,
the PBL moisture
the PBL average
to supply
(1977).
cases.
the PBL heated,
rose,
data from
from the April
behavior of the PBL in all
follows.
above.
data
It was
layer.
boundary
(CNK) and Goodland, Kansas
between model
satisfactory
depth
They are
predicts the
results from Barnard
compared with
model was also
in
in each
discussed.
and results were compared with
tested
O'Neill, Nebraska and model
as
involved
one-dimensional, and
is
(assumed) well-mixed
evolution of an
The
in chapter 2 was used
here.
The model
agreement
factors
the various physical
to quantify
carefully
outlined
the
from
'ose
soil
generallu
its
1he PBL grew
the
inversion
as the PBL
suT'+ace
was wet
PAGE 21
Clouds cut the incident radiation, slowing the heating
and PBL growth, and thereby contributing
-
values.
to higher moisture
Changes aloft tended to be influential as well.
Strengthening of the
inversion slowed PBL growth, allowing
PBL to heat more from below.
less moisture
the
Dry advection a loft contributed
from entrainment and gave lower PBL moisture
contents.
Soil moisture, especially
surface
influenced the PBL by controlling
the
soi
moisture
surface energy balance.
High moisture values gave we t t er, shallower, and cooler PBLs.
Dry soil PBLs were high)
dry and hot.
In each case, the initial conditions were convectively
unstable.
The 12 GMT data generally showed PLI values which
indicated significant instability.
The subsequent BLH
contributed to an increase in convective instability, but the
increase was usually of
less magnitude
These results suggest that the
initial
was the result of large scale forces
than the initial value.
stratification, which
tas the most important
contributor to convective instability.
Just as the various factors contributed to PBL changes
they exerted
influence on the growh
instability.
The BLH by: itse.
Individual y
the othecr
factors
tended
(cept
convective
to
increase the PLI.
soil
moisture) almost
PAGE 282
always reduced
the wettest soil.
convective
The one
of any
of the runs
exception to this was the June
throughout the
Convective
soil run
least inhibitio-n
involved.
for a given case.
The
generallyt
least negative area.
in an
The dryest
It was
"clouds only"
PBL height* surface temperature
others the
model
changes
parameters,
An
runs were of this type.
same,
combination of
gives
in response to
inhibition.
keeping
However, the
soil moisture affected all
lower PBL potential
temperatures.
depended on details
moisture involved;
inhibition was very
etc.
three
in a non-linear way.
gave higher PBL moisture, but
of the changes
values of soil
convective
less
temperature)
one parameter,
in different directions, and
height and
magnitudes
(PBL potential
increase in any
increased soil moisture
PBL
the
wh-en other factors besides BLH were
The negative area depends on the three way
the
sometimes
that the most unstable runs also had
Some of the
and PBL moisture.
involved a
"plain" runs
soil moisture variation.
often had the
case, however,
the most
the June case.
values in
inhibition responded,
opposite manner to the
had
case which
integration.
lowest PLI
the
actually had
Increased soil
"plain" run usually
instability
cloud cover
the
ins.tability.
"plain" runs gave increased PLI values, so
in the
moisture
that
the convective
lower
the
of the sounding,
So the behavior of
comraplicated 1with respect
to
4
--
PAGE: 283
soil moisture.
Th e Apr il and May cases were similar in the sense that
c Iuds were only present during the early morning.
Thi
determined the effect clouds had on the growth: of c:onvective
instability and
inhibition.
reduce th.e growth of
instability by almost I0.
were added to the runs which
however, .:he
Clouds by: themselves tended to
included imposed changes aloft)
effect of the clouds was
instability.
When clouds
to increase the
This .was true for both April. and May cases.
reason for this behavior is as follows.
The
When clouds are added
their effect .is to reduce the 'net radiation,. and
to the. mod:el,
so reduce the gro wth
of the PB3L.
When added tothe "plain".
runs, the reduced growth gave increased PBL mi.xing ratios but
cooler PBLs,
and. the net result was slightly, lower convective
instabilities.
The imposed changes usually
lead to higher-
o the "plain" runs, and. so much.dryer
growth rates wh-e-n added to
PBLs due to the entrainment of dry air' alof.t
lower PLI values.
imposed changes0
When clouds were added
This produced
to the runs with the
the result- was less growth of the PBL-(than
the imposed. changes only runs),. and so higher mixing ratios,
but only slightlg cooler temperatures in the PBL.
The net
result of these changes: was an increase in PLI due to the
higher mi x ing ratios. ailthough the
high as in
the "plain" runs.
.finalvi.ues. w1er e not as.
Hence the clouds exerted a
positive influence on co nverctive instabiiit
wn~en added to the
PAGE 284
runs with
imposed
I
changes.
This non-linear result for both
cases can be stated symbolically
in
the April
and May 23. MT
decreasing order of.
importance:
PLI
clouds -
= Initial
conditions + BLH + soil
impos-ed changes + bulk
surface moisture +
soil moisture.
I
The 21. GMT May case symbolic relation is ientical
order of
As. can
imposed
changes and
be seen, surface soil
the growth
have only
and bulk
bulk
soil moisture is switched.
early morning cloudiness.
moisture were much
conditions which
The imposed
changes aloft
less important.
The convective inhibition was more variable in its
response.
The relation can
I
moisture was very influential on
of conv.ective instability under
soil
except the..
be stated for the April
I
and May
cases as:
NA =
Initial
conditions - BLH + ...
The order and signs
and
of the
other
-
Imposed
factors were
changes +
case dependent,
so cannot be generalized without more case studies.
I
__
_
.PAGE 285
The June case was at least partially cloudy throughout
the
days
and so
its behavior was different
cases.
The symbolic relations will
Suffice
it to say,
however,
not be repeated
inhibition was decreased by all
changes- led
and the convective
of the applied factors.
layer
to lower boundary
tops
effect was to increase the PBL mixing ratios
of dry air)
special
in June
changes which
PBL,
or the
present.
The nature of the
too.
All
of the
initia-l
conditions had
The April and May
later
imposed
an
changes was
imposed
inversion already
cases also had
changes aloft later
strength
of the
The June. case did not have this
aloft).
effect, so the
(less entrainment
an inversion above the
in the runs which tended to reduce the
inversion .(coolin
so the
cases either had
led to the creation. of
The.
temperatures, thereby
and raise the surface layer
increasing the PLI.
here.
that the convective instability
was reinforced by all of the factors,
imposed
from the other two
imposed
changes had
an effect opposite in
sign on the PBL evolut-ion.
Many of the runs without clouds grew fast enough to
create oversaturation at the PBL top.
This suggests that in
the real atmosphere, the cl ou d s act as a feedback
mechanism on PBL. development.
The PBL can groa
enough to produce saturation at the PBL :top.
speeds up.
the
growth
control
only
If
fast
the growth
the cloud cover. increases, and slows growth-.
is
t
o slow and
the
P.L top
is
not
saturated
If
then
PAGE 286
the clouds may
until
(depending upon larger scale conditions) thin,
speeds up
the PBL growth
again.
enough to produce saturation
i
This balance between clouds and PBL growth was
required
to allow the model to simulate the
case study
outbreaks, even those with only early morning clouds.
The model
was able to successfully
especially.for
region PBL behavior,
These results
convective
indicated
that
in
to very
seemed to start
convective
inhibitin.
run
that the
showed
about 16 m*a*2/s**2
morning
clouds
inhibition in
the outbreak
I
region was only
(50-50 run with modified GLD changes and
The analysis of observed values
Table 3. 6).
inhibition
less
than 50, with
convergence
began
in a region of
the nearest observed value being
The mesoscale surface analysis
outbreak region.
The
In tre April case, the "simulation"
showed that the convection
there was
low levels.
the
in the areas' of lowest
(Fig. 3.42)
26.
forcing,
in regions of maximum
convection did not necessarily begin
but
4
the April and June cases.
fact, for weak
inhibition was reduced
instability
simulate the outbreak
(Fig.3. 13) showed that
in the surface wind field
The model
results and
in the
the analysis showed
that the negative area was reduced by thermodynamic processes
to a minimum value which allowed the available forcing
initiate
to
the convection.
In the May
case,
the 21 GMT outbreak was strongly
C
linked
. ".-
I
I
I e
.
PAGE 287
to a wind shift line (see Fig. 4. 16).
Fig. 4. 37 showed that the
convection began just west of a minimum in inhibition.:
"simulation'" run for the outbreak
(5-70 run with
The
clo.uds and
imposed changes, Table 4. I) had an inhibition of about. 13
mr*2/s**2., a low value compared with the .observations. Again,
the thermodynamic processes lowered the inhibition enough to
to initiate the convection.
allow the available forcing
The 23 GMT outbreak
obvious wind
shift
the Mai
in
line or convergence at
inhibition of b-e-tween
30-70 GCH runs modified
Table 4.6.
In the June case,
on Fig.5.31
The
inhibition of zero
enclosed by
"simulation" run showed
(30-60 run with
In
cIoud's and
this
b
a region
zero.
the pattern was repeated.
to be
in
case
The analysis
the
10
convectiv.
imposed changes,
the forcing
was
complete
by thermodynamic processes had taken place.
pattern also
as analyzed
(30-30 and
and corvection did not begin until nearly
destabilitation
This
in th.e area
for this outbreak
1 m+*2/s**2
in this case nearly
Table 5.3 and also Table 5. 5.
very weak.
(no
and the
surface)
the convection broke out
showed the outbreak
m**2/s**2 line.
and
forcing
for surface temperature advection,
Once again:
of minimum inhibition,
the
"simulation" run
The
of minimum inhibition.
less
the o tbreak occurred
analysis on Fig. 4. 40 showe
showed
case had
occurred
San ders
(19
in the
10-11 April
Hin work
e).
1979 SESAME case
showed that
the
PAGE .268
initial convection began in a region of minimum convective
inhibition and modest
instability.
A
The major finding of this thesis is
simply stated.
Convection requires convective instability
convective inhibition.
and low values of
When the inhibition can be overcome by
the available forcing or trigger mechanisms (21 GMT May case,
for example) or the thermodynamic evolution of the PBL reduces
4
the convective inhibition to near zero, the convection will
start.
The forecasting of afternoon convection is then, the
forecasting
of the removal
might prove to -be a
useful
of convective
inhibition.
forecast aid,
but the model used in
this research required much
runs.
This
labor to produce data simulation
Models -which use available data such as rainfall
amounts,
climat ological
soil
satellite data to predict
operational
use -o
data:
past weather and possibly
soil moisture would allow
this model.
The rest of the
input
parameters are readily available in real time for forecasting
purposes.
It would also
be possible to use the model
local mode, allowing local
moisture data they
would be
forecast offices to
deem appropriate.
make a forecast based
input the soil
A third mode
to run the mdel with a. range of
in. a
of usage
soil moistures, an-d
on the range of output.
• .• .• ..
PAGE 289
PBL behavior is clearLy a complicated phenomenon,
central
and is
to the understand:'ing of initiation of convection
especially when dynamic
forcing is weak or ill-defir'ed.
The.
interaction b etwen PBL variables and the effect on convective
instabilitu and inhibition has been quantified for these three
cases., but more work is needed before a complete understanding
is obtaine.d.
Intees
in ths problem i
gr owing as evidenced
by recent work by McCumber and Pielke (1981)
(1982), and Garrett (1982).
. Cooper et
..
As the importance and nature of
the interractions of these variables becomes
better defined.
accurate short term forecasting of convective outbreaks will
become possible.
I
I
III
I
I
I
-
:r -. L
.
PAGE 290
.' .
'
Derivation of Radiative Parameterization
APPENDIX 7. 1:
The following is taken from Katayama
(1972).
IR RADIATION
I
The downward flux at level z is
T
dB (T)•
fRd
I.
dT
d
tf{v
+
z
4
IwrB (Tz)dv
.
irB (T )T {i (u -u ) dv
-
f
V
0
z
V
where the second .and third
of
(u-uz)}dTdv
0
integrals correct
for the
endpoints
the first and
B
(T)
f=lux
V-
u(z Y
of black body radiation of frequenc y
temperature T
4
-amount of-absorbing medium in the
vertical
air column from the ground
I
to height z
IV4
= the generalized absorption coefficient
u.
fri
= the transmission
function of a slab at
frequenc y V
Subscripts
g
z
represent
the ground
the atmosphere respectively.
dependence
amount
on pressures, K
To
tayama
height. z and
correct
the top of
for absorption
defines an effective absorber
Si
PACE 291
U
PoO
1 p(Z)
where
2)
dp
Vqj
density of absorber
a
empirica
constant
q- = mixing ratio of absorber
i
absorb er identifier
He defines two types of weighted mean transmission functions
as follo
.1s:
d
*
T(u ,T)
1
dB(T)
dT(u Jo
T)
wh ere
and t.
wB(T) =
iB
r
dB( (T)3
T
X u) dv
( ,u)dv
T
(T)dv
Dv
is the Ste an-B ltz..
ann constant.
It i
as.:asued that the total transmission function is a
product of the indiv idual transmission functions.
and 1( u,
T)
r(ut T)
The
used hereafte-r will be assumed to be the total
transmission functions.
U sing these Mean functions simp li fies
the equations for the fluxes greaty.AI
becomes. ..
S.
z
d
c
I..
.
+
T (u
iB
z
where i B.
T.
u
,
d)a()
.
(A vs)
PAGE 292
Similarly.
IRu =
the
g
f oJ
equation for upward flux. at a height
T dB (T)
dT
+
f{Z v(Uz-u) }dTdv
z
z
fI B (T )dv
(61-4)
b ec omes
r'lB
IR
=
u
wB
z
+
(U* - u
z
, T) d(wB)
z
These are the equations in the text, a-4 and
5.
Incident Radiation
Sa
, the solar constant is taken to
Ly/min.
The hour angle H
compared with
the
is computed
be equivalent to 1.94-
from the time of day as
length o-f day. at the particular
location.
albedo is modified for the sun angle by a method
The ground
from Wetzel
(197-).
(eo- H
where H
is
O(
,
So
=So* (ZT)(SMOD)
(oO-oQb
in degrees
limits specified for
each site.
where
ZT
=
zenith
angle, a function of H,
SMOD = correction
S
is
divided
day of year
for the distance to the sun
into scattered and
absorbable parts, 65%
4
PAGE 293
scattered, 35% absorbable.
the scattering albedo
The scattered
0- 2'7 lo,
,
foo d
S-
.
the
The scattered
scattering albedo depends on them.
Clowd4op a orabove (oo
06
0D
t
radiation incident on the ground is modified
Absorptian
S
46
+p beour (oDo
00
1o
multiple reflection between the ground and
*
A)
T
surface pressure (t6h)
If clouds are present,
for
modified .for
of the atmosphere,
035
where PG
is
part
the atmosphere:
by water vapor is calculated using
effective absorption of H20 as aiready
the values
computed
for the IR
Fractional :absorption for the whole atmosphere
Anu
. clouds encoun ter ed absrb accorud ing to their
.
.. . .. .
the level i is:
flux.
• utpe.'0eto
where
i.
ewe
hegon
n
for
h
down to
topee
is taken f&om the top of the atmosphere down to level
Absorption of a given layer
is then a function of
available radiation ad EH20 in the
layer.
PAGE 294
thickness
absorbable
ground
yield
and
height.
radiation.
(GLW,)
is
cloud tops also reflect some of the
An
absorbable radiation
combined
the incident
modified for
The
with
solar radiation on the
the ground
soil.
is
This is
then
RB
AI-IS)
MR'
~;I
radiation to
albedo to give the ground absorption
The net radiation at the ground
I.
the scattered
left at the
G8
tT~llreruI
I
I
I
1-
PAGE 295
APPENDIX 7.2: Derivation of Ekman Laer Similarity. Equatiions
The Ekman layer similarity theory is used to find SH and
LH as follows.
The equation for PBL wid. is
k V
SCA
.
(from Ara. 1975...
.
where
v
Von Karman constant
u.
friction velocity
zo.
roughness length.
OT
= structure-funt
tion
For :unstable condition-s..
(lrf2)
Sao
letting
VS
ind
The sensible heas
su
!
at .speed
th e. ta-p of the
PB-L*I
91ux into the PBL: is :given. by
•.
..
G
:
..
•
.(A
.f)
.
.
.
.
.
where
density of air
c
0
.specific heat of.air
.
(ground)
-
(PBL)
0
r:::~
I
.
PAGE 296
(Az2 s)
-L
where
.
,
( 2.()
which gives
,
Jk
'(pt~
L
C,3-~
Monin/Obukhov similarity theory defines the M/O length L as:
v
where g
-
(426)
.3
acceleration of gravity
TS = surface temperature (airnot ground)
Solving for SH gives
l
3 TS . C
llll1 Z,) l
p-
Equating.k*? and W2,, gives
anCl
L5 Vcj
. -
%Ig
Inserting for u
/41 loS
,
I
Ts
from U-*3 and rearranging
L-O
~idi
LL ~
L
e
~
_0
CA Z.I
I
I
-
----------
- ----- ---------- -----II--
~ -~~-~--~-I-~ ~~~~ I-~- -----
PAGE 297
For a given
8tVS this equation is in one variable, L.
solved by iteration to yield the correct value of L.
value is then inserted in
This
V.3 to find u , and L u* are inserted
in ~1*. to find SH.
Finally,
where
the companion equation for LH is solved
X = latent heat of evaporation
9q = q(PBL) -
q(purface)
q at the surface is found according to Wetzel
(1978)
where OSAT = saturation value of q for ground conditions
(TQ,PG).
6w
It is
o
4%m4+
S 4u YwKL50;
TUY
U'W
&C
PAGE 293
APPENDIX 7.3: Derivation of Soil Heat Flux Parameterization-
i•
The following is from Bhumralker(1975).
Heat conduction is described by
vr
(zt
.I
T5
C.
where T (zt) = soil temperature at depth z, time t::-
I1
:
it = thermal conductivity of soil
Ivolumetric heat capacity
c
Assume that the surface soil temperature T (Ot) is described
L
- -
T
TBAR + AT, sin(4et)
T (O t)
3.2.
a.verage- temperature of the soil. assumed to be
where TBAR
invariant with -depth
I.
vamplitude
of the variance
ATo-
2T/period
) = frequency of the variance
The solution -of
43
T
where d =
negligible.
(z.t)
C2c/c]
is
A3.3)
:
= TBAR + bTexp(-z/d)Esin(bt-z/d)3
= depth at which the amplitude of A ,is
thin soil
For an infintesimally
layerp
the heat.
flux into the soil is
"
Combining 43-3 and A%3.
k .i)
(A
"p
I
gives
I~i
'i
.A
CDS
.T
c4))
e
..
, . .. .
.. . . .
.
...
:
I1
_____1
_
~__ __I ___I_
~
~
_1~~_1__
_
___~
PAGE 300
APPENDIX 7.4 Derivation of Ground Variable Equations
The following is from Bhumralker (1975) and Deardorff
(1977).
Consider a layer of soil from the surface (z=O) to
some depth ( =z).
The time rate of temperature change for
this layer is given by
where c is volumetric heat capacity and all other variables
are as previously defined.
G(O.t)-= soil heat flux at the surface
(4 Z)
NR - SH - LH
G(z,t)
s-oil heat flux at depth z from Appendix 7.3
If the approximation is made that
T (z=I
then A4I
becomes, with A4-7
and AI.3
=
where -HA = NR -
LAq')
cm,t) = T (0.t) = TGO
-,-
SH - LH
Rearranging, gives the equation 2-10 in the text,
(4 4)
W
PAGE 301
-Y.
44-
=
(&=
-c
(f&r
(4
(A4.,)
to
From Appendix 7.3, d is defined as
: t2 Kz
--
- /
o
i'"
7(
A -7)
The denominator can then be written
(Aq-8)
Let lcd = c(-2cm)
for d
_5
2C
cd , where 1
+4
cm, 1 = 1.04
4L.(, is then
A_
a76-- LatS
( 4)(F1
Ll)
Using 44.7 for the second term, this becomes
3T(.-
.
'aa.
Since
&=
.
Y
T
•
24/' , we have
wA
(Y,)
pe
eAod
t'
2R
-
C,,
_ C,,. ( -T-,-'ra)
cI
C
di S0 t )+
PAGE 302
Deardorff (1977) suggests that a similar equation be written
for soil moisture
where
density of water
d
= depth of diurnal moisture cycle
E =-evaporation
P = precipation-
WG = volumetric soil content
WBAR-=-bulk moisture in soil
S
p=
eriod- of diurnal cycle
c ,c
are constants
Dividing through by--field capacity moisture, WMAX and taking
E = LH/
P=O
where
GW
W=G/W
X
GWB = WBAR/WMAX
= latent heat of evaporation
This is the equation used in this model, using (from
Deardorff. 1977)
d,
= 10 cm
= 1 gm/cm
II
I
1
T
I
1
I
PAGE 303
= 1 day
I
.
,W 5. I~/
c.
=0.9
Datea for these constants comes from Jackson (1973).
I
I
I
.
III
I
I
I
--
PAGE 304
APPENDIX 7. 5: Derivation of Inversion-:Equations
The following is taken from Zeman and Tennekes (1977).
A
complete discussion of the equations would be lengthy and
unnecessary.
It
mentioned above.
is found in its entirety inthe source
Briefly, the equations result from.
consideration of the turbulent energy budget at the inversion.
This balances kinetic energy change with buoyant production#
turbulent flux divergence, mechanical production#
dissipation.
For a convective boundary layer
and
such as will be
considered for this model. mechanical production can safely be
ignored.
Buoyant pro-tctian is proportional: to temperature
and heat flux &t
already.
the inversion.
The temperature is modeled
The sensible heat flux at the inversion is equal to
the temperature jump AG8times the rise of the inversion
h/bt.
as proposed by Ball (1960) and Lilly (1968).
where H
according
( = sensible heat flux at the inversion) is changed
to Tennekes (1973)
where ea /It
is a function of
etentrainment of stable air from above, and net sensible heat
transfer inside the boundary layer
The turbulent transfer is parameterized according to Tennekes
I
PAGE 305
(1973).
He maintained that the large eddies which transfer
most of the kinetic energy scale on h and aw, the inversion
height and the convective velocity scale.
i
44
'V*4
Zilitinkevich
~T
3
TST
-54k
g
is
LJk.~
Shk
-e(3)
~ ~Mewi~(cL ~aLC~bSI
(1975) showed that the time rate of change of
TKE should be parameterized as
- C
4
-
wkre.
%
s
q~
ei.pl
Cusr~bw+
Part of the dissipation at th-e inversion scales on h, and can
be included with tie turbulent flux term.
The rest, according
to Zeman and Tennekes (1977) can be written as
Uh*, Ubvuwhere wb= Brunt-Vais'al
( CS -S)
frequency in the air above the
inversion.
The parameterization is based on a mixing length which
depends on the stability above the inversion.
Putting #S-3,
5 -4 and A-5 together,
h
TS
k
4
:
4-C
II
~
k.
e
C~5.L)
II
PAGE 306
Substituting for Oh/Wt from 4 S'lyields
3
Substituting
C -
tC
,, (
'
IVIt
Substituting for w from k5-3 gives
S-4
+
6WA4
'TS
- TC --A,
which is
equation 245in the text.
5,a
(As
PAGE 307
RE ER ENCES
1975: Geostrophic d rag and heat transfer
Aria, S.P. S.,
relations for the atmospheric boundary layer. Guart. J.
147-161.
Rog. Meteorol. Soc., v. 101. pp.
Ball,
F.K. , 1960: Control of inversion height by surface
heating. Qua-rt. J. Roy. Meteorol. Soc., v. 86, pp
483-494.
1977: On the sensitivity of an atmosphere in
Barnard, J. C.,
radiative-convective equilibrium to soil moisture. M. S.
Thesis, M. I. T.
July, 1977.
1975: Numerical experiments on the
Bhumralker, C.M.,
computation of ground surface temperature in an
atmospheric general circulation model. J. Appl.
Meteorol., v. 14, pp. 1246-i258.
1950: A tabular method for the computation of
Brooks, D.L.,
temperature change by infra-red radiation in the free
atmosphere. J. Meteor., v. 7, pp. 313-321.
izumi. and E.F. Bradley,
Businger, J.A., J.C. Wyngaard, Y.
in the atmospheric
1971: Flux-profile relationships
v. 28, pp. 181-189.
surface layer. J. Atmos. Sci.
Caracena, F. , R.A. Maddox, L. R. Hoxit, and C.F. Chappell,
1979: Mesoanalysis fo the Big Thompson Storm. Mon. Wea.
Rev., v. 107, pp. 1-17.
1982: A severe frontal rainband. Part I:
Carbone, R. E.,
Stormwide hydrodynamic structure. J. Atmos. Sci., v.
pp. 258-279.
39,
Carlson, T.N., R. A. Anthes, M. Schwartz, S.G. Benjamin, and
D.G. Baldwlin, 1980: Analysis and prediction of severe
storms environment. Bull. Amer. Meteor. Soc., v. 61, pp.
1018 to 1032.
1982: The diurnal
Cooper, H. J., M. Garstanq, and J. Simpson
interaction between convection and peninsula-scale
forcing over south Florida. Mon. Wea. Rev., v. 110, pp.
486-503.
1982: An intense,
R. L. George, and K R. Knupp1
Cotton, W. R.,
terrain. Part
mountainous
over
quasi-steady thunderstorm
mesoscale
I: Evolution of the storm initiating
. 39
pp. 32.8-342.
c.
Atmoscirculation. J.
Darkow, G. L.,
patterns
P.M
1. Kuhn, and V. G. Suomi, 1958:
as a tornado 4-orecast aid. Bull.
Surface thermial
Amer. Mletec'rol.
-----P----.l.-
.--.- '-
-
-~5--~-
---
PAGE 308
Soc.,
v.
Davis, J.. .
39, pp.
and
J.R.
532-537.
1981:.
Scoggins
The development of
convective instabi1ity, wind shear, and vertical motion
in relation to convective activity and synoptic. systems
in AVE IV. NASA Contractor Report 3386, contract.
NAS8-31 773.
Deardorfft
J.W., 1977: A parameterization of ground surface
moisture con-t-ent for use in atmospheric prediction
models. J. Appl. Meteor.
v. 16, pp. 1182-1185.
)
Elsasser
:W.,
1942: Heat transfer
by infra-red radiation in
the atmosphere. Harvard Meteor. Stud., No. 6, 105 pp.
Fawbush, E.J., R. C. Miller, and L.G. Starrett,: 1951: An
empirical method of forecasting tornado development.:
Bull.
Amer. Meteorol. Soc. , v. 32, pp.
1-9.
Galway, 1956: The lifted index as a.predictor of latent
instability. Bull. Amer. Meteorol..""Soc.
v. 34) pp.
528-529.
*t
1982: A parameter study of interactions between
Garrett, A. J.,
convective- clou-ds, the convective boundary layer, and a
forested surfac-e. Mon. Wea. Rev., v.
110: pp. 1041-1059.
1980: Tornado formation from
Holle, R.L., a-nd M.W. Maier,
downdraft interaction in the FACE mesonetwork. Mon. Weae.
Rev..,
v. 108i pp.. 1010-1028.
Jackson, R. D.
1973: Diurnal changes in soil water content
during drying. Field Soil Water Regime, Soil Sci. Soc.
Amer. , pp. 37-55.
M.J. and R.EI Turner, 1980: A preliminary look at
AVE-SESAME IV conducted on 9-10 May 1979. NASA Tech. Mem.
(TM-78314). Marshall Space Flight Center, Ala., 51 pp.
July
1972: A simplified scheme for computing
Katayama,. A.,
radiative transfer in the troposphere. Tech. Rept. No 6,
Dept of Met. UCLA. 77 pp.
Koch,
S. E. and J. McCarthy, 1982: The evolution of an Oklahoma
Part II: Boundary-layer forcing of
dryline.
mesoconvective systems. J. Atmos. Sci
v. 39, pp.
237-257.
Lettau,
H.
and
Davidson,
First M.iile.
Lili
;
D.
.
strong
1957:
Exploring
the Atmospher's
Pergamon Press, New Yor'i.
1968:
N Y.
volumes
I,II.
Models of clouvd-topped mixed-layers under a
Guart.
inversion.
.
.
Roy.
J.
.
,
Meteorol.
,
' . .,.
Soc.
V.
94,
. ,
"
.
.
PAGE 309
pp.
292-309.
London, J.,
change
1952: the distribution of radiational
temperature
in the Northern Hemisphere. J. Meteor.
v. 9, pp.
145-151.
McCumber, M.C., and R.A. Pielke, 1981: Simulation of the
effects of surface fluxes of heat and moisture in a
mesoscale numerical model.
1. Soil layer. J. Geophys.
Res., v. 86, pp 9929-9939.
Maddox, R.A., L.R. Hoxit and C.F. Chappell, 1980: A study of
tornadic thunderstorms interacting with thermal
boundaries. Mon. Wea.. Rev.
v.
108, pp. 322-336.
Matthews, D.A., 1981: Observations of a cloud arc triggered
thunderstorm outflow. Mon. Wea. Rev.
v. 109, pp.
2140-2157.
by
Miller, D.A., and F. Sanders) 1980: Mesoscale conditions of
the severe convection of 3 April 1974 in the east-central
United States. J. Atmos. Sci., v. 37, pp. 1041-1055.
Miller, R.C.
1972: Notes on analysis. and severe storm
forecasting procedures of the A-ir Force Global Weather
CentraL. Air Weather Service. Tech. Rept. 200(rev. )
102
13P.
Modahl, A.,
1979: Low-level wind and moisture variations.
preceding and following hailstorms in northeast Colorado.
Mon. Wea. -Rev. v. 107, pp. 442-450.
Moore, (unpubli-shed): The development of a severe weather
index for use in a -nested grid mesoscale primitive
equation model. Report of research at NASA Langley under
NASA-ASEE Summer Faculty Fellowship Programs 1979.
Ogura
Y., H. -M. Juang, K.-S. Zhang, and S. -T. Soong
1982:
Possible triggering mechanisms for severe storms in.
SESAME-AVE IV (9-10 May 1979). Bull . Amer. Meteorol.
Soc., v. 63, pp. 503-515.
Raymond, D.,
and M. Wilkeninci, 1980: Mountain-induced
convection under fair
weather conditions. J. Atmos. Sci.
v 37, pp. 2693-2706.
and C.D. Walshawp 1966: The computation of
Rogers, C.D.,
infra-red cooling rate in planetary atmospheres. Quart.
J. Roy. Meteorol. Soc.
v. 92, pp. 67-92.
Sanders, F.1982.: The origin and maintenance of the severe
thunderstorms of 10 Apil 1979. Twelfth Conference on
January
Severe Local Storms of the Amer. Meteorol. Soac.,
4
PAGE 310
12-15,
1982, San Antonio, Texas.
Sho walter
AK. .1953: A stability index for thunderstorm
forecasting. Bull. Amer. Meteorol. Soc., v. 34, pp.
250-252.
Sun,. W-Y., and Y. Ogura. 1979: Boundary-lager forcing as a
possible trigger to a squall-line formation. J. Atmos.
Sci.. v. 36, pp. 235-254.
Tennekes, H.
1973:. A model for the dynamics of the inversion
above a convective boundary layer. J. Atmos. Sci., v. 30,
pp.
558-567.
Uccellini, L;W-.
1975: A case study of apparent gravity wave
initiation- of severe convective storms. Mon. Wea. Rev.
v. 103, pp. 497-513.
Wetzel, P.J., 1978: A detailed parameterization of the
atmospheric boundary layer. Atmos. Sci. Paper 302,
Colorado State University.
1982: Breakup of temperature inversions in
Whiteman, :C.D.
deep mountain valleys: Part I. Observations. J. App..
Meteorol., v. 21, pp. 270- 289.
Willi.iams, S.F., N. Horvath
and R.E. Turner, 1980: A
preliminary look at AVE-SESAME II conducted on April
19-20, 1979. NASA Tech. Mem., NASA TM-78280.
Yamamoto,
.
1952: On a radiation chart. Sci.
v. 4
pp. 9-23..
Univ. 3. 5 Geophysics
Zeman
0, 1975:
Rept.
Tohoku
The dynamics of entrainment in the planetary
boundary layer: A study in turbulence modeling and
parameterization. Ph.D. Thesis, Penn. State Univ.
Zeman, O, and H. Tennekes, 1977: Parameterization of the
turbulent energy budget at the top of the daytime
atmospheric boundary laLer. J. Atmos. Sci.
v. 34, pp.
111-123.
S.S. , 1975: Comment on "A model For the
dynamics of the inversion above a convective boundary
Zilitinkevich,
lauer".
J.
Atmos.
Sci.
v. 32, pp.
991-992.
PAGE 311
AC kNOWLEDEMENTS
There are many
leading to
Tesearch
preparation
wif.e,
people who have
project.
as well
thesis,
of the thesis
Beverly,
financially
this
itself.
for supporting
and
emotionaily
Without that.
contributed to the
as the actual
Firstly,
our
I thank
family and me,
my
both
throughout this four year
the work would have been
s.uppor t,
impossible.
I
Secondly.
thank
aru.st
Professor Sanders for
providing the grant money which
I benefitted
Professoar
a.lso
greatly
fyrom the many
Sanders provided
grateul
to my
finish
the
first
of the
Thirdl,
I
inhabitants of
rewarding
the
c ompan ionship.
his
the funds
the
term of
expres
residence
Brad C olman fOr
at
my
-or
so long.
for travel.
I am
to Jane McNabb,
last possible moment before
1982-1983.
appreciation
16th Floor
to the
for an enjoyable and
in b;uildin
fr
me
conferences to which
Thesis Committee and
For helping me
end
supported
end ship:
54.
1 especlallh
support
and
t hank
_
____
__
_~~____~
~~___
_ _
_1
_____1___1_~__~_1________ ___~___~_1
PAGE 312
Finally,
allowed me
I
I
am grate-ful
to begin
finished my
thesis
post-doctcral
at MIT.
to Don Grantham
rese;ar.ch
of AF(L who
at AFGL while
____I~