Why observe temperature, water vapor, and cloud structures with high-­‐spectral-­‐resolu9on infrared observa9ons at 1-­‐km horizontal scales? Brian H. Kahn and Evan Fishbein Jet Propulsion Laboratory, California Ins9tute of Technology The 20th Interna9onal TOVS Study Conference Lake Geneva, WI, USA November 2nd, 2015 Copyright 2015 California Ins9tute of Technology. Government sponsorship acknowledged. What this talk is not about Ver9cal resolu9on Temporal resolu9on Spa9al coverage Sampling in thick/uniform clouds/precipita9on “Value tradeoffs” among these topics and with spa9al resolu9on Dependence of spa9al variability on cloud regime Global NICAM model (7-­‐km resolu9on) A large diversity of T structures Scales of variability dependent on cloud regime Boundary Layer Kine9c energy power spectra in NICAM Break from roughly –3 to –5/3 in mesoscale range Diffusive behavior at smallest scales Terasaki et al., Characteristics of the Kinetic Energy Spectrum of NICAM Model A
Terasaki et al. (2009), SOLA Global NICAM model (7-­‐km resolu9on) Even larger variability of q compared to T Scales of variability oEen inversely related to T Boundary Layer Global NICAM model (7-­‐km resolu9on) A large diversity of T structures Scales of variability dependent on cloud regime Lower Free Trop Global NICAM model (7-­‐km resolu9on) Even larger variability of q compared to T Scales of variability oEen inversely related to T Lower Free Trop A brief focus on trade Cu regime averaging kernels is about 2–3 km [Maddy and Barnet, 2008],
leading
to vertically
AIRSyield profiles
AIRS
Greatest AIRS (smoothed
IASI? CrIS?) in land
ow limited
lat oceans vertical resolution [Martins et al.,
2010]. Therefore, the fine
Very high skill in trade cumulus Cloud regime very important for cloud-­‐climate feedback gure 2. AIRS yield as a function of CloudSat Sc cloud fraction following Figure 1. The gray shadi
er the ocean indicates grids containing less than 5 AIRS retrievals.
5 of 12
Yue et al., 2011, J. Geophys. Res. JPL-­‐LES near Barbados during RICO campaign q in PBL highly variable at scales < 1 km Domain size very similar to AIRS, CrIS, and IASI FOV Some key issues (horizontal resolu9on) Climate/NWP model evalua9on/parameteriza9on Confron9ng a new genera9on of high spa9al resolu9on models with low spa9al resolu9on satellite soundings Scale-­‐dependence of pdfs related to cloud regime, al9tude, la9tude, etc. NWP model data assimila9on “Hole hun9ng” more successful at fine spa9al resolu9on High value per pixel in cloudy scenes within large horizontal T/q gradients Representa9on error of q (regime, la9tude, height dependence) (c.f., Hyoung-­‐Wook Chun talk Thursday) CRM + parameteriza9on for BOMEX Variance depends on parameterizaTon (CLUBB) and resoluTon of CRM Total water (vapor, cloud, precip) Larson et al., 2012, Mon. Wea. Rev. Small-­‐scale variance and cri9cal RH in climate models: Substan9al regime dependence Lower criTcal RH suggests larger variance of T and q Figure 2. Geographical distribution of the annual mean critical relative humidity
rc for(2012), selected vertical
Quaas J. Geophys. Res. Comparing AIRS and model variance Scaling exponents & breaks depend on al9tude Kahn and TKahn
eixeira (2009), J. CJ.limate and Teixeira
(2009),
Climate
Mesoscale “break” in AIRS T apparent but negligible for q –3
–2
–5/3
Kahn and TKahn
eixeira (2009), J. CJ.limate and Teixeira
(2009),
Climate
100
0.8
Pres (hPa)
ECMWF
(YOTC)
Models with data assimila9on more comparable to AIRS 1000
2
0.7
0.8
0.5 0.6
0.5
0.5 0.8 0.7 1 0.8
0.5 0.60.7 0.9
0.6 0.6
0.5
0.8
0.5
0.50.7
0.8
0.7
0.8
0.7 0.6 0.6 0.7
0.5
0.5
Pres (hPa)
MERRA
0.8
1
3
0.7
0.9
0.8
0.9
0.7
0.7 0.7
0.8 0.8 0.6 0.8
0.9
0.6
0.6
0.7
0.6 0.5
0.70.6
1000
Pres (hPa)
AIRS
0.8
2
0.7
4
5
6
7
8
0.8
0.5
0.80.6
0.6
0.7
0.5
0.6 0.5
0.7 0.5
0.70.6 0.5
0.5
0.7 0.8
0.5
0.6 0.5
0.7
0.5
0.8
0.50.3 0.6
0.6 0.8 0.4 0.4 0.7 0.7 0.5
0
AIRS
0.5
50
0.4
0.4
0.4
0.6
0.5 0.4
0.5
0.6
0.6
0.3
0.5
0.6 0.5 0.4
0.5
0.7
0.6
0.3
0.4
0.5
0.6
0.3
0.4
0.5
0.4
0.5
0.5
MERRA
0.5 0.2 0.4
0.3
0.4
0.4
0.2
0.3
0.3 0.3
0.4
0.6
0.6
0.6
0.7
0.5
0.6
MERRA
0.9
0.6 0.9
0.9 0.3 0.3
0.8
0.6 0.20.4 0.7
0.5
0.7 0.3 0.5 0.6
0.9
-50
0.5
0.6
0.3
0.8
0.40.4
ECMWF
0.6
0.3
MERRA
100
0.4
0.4
0.8
5
6
7
8
0.5
ECMWF
0.8
0.7
0.8
0.8
0.7
0.4
0.4
0.5 0.5 0.6 0.7
0.5
ECMWF
4
3
0.6
0.6
0.4
0.5
0.7
0.5
0.7
0.5
0.6
0.6
ECMWF
2
0.6
0.6
100
1000
0.5 0.5
0.6
4
5
6
7
8
9
0.4
0.5
0.6
0.7
0.9
3
0.5
0.6
0.6
0.4
MERRA
0.2
0.3
0.4
0.3
0.3
0.3
0.2
0.3 0.2
0.3
0.3
0.4
-50
0.5
0.3
0.5
0.5
0.3 0.4
0
AIRS
50
-50
0.3
0
AIRS
50
-50
0.4
0
50
AIRS
Kahn et al. Kahn
(2011), J. Atmos. ci. et al. (2011),
J. Atmos.SSci.
“Free-­‐running” models not as comparable to AIRS Pres (hPa)
NCAR
CAM3
100
2
1
0.7 0.6
0.9 0.6
1
0.5
3
0.8
0.9 0.8
0.8 1
0.9
0.6
0.8
0.90.6 0.5
0.6
0.9 0.8 0.5
1000
0.7
0.7
0.7
Pres (hPa)
GFDL
AM2.1
0.7
1
1
0.9
1 0.8
1
1
2
0.80.6
5
6
7
8
9
1
0.8
0.9
0.6
0.7 0.9
0.6 0.7
0.6
0.7 0.8
4
Pres (hPa)
AIRS
0.8
1000
2
0.7
3
4
5
6
7
8
0.9
0.8
1
0.9
0.5
0.6 0.5
0.7
0.5
0.8
0.50.3 0.6
0.6 0.8 0.4 0.4 0.7 0.7 0.5
0
AIRS
50
0.6
0.6
0.4
0.6
0.6
CAM3
0.8
0.8
0.5
0.7 0.6
0.6
0.6
0.6
0.5 0.2 0.4
0.3
0.4
0.4
0.2
0.3
0.3 0.3
0.6
0.6 0.7
0.6
0.6
0.6
0.5
0.5
0.6 0.7
0.6
0.7
0.7
0.8
0.6
0.6
0.7 0.5
0.5
0.6
0.70.5
0.6
0.5
C180HIRAM2.1
0.3
0.4
0.5 0.5
CAM3
0.7
0.7
0.6
0.6
0.4
0.4
0.6
0.6
0.4
0.5
0.8
0.5
0.6
0.6
0.7
0.7
0.7 0.7
0.7
0.6
0.6
0.6
0.7
0.7 0.6 0.7
0.6
0.5
0.7 0.6
0.6
C180HIRAM2.1
0.9
0.6 0.9
0.9 0.3 0.3
0.8
0.6 0.20.4 0.7
0.5
0.7 0.3 0.5 0.6
0.9
-50
0.7
0.8
0.6
C180HIRAM2.1
100
0.5
0.7
CAM3
0.6
0.7
0.6
0.7
0.6
0.7
0.7
0.5
0.6
0.7
0.7
0.7
0.7 0.6
0.6
0.6
CAM3
1000
0.8
0.7 0.8
0.7
0.9
5
6
7
8
3
0.7
0.6
4
100
0.9
0.7 0.8
0.9
0.6
0.6
0.5
C180HIRAM2.1
0.2
0.3
0.4
0.5
0.3
0.3
0.3
0.2
0.3 0.2
0.3
0.3
0.4
-50
0.5
0.3
0.5
0.5
0.3 0.4
0
AIRS
50
-50
0.3
0
AIRS
50
-50
0.4
0
50
AIRS
et al. (2011),
J. Atmos.SSci.
Kahn et al. Kahn
(2011), J. Atmos. ci. Can we use current 1-­‐km CWV observa9ons to address scaling? Ambiguity between CWV and height-­‐resolved q ExisTng 1-­‐km resoluTon CWV observaTons might fall short on this issue of CWV resembles E R I C S C I E NScaling CES
OLUME
68T, height-­‐resolved q close to –2 K A H N E T VA
L.
2165
157.58–1708W
(68–188N,
(b) 48S–68N,
.8–208N,
As in Fig.
5, but 1-s values
of q1568–1688W),
shown at constant
pressure levels in SP-CAM and AIRS.
S, 848–968W). All spectra are from September–November
–08 (AIRS). The scaling exponents for 25/3 (0.33) and 23
orous evaluation of the model varrrants further investigation.
4. Discussion and conclusions
Kahn et al. (2011), J. Atmos. Sci. Scaling of qt approximately –2 at all scales SimulaTons based on trade cumulus regime (RICO) Averaged over height (column) 3622
JOURNAL OF THE ATMOSPHERIC SCIENCES
FIG. 6. Power density spectra of the intermediate wavenumber
et ofal. (F2013), J. Fig.
Atmos. Sci. IG. 7. As in
6, except
the wh
space of total water mixing ratio. The integralSchemann over the spectra
that serve as sub
not necessarily ens
constraints from s
v
continue to play a
zation developme
2166 aircraE observaTons J Ow
U ithin/above R N A L O F TsH
E ATMOSPHERIC S
VOCALS-­‐REx tratocumulus This work show
lution
observation
necessary
to de
necessary to obser
climate sensitivi
scale ‘‘turbulence’
and cloud feedb
same time, current
tests of ECMW
atmospheric sound
tialization.
sessing
climateAdd
pro
(1999)
useco
marks
forthat
model
from o
ofobtained
more rigorous
the SP-CAM res
parameterizations;
that
as re
s
egy
to serve
monitor
not necessarily
spectrum
of T and e
constraints
from
spectrum
observed
continue
to play
previous
observati
ieszation
that demonstra
developm
in the
neighborhoo
This
work sho
However,
the scal
lution observatio
ature
(not heightnecessary
to obs
pronounced
in th
scale ‘‘turbulenc
boundary
layer
a
same time,
curre
FIG. 7. Shown are composite variance spectra for
(top)et q aand
Kahn l. (2011), J. Atmos. Sci. popause.
Temper
Height-­‐dependent scale break in q near 5–10 km What about clouds? Scale dependence of cloud number, coverage, and reflectance sensi9ve to spa9al scale Wood and Field (2011), J. Climate scene boundary and aircraft flight tracks are shown
Electra repeatedly gathered data across a stratocudown from the 1000 km baroclinic forcing scale,
- 6
in
Fig. 2. The ER-2 flew a "racetrack" pattern at
mulus-fair
weather0 cumulus
transition.
1
2
3
- 5 /curve
3 atina Fig.
few2),hundred kilometers,
about and
60,000change
feet (the todashed
The large-scale cloud pattern on 7 July may be
LOG WAVENUMBER (1/km)
seen in Fig. 1, which shows three channels of a
while the
BMO
C-130
and
the
NCAR
Electra
flew
continuing to the smallest observed
scale of a few
AVHRRspectrum
scene acquired
16432GMT
Figure lOa).NOAA-10
Wavenumber
of TM atBand
reflectiv-in and just above the cloud layer. The Electra
The
stratocumulus
(9:43 am cumulus
local time).
The thermal
(Channel
flew a "butterfly"
pattern
(the solid curve brightness spectra
ity in fair weather
averaged
over band
10 scan
lines of initiallykilometers.
a very
uniform
cloud-top
2), and thenanother
repeatedly change
traversed the
totran-- 3 at small scales
subscene a. 4)Asshows
expected
from
Fig. 6a),
there temperature,
is a change of in Fig. suggest
while
the
near-infrared
and
visible
reflected
bands
sition
region
on
east-west
flight
legs,
as
the
strascaling properties near 1/2 km. The smaller scales follow a
associated with convective forcing.
It will be intershow
patchy
stratocumulus
over
most
of
the
area.
k-3 power-law, expected for three-dimensional turbulence oftocumulus boundary advected to the east.
esting to see if this break occurs in stratocumulus
The right
the scales
visible approach
band (Channel
a passive scalar,
whileside
the of
larger
a fiat 1)white Two 60 x 60 km (20482 pixel) subscenes, one
shows apparently solid stratocumulus hugging the
in the fair
weather
cumulus
and onewater
in the spectrum
straliquid
water.
A liquid
from a single
noise spectrum.
coast, and bounded on the ocean side by an appartocumulus, with boundaries labeled in Fig. 2 as
cumulus cloud (King et al., 1981) gives a power of
ently clear patch, with the interface between the
"subscene a" and "subscene b," respectively, were
(km)"clear-cloudy"
- 1 .analysis.
8 down
the (km)
- 3 is
WAVELENGTH
chosen for
The to
Banda 7few
(2.2 meters.
/.tm) subim-Perhaps
two appearingWAVELENGTH
quite sharp. This
interface occurs
at the1.00
center 0.10
of the Landsat overages in not
Figs. found
3a) and in
3b), such
respectively,
showevolving
con10.00 clouds
1.00 0.10
10.00
rapidly
because
0
pass at 1800 GMT (11 am local). The solid square
siderable structure not observable in the AVHRR
the cloud
thickness
scalein isBand
not2, well-defined as it is
° Ithe approximate
k region of cloud seen by image. (The
same structure
is observed
indicates
"-*,,."~,",,
,./4
km
Landsat. As we shall see, the "clear" region is
which isinused
for the numerical analysis, butF the
stratocumulus.
-- 1
'
In
Band 7 image appears somewhat sharper
actually filled with a large number of fair weather
u.I because
it has no saturation, and possibly because of its
cumulus clouds, all well below AVHRR resolution.
k:3
i
r-,~
¢3Daddition, we shall
/ " see
,,
"
In
a large
t.ul-- - 2
'"F
2
[
~ that in the stratocumulus greater absorption.) Subscene a shows
i,l z j
!
N~.
number CONCLUSIONS
of small fair weather cumulus clouds, with
region the Landsat TM reveals considerable spatial
k-O.6
structure, again not resolved by the AVHRR.
diameters typically less than 1 / 2 km, aligned in
rr,N<<N
S
~j
-3
Three aircraft gathered in situ data across the
streets oriented parallel to the wind, r'r'u_
and separated
O omuch larger
We1-2have
provided
a detailed
study of cloud spatial
stratocumulus
boundary within the Landsat area
by about
km. Subscene
b shows
Oo
Z,,,
indicated in Fig. 1, using the Landsat TM center
structure in the California stratocumulus regime
Scale Break at 0.5 km in Sc with Landsat’s Thema9c Mapper (TM) Data \
O'< - 4 !
~D
!
-4
with
the7) fair
help
ofcumulus
data
fromon the FIRE Marine
Figure 3a).
TM (Band
weather
cJsubscene
Figure 2. Boundary of the thematic mapper (TM) scene on 7
O3 on the left
7
July
1987.
The
subscene
boundaries
are
shown
Structure
of
Marine
Stratocumulus
99
July 1987, and aircraft flight patterns at about the time of the
Stratocumulus
Intensive
Field
side of Fig.
2. Note the cloud streets
oriented parallel
- 5 to the Observations, focusLANDSAT overpass at 1800 GMT (11 am local). The strawin& ing on the 7 July 1987 Landsat Thematic Mapper
tocumulus boundary is about in the center of the scene, so
that the two 60 km subscenes labelled 3a and 3b are on the
Landsatreferred
TM Reflected
Bands
WaveWe have shown that
the surface, cloud base,
clear mad cloudy sidesTable
of the 1.boundary
to in Fig.
1. with Nominalscene.
- 6
length
Ranges
andatMaximum
Reflectances
0
1
2
3
The boundary
moved
steadily
eastward
about
60
kph,
so
-1
0
1
2
3
and cloud top temperatures LOG
determined
by spatial
that the Electra soundingBand
shown in Fig.
5,
taken
across
the
WAVENUMBER
(1/km)
Wavelength (izm)
Rs,,t
Saturated?
WAVENUMBER
boundary anLOG
hour later,
was at the right (1/km)
edge of the TM
coherence analysis of the TM thermal band are
1
0.45-0.52
0.25
t/
scene.
Figure lOa). Wavenumber spectrum of TM Band 2 reflectivFigure lOb). Wavenumber spectnlm
of TM
Band 2 reflectiv2
0.52-0.60
0.58
validated by
situ aircraft
soundings.
33.5.
ity the
in fairinweather
cumulus averaged
over The
10 scan lines of
3
0.63-0.69
0.53
t
ity in stratocumulus averaged over 10 scan lines of subscene
4
0.76-0.90
0.73
agreement subscene
is especially
close for
base
a. As expected
fromthe
Fig. cloud
6a), there
is a change of
NASA ER-2
b. The scales smaller
a path
k 3.0,similar
33 than about5 200 m follow
flight
1.55-1.75
0.47
t/
scaling
properties
near
1/2
km.
The
smaller
scales
threshold, upon which previous fractal studies are follow a
7 Tthe
2.08-2.35
0.69 the
to the fair weather cumulus,LANDS~
while
scales follow
TM Slarger
k-3 power-law, expected for three-dimensional turbulence of
/ ~ ~
CE~
,
same k -3/3 law seen
water
Fig.
9. corrected
Peaks for
at solar
4 zenithbased.
32.5 in the
/ ^ \liquid
\ Source:
- _ in
/(1986),
From
Clark
angle of We have contrasted the fractal properties of
a passive scalar, while the larger scales approach a fiat white
30 ° .
and 8 km correspond to / fthe
cloud streets
seen inindicates
Fig. 3b).
the last column
that most cloud pixels
the arefair weather
cumulus and stratocumulus sub/ - - % - 5A, check in.....
noise spectrum.
D 32
BMo c13o
saturated, so that the gain settings of these bands are of limited use in
cloud analysis.
scenes of the 7 July scene. The fact that such
divergent spatial structures areWAVELENGTH
observed in(km)close
Thus at least for the
larger
we expect a
Sub......
%\ "X~ scales
~/
saturated. Bands 4 and 7 are well below saturation.
1.00 "mixing"
0.10
proximity underscores the 10.00
potential
direct correlation
reflectivity
and liquid
31 1 between
\ ~ NCAR ELECTRA
The" - Jhistogram
of the thermal band (Band 6,
'~ flight path
° I results ink the analysis
water, though 30.5
details of
the structure
two at thewhich
104-12.5
m/x) for a 15ofkmthe
subscene
stra- can lead to misleading
12-3
-1:92
-1'21
"-*,,."~,",, ,./4 km
edge is shownspectrum
in Fig. 4a). The
of two
coarse-resolution meteorological
fields will likely differ.tocumulus
The
reflectivity
LONGITUDE
-- 1
'
I n data. We have
narrow peaks, separated by about 5°C, correspond
emphasized
importance
of the darker,
optically
also shows peaks at 4 and
8 km corresponding to
Cahalan the
and Snider (1989), Remote. Sens. Env. 7) stratocumulus subscene on 7 July
to the surface and cloud top. A dry adiabatic lapse
_J
31.5
Figure 3b).
T M (Band
What about the future? CubeSat Infrared Atmospheric Sounder
(CIRAS)
CIRAS will significantly reduce the cost
of atmospheric sounding in the infrared
and enable improved timeliness
through constellations
For NASA InVEST
PI: Tom Pagano (JPL)
Sponsor: NASA ESTO
CIRAS Technologies
CIRAS Mission
Micro Pulse Tube
Cryocooler (Lockheed
Martin)
•  Demonstrate Key Technologies needed for Infrared
Instruments on CubeSats
•  Demonstrate fidelity of Hyperspectral Mid IR
radiance measurements to retrieve Temperature and
Water Vapor Profiles
•  Fill Coverage Gaps and Improve Timeliness of
Operational IR Sounders
•  TRL in: 5-6, TRL out: 7
•  Build: 2016, 2017. Launch 2018 (TBD)
Parameter
Spatial
Orbit Altitude
Scan Range
Horizontal Res’n
Spectral
Method
Band 1
Res’n / Sampling
Total Channels
Radiometric
NEdT (@250K)
Resources
Size
Mass
Power
Data Rate
CIRAS
600-850 km
0.84° - 57°
1.6 km - 13.5 km
Grating
4.78-5.09 µm
0.5 / 0.2 cm-1
625
<0.25 K
6U Cubesat
8.5
37.5
2 Mbps
© 2015 California Institute of Technology. Government
sponsorship acknowledged.
JPL GRISM
Spectrometer
CIRAS Measurements
•  Lower Tropospheric Temperature Profiles
•  Lower Tropospheric Water Vapor Profiles
•  Goal: Experimental Demonstration of 3D Winds
JPL HOT-BIRD
Detector