Click
Here
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 111, D19203, doi:10.1029/2005JD006802, 2006
for
Full
Article
Cloud-to-ground lightning and surface rainfall in warm-season
Florida thunderstorms
Bruce Gungle1,2 and E. Philip Krider1
Received 23 October 2005; revised 27 February 2006; accepted 20 April 2006; published 11 October 2006.
[1] Relationships between cloud-to-ground (CG) lightning and surface rainfall have
been examined in nine isolated, warm-season thunderstorms on the east coast of
central Florida. CG flashes and the associated rain volumes were measured as a
function of time in storm-centered reference frames that followed each storm over a
network of rain gauges. Values of the storm-average rain volume per CG flash ranged from
0.70 104 to 6.4 104 m3/CG flash, with a mean (and standard deviation) of 2.6 104 ± 2.1 104 m3/CG flash. Values of the rain volume concurrent with CG flashes ranged
from 0.11 104 to 4.9 104 m3/CG flash with a mean of 2.1 104 ± 2.0 104 m3/CG flash. The lag-time between the peak CG flash rate and the peak rainfall rate
(using 5 min bins), and the results of a lag correlation analysis, show that surface rainfall
tends to follow the lightning (positive lag) by up to 20 min in six storms. In one storm the
rainfall preceded the lightning by 5 min, and two storms had nonsignificant lags. Values of
the lagged rain volume concurrent with CG flashes ranged from 0.43 104 to 4.9 104 m3/CG flash, and the mean was 1.9 104 ± 1.7 104 m3/CG flash. For the five storms
that produced 12 or more flashes and had significant lags, a plot of the optimum lag
time versus the total number of CG flashes shows a linear trend (R2 = 0.56). The number of
storms is limited, but the lag results do indicate that large storms tend to have longer lags. A
linear fit to the lagged rain volume vs. the number of concurrent CG flashes has a slope
of 1.9 104 m3/CG flash (R2 = 0.83). We conclude that warm-season Florida
thunderstorms produce a roughly constant rain volume per CG flash and that
CG lightning can be used to estimate the location and intensity of convective rainfall
in that weather regime.
Citation: Gungle, B., and E. P. Krider (2006), Cloud-to-ground lightning and surface rainfall in warm-season Florida thunderstorms,
J. Geophys. Res., 111, D19203, doi:10.1029/2005JD006802.
1. Introduction
[2] The processes that electrify clouds and produce
lightning depend on complex interactions between the
cloud microphysics and dynamics, characteristics that in
turn depend on the type and location of the storm and the
larger meteorological environment [Williams et al., 1989;
Zipser, 1994; Baker et al., 1995; Boccippio et al., 2000,
2001; Blyth et al., 2001; McCaul et al., 2002; Lang and
Rutledge, 2002; Toracinta et al., 2002; Williams et al.,
2002; Brown et al., 2002; Carey and Rutledge, 2003;
Seity et al., 2003]. Since precipitation is thought to play a
key role in electrifying clouds on both small and large
spatial scales (see MacGorman and Rust [1998] for a
review of the literature on thunderstorm electrification),
and since there are still many questions about how
1
Institute of Atmospheric Physics, University of Arizona, Tucson,
Arizona, USA.
2
Now at U. S. Geological Survey, Water Resources Discipline, Tucson,
Arizona, USA.
Copyright 2006 by the American Geophysical Union.
0148-0227/06/2005JD006802$09.00
lightning relates to rainfall on various scales of space
and time [Petersen and Rutledge, 1998; Tapia et al.,
1998; Baker et al., 1999; Soula and Chauzy, 2001; Seity
et al., 2001], we have undertaken a study of relationships
between cloud-to-ground (CG) lightning and surface rainfall in Florida. The motivation for this work has been to
determine whether real-time measurements of the location
and intensity of CG lightning can assist forecasters in
making quantitative estimates of the location and amount
of convective rainfall, especially in regions where radar
coverage is poor or incomplete.
[3] Tables 1 and 2 summarize some recent studies of
relationships between lightning and the associated precipitation volume (or mass) that is inferred from rain gauge and
weather radar measurements, respectively. The spatial and
temporal scales of these studies range from individual
thunderstorms, or even the cells within storms, to synoptic
and climatic scales where the pattern of lightning and
rainfall over broad regions is considered for days, months,
and/or years. The studies of individual storms tend to focus
on intervals when there is active convection and lightning,
whereas the synoptic and climatic studies tend to include all
precipitation that is detected within the region of interest,
even that which is not associated with lightning. Because of
D19203
1 of 15
D19203
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
D19203
Table 1a. Precipitation Volume (PV) Per Lightning Flash Reported by Investigators Who Used Gauges to Measure the Precipitation
PV Per Total Flasha
(104 m3/flash)
PV Per CG Flash
(104 m3/CG Flash)
Author(s), Location, and Date
Battan [1965]: Tucson, AZ
3.0
–
Piepgrass et al. [1982]: KSC, FL
Holle et al. [1994]: Kansas and
Oklahoma
Gosz et al. [1995]: New Mexico
Sheridan et al. [1997]b: Gulf
Coast and Central Plains
Soriano et al. [2001]: Iberian
Peninsula
Ezcurra et al. [2002]: Northeast
Spain
1.8 – 2.2
20
0.67 – 0.85
–
3.2 – 4.0
<0.15; 0.15 – 1.0; >1.0
–
–
12; 21
–
21, 23, 38 68, 120, 136
–
4.0 – 430
–
0.11 – 6.4
–
Kempf and Krider [2003]: Upper
Mississippi River Basin
Present study [2006]: KSC,
Florida
Comments
Visual CG count, over four summer
seasons, 800 – 1000 km2
Two isolated storms
Mesoscale Convective System (MCS)
Over three summer seasons and 1000 km2
1989 – 1993 warm seasons, six areas
2.58 – 13.1 103 km2, by day
1992 – 1994 warm seasons, 102 km2, by
months: semiarid; humid
5 year totals at three locations; daily totals
at three locations; CGs in 20 km2 box
centered on gauge
1993 ‘‘Great Flood,’’ 1.3 109 km2; daily
averages
Nine isolated storms
a
The total flash count is the sum of the intracloud (IC) and cloud-to-ground (CG) lightning flashes.
Sheridan et al. [1997] did not include lower and upper values of PV/CG flash; however, thunderstorm days were divided into one of three PV/CG flash
categories.
b
meteorological conditions that produce thunderstorms over
the Florida peninsula [Piepgrass et al., 1982; Watson et al.,
1987; Watson et al., 1991; Reap, 1994; Laird et al., 1995;
Jameson et al., 1996; Bauman et al., 1997; Wilson and
this, the synoptic and climatic studies tend to find larger
precipitation volumes per flash than the studies that focus
on individual storms or the cells within storms.
[4] In recent years, there has been much research on the
Table 1b. Precipitation Volume (PV) Per Lightning Flash Reported by Investigators Who Used Radar to Measure the Precipitation
Author(s), Location, and Date
Kinzer [1974]a: Oklahoma
Grosh [1978]: St. Louis, Missouri
Maier et al. [1978]: South Florida
Goodman et al. [1988]: Huntsville,
Alabama
Goodman and Buechler [1990]:
Alabama/Tennessee line
Buechler et al. [1990]: Tennessee
Valley
Nielsen et al. [1990]: Oklahoma
PV Per CG Flash
(104 m3/CG flash)
PV Per Total Flash
(IC Plus CG) (104 m3/flash)
Comments
1.6*
–
1.1 – 9.2
2.0
–
1.4
–
0.1
One squall line storm
One isolated storm
Over 20,000 km2 and 24 days
One microburst-producing storm
1.1 – 9.9
0.12 – 0.26
0.85 – 54.1
–
Six isolated storms (PV/CG), and two
isolated storms (PV/TF)
21 storms over 6 days
4 – 10
–
Lopez et al. [1991]: Cape
Canaveral, Florida
Williams et al. [1992]: Darwin,
Australia
Holle et al. [1994]: Kansas and
Oklahoma
Buechler et al. [1994]: South
Florida
Senesi et al. [1996]: Southern
France
Petersen and Rutledge [1998]:
several locations
Soula et al. [1998]: Spain
Tapia et al. [1998]: KSC,
Florida
Molinie et al. [1999]: Spain
0.1 – 1000
–
PV/CG varies over life of 1 Mesoscale
Convective System (MCS)
Over 14,800 km2; hourly PV/CG ratios
50 (continental),
500 (monsoon)
76.9
–
Two wet seasons; about 40,000 km2
–
MCS
10.8
–
2 months over 90,000 km2
3.0
–
Squall line system
5.7 – 2000
–
11 regimes over months and 100,000 km2
3.1
2.4 – 36.5
–
–
One storm, possibly supercell
22 storms in August 1992 and 1993
0.32 – 4.7
–
Soula and Chauzy [2001]: Paris,
France
Seity et al. [2001]: Bordeaux,
France
Zhou et al. [2002]: East Gansu,
China
5.10 – 9.94
0.75 – 2.04
Three isolated cells and three cells embedded
in an MCS
Three summer 1997 and one spring 1999 storm
0.92 – 113.0 1.15 – 106.6
0.204 – 12.89 (land)
0.156 – 31.75 (ocean)
–
2.4
a
21 stormy days over 240 km2 of land and ocean
Median value observed in thunderstorms in East
Gansu
Kinzer notes there may have been a systematic 2 dB error in the NSSL radar data; in that case, the PV/CG flash ratio would be 2.6 104 m3.
2 of 15
D19203
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
D19203
Table 2a. Lag Times Between Lightning and Surface Rainfall Reported by Investigators Who Used Gauges to Measure the Rainfall
Author(s), Location, and Date
Bin Size, min
Type of Lightning
(CG or IC)
Piepgrass et al. [1982]: KSC,
Florida
Kane [1992]: Northern Virginia
Watson et al. [1994]: Arizona
5 and 1
CG and IC
4 – 10
5
60
CG
CG
5 and 3
CG
5 – 15
Peak precipitation lags peak CG flash
rate (three of six regions) or is concurrent
(three of six regions)
0 – 20
Present study [2002]: KSC,
Florida
Megenhardt, 1997; Bringi et al., 1997; Hodanish et al.,
1997; Camp et al., 1998; Gremillion and Orville, 1999;
Lericos et al., 2002; Mazany et al., 2002]. Here, we describe
a storm-scale, gauge-based study of CG lightning and
surface rainfall that was produced by nine warm-season
thunderstorms at the NASA Kennedy Space Center (KSC)
and Cape Canaveral Air Force Station (CCAFS). Our
methods of analysis will be similar to those described by
Piepgrass et al. [1982].
[5] For each storm in our data set, we have measured the
CG lightning and area-average rainfall as a function of time;
we have computed the lag times between the lightning and
rainfall rates; and we have estimated the average rain
volume per CG flash with and without a correction for
the lag time between the lightning and rainfall rates.
Originally, 12 thunderstorms were considered for analysis
because they occurred over or near the rain gauge network,
but as the analysis progressed, three of these storms were
found to be unsuitable, either because the storm produced
substantial lightning and rainfall outside the perimeter of the
network or because the storm was not sufficiently isolated,
i.e., it was part of a larger system that produced substantial
lightning and rainfall outside the network. We will begin
with a description of the sensors that were used to measure
the lightning and rainfall, and then we will describe the
methods that have been used in the data analysis. Finally,
Optimum Lag in Precipitation, min
we will describe our results and how they compare with
other measurements.
2. Instrumentation
2.1. Cloud-to-Ground Lightning Surveillance System
[6 ] The number and locations of CG flashes were
obtained from the Cloud-to-Ground Lightning Surveillance
System (CGLSS) that is operated by the CCAFS. This
system contains five gated, wideband IMPACT sensors
similar to those used in the U.S. National Lightning Detection Network (NLDN) [Cummins et al., 1998a, 1998b] but
with lower gain. The location accuracy of the CGLSS is
approximately 0.6 km over the KSC-CCAFS, and the
detection efficiency is estimated to be better than 90%
[Maier, 1991; Maier and Wilson, 1996; CSR, 2005; Boyd
et al., 2005]. Prior studies have shown that a large fraction
of the low-amplitude, positive flashes reported by IMPACT
sensors are actually cloud discharges [Cummins et al.,
1998a], so positive reports have been included in our counts
only if the estimated peak current is greater than 15 kA.
2.2. Rain Gauge Network
[7] Figure 1 shows the locations of 31 tipping-bucket rain
gauges that were originally installed at KSC as part of the
TRMM ground validation program [Habib and Krajewski,
Table 2b. Lag Times Between Lightning and Surface Rainfall Reported by Investigators Who Used Radar to Measure the Rainfall
Author(s) and
Study Location
Workman and Reynolds
[1949]: New Mexico
Shackford [1958, 1960]:
Milton, Massachusetts
Carte and Kidder [1977]:
South Africa
Rutledge and MacGorman
[1988]: Oklahoma and
Kansas
Goodman et al. [1988]:
Huntsville, Alabama
Buechler et al. [1990]:
Tennessee Valley
Tapia et al. [1998]: KSC,
Florida
Molinie et al. [1999]:
Southern France
Soula and Chauzy [2001]:
Paris, France
Bin Size, min
Type of Lightning
(CG or IC)
–
CG and IC
–
CG and IC
First IC flash concurrent with appearance of precipitation at cloud
base; CGs follow ‘‘a few minutes later’’
3
–
CG
Precipitation both lags and precedes CGs
30
CG
30 to 0 (negative CG-convective rainfall) 45
(positive CG-stratiform rainfall)
1
total
2
10
CG
10
–
CG
‘‘significant lag’’
5
CG
‘‘precipitation occurs after the lightning flashes’’
5
CG and IC
10 to 0
3 of 15
Optimum Lag in Precipitation, min
D19203
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
D19203
Figure 1. Outline of the precipitation area (black lines) in four consecutive 15-min precipitation intervals
and the concurrent CG flash locations (black dots) during a thunderstorm on 28 April 1999 (990428).
2002]. The gauges were manufactured by Qualimetrics, Inc.
(Model 6011-A), and gauge tips were recorded in real time
by the CCAFS. The manufacturer states that the accuracy of
each gauge is about ±0.5% at a rainfall rate of 0.5 inches per
hour and that the resolution of each gauge is 0.254 mm
(0.01 inch) of rainfall per bucket tip. The network covers a
total area of about 450 km2, but the sensor spacing is
irregular with the distance between gauges ranging from
about 2 to 6 km. The average distance between gauges is
3.3 km, and the average network area per gauge is about
14.5 km2.
[8] During a storm on 28 April 1999 (990428), data from
gauges 10, 15, and 18 were not available, and on 10 June
1999 (990610A), the data from gauge 14 were missing. In
the analyses that are discussed below, any gauges that had
missing data were treated as if they did not exist, i.e., those
sensors were not included in our estimates of the precip-
itation areas or the associated precipitation volumes.
Sometimes gauges reported two or more bucket tips during
a 1 s interval, most commonly two tips but occasionally as
many as seven tips per second. The manufacturer did not
have information about the maximum number of tips that
would be produced by a heavy downpour, but when a gauge
was held under an open faucet, the maximum tip rate was
two per second, or about 30 mm min1 (Qualimetrics,
private communication, 2000). Therefore in our analyses
we have limited the maximum tip rate to two tips per
second. Multiple tips occurred in fifty-three 1-s intervals
in this study.
3. Data Analysis and Results
[9] Because thunderstorms are rarely stationary, their
evolution and movement must be taken into account when
4 of 15
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
D19203
analyzing relationships between lightning and surface
rainfall. In the following, we have used quasi-Lagrangian
(i.e., moving, storm-centered) reference frames for the data
analysis.
3.1. CG Lightning
[10] CG lightning flashes that were located within or just
outside the perimeter of a storm precipitation area (see
section 3.2) were counted over consecutive 5 min intervals.
Whether a flash was included or excluded from this count
depended on whether it was located within or near the
precipitation area and how that area evolved with time. The
times and locations of the lightning were used to determine
which precipitation area the flashes belonged to, and if the
storm was completely isolated, flashes up to 10 km from
the perimeter of the area were included. If two storms
occurred over the network at the same time, flashes were
assigned to the closest precipitation area.
3.2. Precipitation
[11] The precipitation area, Ap, was estimated for each
storm in consecutive 15-min ‘‘precipitation intervals,’’ and
within each of these intervals, the rainfall was grouped
into consecutive 5-min bins, the smallest interval that
could be used for averaging without producing significant
sampling error [Habib et al., 2001b]. If any gauge
produced one or more bucket tips (0.254 mm or more)
during a precipitation interval, that gauge was deemed
‘‘active,’’ and it was included in the precipitation area.
The precipitation area included all gauges that detected
rainfall in each 15-min precipitation interval, or the sum
of the areas of two or more groups of gauges that were
separated by a short distance (usually less than 5 km). In
cases where a second storm was active over the network,
the decision to include or exclude precipitation was made
using factors such as proximity to a cluster of CG
lightning, how the shape of the precipitation area evolved
with time, and the movement of the storm. In order to
compensate for edge effects, we assumed the perimeter of
the precipitation area extended 2 km beyond the locations
of all gauges that detected rainfall in that precipitation
interval. Figure 1 shows an example of how the spatial
pattern of CG lightning and the associated precipitation
area moved over the network during a storm on 28 April
1999 (storm 990428).
[12] The average depth of rainfall over the precipitation
area, Dp, was computed for each 5-min sampling interval
by averaging the rain reported by each active gauge in the
precipitation area,
Dp ¼
1X
di
n i
ð1Þ
where di is the rain depth [di = (0.254 mm) (no. tipsi)] that
was reported by gauge i, and n is the number of active
gauges. The precipitation volume, Vp, was computed by
multiplying the value of Dp in each 5-min interval by the
precipitation area, Ap, i.e.,
Vp ¼ Dp Ap
ð2Þ
D19203
The total precipitation volume, PV, was obtained by
summing the values of Vp from the time of the first CG
flash to the time of the last gauge tip.
[13] In four of the nine storms that were analyzed (see
Table 4 to follow), a small amount of precipitation was
detected at one or more gauges before the first CG flash
was detected, and there were no cases where the precipitation ended before the last CG flash. In storms 990428,
990509, 990610A, and 990611A, the time of the last
gauge tip could not be determined precisely because the
storm had moved outside the network or merged with
other areas of precipitation. In these cases, the stop times
were estimated from the movement of the lightning and
precipitation relative to the perimeter of the network (and
relative to other areas of lightning and precipitation). This
procedure may have underestimated the true precipitation
volume, but the rainfall rate peaked well before the rain
volume became indeterminate in all cases, and the CG
lightning stopped at least 10 min before the last gauge tip
in all storms.
[14] Sampling biases can be significant in rainfall
measurements, especially when the spacing between the
gauges is larger than the area of intense rainfall; indeed, it
is rare for the maximum or even near-maximum rain
depth to be recorded by a network of gauges [Weather
Bureau, 1947; Huff and Shipp, 1969; Huff, 1970]. The
spatial correlations and sampling errors to be expected in
rainfall measurements in central Florida have been discussed by Woodley et al. [1975], Habib et al. [2001a],
and Habib and Krajewski [2002]. If we ignore the
irregular spacing of the gauges and apply the factor-ofdifference method of Woodley et al. [1975], assuming that
the average area per gauge is about 14.5 km2, then we
expect our measurements of area-average rainfall should
be accurate to within a factor of 2 about 98% of the time,
and within a factor of 1.5 about 92% of the time.
[15] Figure 2 shows two examples of how the CG
lightning and surface rainfall rates evolved with time during
a small (Figure 2a) and a large (Figure 2b) thunderstorm.
Our estimates of the total PV in these storms were 6.7 105 m3 and 3.8 106 m3, respectively, and the average
PV per CG flash was 1.1 104 m3/CG flash and 2.3 104 m3/CG flash, respectively. The average rain mass per
CG flash or ‘‘rain yield’’ was 1.1 107 kg/CG flash and
2.3 107 kg/CG flash in Figures 2a and 2b, respectively.
3.3. Lag Times Between Lightning and Rainfall
[16] Prior studies at the KSC-CCAFS have shown that
the frequency of CG flashes tends to peak a few minutes
before the surface rainfall [Piepgrass et al., 1982; Tapia
et al., 1998], and a similar behavior is evident in Figures
2a and 2b. The third column in Table 3 summarizes the
lag times between the peak CG flashing rate (tCG,MAX)
and the peak rainfall rate (tp,MAX)
Dtpeak ¼ tCG;max tp;max
ð3Þ
(using 5-min bins) for all storms in our data set. In storms
where Dtpeak is positive, the lightning rate peaked before the
rainfall rate, and in the storm where Dtpeak is negative, the
rainfall rate peaked before the lightning rate.
5 of 15
D19203
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
D19203
Figure 2. Surface rainfall (shaded) and the number of CG flashes (black) in 5-min bins for (top) a small
storm on 5 August 1998 (980810B) and (bottom) a large storm on 10 August 1998 (980805B). The times
on the x-axis show the beginning of each bin.
[17] We also performed a lag-correlation analysis on each
storm to find the optimum lag time and to determine whether
this optimum was significant. We began the lag analysis by
computing a linear regression between the lightning and
rainfall rates (using 5 min bins), and then we shifted the
lightning rate forward and backward in 5-min steps and
recomputed the regression until the best fit was obtained.
Plots of the optimum fit between the lightning and rainfall rate
6 of 15
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
D19203
D19203
Table 3. Lag Times Between the CG Lightning and Surface Rainfall Rates
Storm
Number of CG Flashes
Optimum Lag Time, min
Lag in Peak Rates, min
PV/CG Flasha (104 m3/CG)
R2
P-Value
980805A
980805B
980810B
990428
990509b
990521
990610A
990611Ab
990611B
3
167
61
42
13
12
6
81
54
10
20
10
5
(0)
5
15
(5)
5
10
20
10
5
–
5
0
–
5
0.32
1.39
0.73
2.94
(1.18)
0.69
0.80
(0.30)
0.48
1.0
0.66
0.87
0.63
0.06
0.87
0.80
0.17
0.53
–
<0.0001
<0.0001
0.01
0.41
0.0021
0.041
0.13
0.0009
a
PV/CGF ratios are taken from the regressions shown in Appendix A.
P > 0.05; regression lags and lag in peak rates are indeterminate.
b
for each storm in our data set are given in Appendix A. The
values of the optimum lag-time and the associated R2 and Pvalues are summarized in Table 3. The P-value indicates the
likelihood that the correlation coefficient will occur at random
given a normal distribution of random samples from the parent
population, but because consecutive samples of both the
lightning and rainfall rates are not necessarily independent,
the R2 and P-values in Table 3 will tend to overestimate the
true significance of the regressions. Seven of the nine storms
in Table 3 had acceptable regressions, i.e., P-values less than
0.05. It should be noted, however, that the lagged correlation
for storm 980805A is based on just two points and thus R2 =
1.0 and P is undefined. Because the lag in the peak rates for
this storm agreed with the optimum (statistical) lag, this storm
has been retained in Table 3 and included in the ‘‘All Storms’’
plot in Figure 3.
[18] Figure 3 summarizes the optimum lag times as a
function of the total number of CG flashes for the seven
storms that had acceptable regressions. Although the dashed fit
in Figure 3 is clearly positive, the fit is not conclusive (R2 =
0.16, P = 0.37). If the two storms that produced less than
12 CG flashes are omitted from the analysis, because it is
likely that there were sampling biases in the rainfall measurements, then the fit in Figure 3 improves (solid line), but it
is still not statistically significant (R2 = 0.56, P = 0.15). It
should be noted that storm 980805B is the only storm in our
dataset that produced a large number of flashes (167 versus 61
in the next largest storm). Given the small sample, this storm
clearly has a strong influence on the statistics.
3.4. Precipitation Volume Per CG Flash
[19] In section 3.2, we defined the total precipitation
volume to be the sum of all rainfall from the time of the
Figure 3. The optimum lag time versus total the number of CG flashes for the seven storms that had
P < 0.05 (dashed) and for the five storms that had P < 0.05 and 12 or more CG flashes (solid).
7 of 15
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
D19203
D19203
Table 4. Number of Cloud-to-Ground Flashes (NCG), Precipitation Volume (PV) Calculated Three Ways (See Text), and the PV/CG
Flash Ratios Calculated for Each Method
Storm
980805A
980805B
980810B
990428
990509b
990521
990610A
990611Ab
990611B
Total
Median
Mean (PV)
Mean composite
PV/CG (SPV/SNCG)
Mean storm
PV/CG [S(PV/NCG)]/n
Standard deviation
PV Concurrent
Lagged PV
Total
Total PV/CG,
PV/CG Concurrent,
Lagged PV/CG
NCG PV, 104 m3 104 m3/CG With CGs, 104 m3
Concurrent, 104 m3 Concurrent, 104 m3/CG
104 m3/CG
3
167
61
42
13
12
6
81
54
439
42
49
2.90
384
66.9
167a
66.8a
11.2
38.1a
140a
37.8
915
66.8
102
0.97
2.3
1.1
4.0
5.1
0.93
6.4
1.7
0.70
0.32
302
55
158
67
6.5
29
101
27
746
0.11
1.8
0.9
3.8
5.1
0.54
4.9
1.2
0.51
1.7
1.3
337
61.3
151
–
9.1
29.1
–
22.5
611
1.2
83
0.43
2.0
1.0
3.6
–
0.76
4.9
–
0.42
1.0
87
2.1
1.7
1.8
2.6
2.1
1.9
2.1
2.0
1.7
a
Final gauge tip indeterminate; total storm precipitation volume may be an underestimate.
b
P > 0.05; regression lag indeterminate.
first CG flash to the time of the last gauge tip in the
precipitation area. When Piepgrass et al. [1982] computed
their precipitation volumes, however, they summed the
rainfall between the starting and stopping times of the
lightning (either intracloud or cloud-to-ground), and other
investigators have integrated between the starting and
stopping times of the precipitation whether lightning was
present or not [e.g., Kinzer, 1974; Grosh, 1978; Goodman
et al., 1988]. In order to facilitate a comparison of our
results with other investigators, we have integrated the
rainfall over three different intervals: (1) the entire storm
as defined in section 3.2; (2) an interval that is concurrent
with CG flashes, i.e., starting with the first CG flash and
ending with the last CG flash; and (3) over a lagged
interval that is concurrent with the CG flashes, i.e., for
storms that had a significant lag-correlation, the precipitation was shifted forward (or backward) by the optimum
Figure 4. Linear regression of the precipitation volume computed using methods 1, 2, and 3 (see text)
versus the total number of CG flashes in nine Florida thunderstorms.
8 of 15
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
D19203
D19203
Table 5. Precipitation Volume (PV) Per CG Flash and the Optimum Lag Times Between Lightning and Surface Rainfall in Florida
Type of Precipitation
Measurement
PV Per CG
Flash, 104 m3
Lag in Precipitation, min
Storm-Scale Studies
5 – 10 (5 min bins) 4 – 9 (1 min bins)
Comments
Piepgrass et al. [1982]:
KSC, FL
Tapia et al. [1998]:
KSC, FL
Present study [2005]:
KSC, FL
25 rain gauges,
250 km2
WSR-88D radar
1.8 – 2.2a
2.4 – 37b
‘‘significant lag’’
22 storms in August 1992 and 1993
31 rain gauges,
450 km2
0.70 – 6.4b 0.11 – 5.1a
5 – 20 (5 min lag)
Nine isolated storms
Maier et al. [1978]:
South Florida
Lopez et al. [1991]:
Cape Canaveral, FL
Buechler et al. [1994]:
South Florida
WSR-57M radar
1.1 – 9.2c
WSR-74 S-band
radar; 0.5 angle
WSI commercial
radar product
0.1 – 1000d
–
Over 14,800 km2 and 11.75 hours
11c
–
Over 90,000 km2 and 2 months
Large-Scale Studies
–
Two isolated storms
Over 20,000 km2 and 24 days
a
Precipitation during CG flashes only.
Total storm precipitation.
c
Multiple hour PV/CG.
d
Hourly PV/CG ratio.
b
lag time and integrated between the starting and stopping
times of the lightning.
[20] Table 4 summarizes the results. The values of ‘‘PV/
CG flash’’ in Table 4 do depend on which method is used
to integrate the rainfall, and it should be noted that the
large storms tend to have similar values of PV/CG flash
and that the small storms are more variable, probably
because of sampling biases in the rainfall measurements.
The smallest PV/CG flash in Table 4 is 0.11 104 m3/CG
flash (storm 980805A, precipitation concurrent with CG
flashes), and the largest is 6.4 104 m3/CG flash (storm
990610A, total PV). (Note that the final gauge tip was
indeterminate in storm 990610A; therefore this PV is likely
to be an underestimate.) The values of the precipitation
volume and the PV/CG flash ratios are largest when they
are computed using the total PV without corrections for
being concurrent with the lightning or the lag times.
[21] Figure 4 shows plots of the precipitation volume,
computed using each of the three ways discussed above,
versus the total number of CG flashes in the storm.
Although the number of storms is limited, all regressions
appear to be significant with R2 values ranging from 0.75 to
0.83. The best fit in Figure 4 (solid line) is for lagged
precipitation concurrent with CG flashes, and in this case
the slope corresponds to 1.9 104 m3 of (lagged) precipitation per CG flash.
4. Discussion
4.1. Previous Studies in Florida
[22] Table 5 summarizes the results of prior studies of
lightning and rainfall in Florida and our results. Maier et al.
[1978] examined the natural variability of CG flashes over
an area of about 2 104 km2 in south Florida using a
network of gated, wideband magnetic direction finders
[Krider et al., 1976, 1980] to locate and count the
lightning. Rainfall was measured using the National
Weather Service (NWS) WSR-57M radar in Miami. Maier
et al. found that storms that produced the highest lightning
frequency also produced the lowest PV/CG flash (1.1 104 m3/CG flash), and such storms occurred on days that
had normal or slightly above normal convection. The
average PV/CG flash was 2.6 104 m3/CG flash, and
Maier et al. noted that as the number of flashes increased,
so did the rain volume (i.e., the PV/CG flash remained
fairly constant) up to a total volume of 4 107 m3 to 8 107 m3. When storms produced a larger rain volume, the
number of flashes tended to be lower relative to the
rainfall (and the PV/CG flash ratios were larger). Lang
and Rutledge [2002] have noted a similar behavior in large
storms in the midwestern United States.
[23] Piepgrass et al. [1982] studied one small and one
large thunderstorm at the KSC-CCAFS using lightningcaused changes in the surface electric field and tippingbucket rain gauges. The storms were located directly over
the rain gauge network, and the precipitation areas were
about 100 km2 and 350 km2. The peak rainfall rate lagged
the peak lightning rate by 4 and 10 min in the small and
large storms, respectively, and the authors noted that lag
times of this order are consistent with the time it takes
elements of precipitation to fall from 7 to 8 km altitudes
(the 10 to 20oC temperature level), where the electrification is strong, to the surface. The rain volume per total
flash was 6.7 103 m3/flash and 8.5 103 m3/flash, and
the estimated PV/CG flash ratios were 1.8 104 m3/CG
flash and 2.2 104 m3/CG flash in the small and large
storms, respectively.
[24] Lopez et al. [1991] located and counted CG lightning near the KSC-CCAFS using three gated, wideband
direction finders similar to those described by Krider et al.
[1976, 1980] and they measured precipitation using a
NWS WSR-74 weather radar at Daytona Beach. The study
area was 14.8 103 km2, and the PV/CG flash ratios were
computed at hourly intervals under various prevailing
wind regimes. The ratios varied by 4 orders of magnitude,
from 103 to 107 m3/CG flash, depending on the wind
direction. The highest ratios (106 m3/CG flash) occurred
on days when winds were onshore from the northeast, and
9 of 15
D19203
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
the lowest (104 to 105 m3/CG flash) occurred on days
when the winds were calm or there was southwesterly
flow.
[25] Buechler et al. [1994] measured CG lightning over
south Florida using the NLDN and rainfall using a WSI
composite radar product on storms that produced convective
rain rates greater than 10 mm h1. The counts of CG flashes
and inferred rain volumes were grouped into 15-min bins,
and the best linear fit to the data (R2 = 0.86) corresponded to
a PV/CG flash of 1.1 105 m3/CG flash. Buechler et al.
concluded that CG lightning provided a better estimate of
the location and intensity of convective rainfall than a
convective stratiform technique that is based on satellite
measurements of the cloud-top temperature [Adler and
Negri, 1988].
[26] Tapia et al. [1998] studied 22 warm-season thunderstorms near the KSC-CCAFS using the NLDN and the
WSR-88D (NEXRAD) weather radar at Melbourne, Fla.
These authors found a ‘‘significant lag’’ between CG
lightning and the rainfall rate in the late afternoon which
they attribute to the formation of appreciable stratiform rain
(without lightning) late in the storm. Values of PV/CG flash
ranged from 2.4 104 to 3.6 105 m3/CG flash, with a
mean of 7.1 104 m3/CG flash and the composite mean
and median were 5.6 104 and 4.3 104 m3/CG flash,
respectively. Storms that produced the highest lightning and
rainfall rates had the lowest PV/CG flash, and the minimum
PV/CG flash occurred near the maximum CG flashing rate
in all storms. These authors also noted that storms with low
cloud tops tend to produce a higher PV/CG flash than
storms with high tops.
4.2. Precipitation Volume Per CG Flash
[27] As can be seen in Table 4, our values of the PV/CG
flash depend on which method is used to compute the
precipitation volume, and for each method, the results vary
by about a factor of 2 from storm to storm. Our mean ratios
are in excellent agreement with Piepgrass et al. [1982]
(2.0 104 m3/CG flash), and they are a factor of 2 to 6
lower than those of Buechler et al. [1994] and Tapia et al.
[1998], probably because these authors used radar to measure rainfall, and radar estimates of surface precipitation are
often different from gauge measurements [Brandes and
Wilson, 1988; Joss, 1990; Brandes et al., 1999; Habib
and Krajewski, 2002].
[28] Lopez et al. [1991] found a much larger variation in
the values of the PV/CG flash than this or any other
Florida study (a factor of 10,000 versus a factor of 10),
and they attribute this to differences in the flow regime,
with storms originating under a NE flow producing the
highest PV/CG flash and those under a SW flow producing the lowest. The discrete 1-hour sampling intervals may
also have exaggerated the extremes.
4.3. Lag Times
[29] Although Tapia et al. [1998] did report ‘‘significant’’ lags between lightning and rainfall in warm-season
Florida storms, Piepgrass et al. [1982] are the only other
authors who have examined individual storms with enough
time resolution to resolve the lag in the precipitation rate,
and the lag times in Table 3 are in good agreement with
D19203
Piepgrass et al. In our discussion of Table 3 and Figure 3,
we have noted that storms that produce larger numbers of
CG flashes tend to produce longer lag times. This was also
the case in the two storms that were studied by Piepgrass
et al. [1982]; their small storm had a lag time of 4 min and
the large storm had a lag time of 10 min.
[30] Williams et al. [1989], Rutledge et al. [1992], Zipser
[1994], and Petersen and Rutledge [1998] have noted that
storms that have high updraft speeds also have high CG
flashing rates; therefore it is reasonable to suppose that the
longer lag times in large storms are caused by higher
updrafts carrying precipitation particles to higher altitudes
where they are suspended longer and have a larger distance
to fall to the surface.
5. Summary
[31] We have examined relationships between CG lightning and surface rainfall in nine relatively isolated, warmseason thunderstorms in Florida, and we have found that
there is a linear relationship between the precipitation
volume and the number of CG flashes. Values of the lagged
rain volume per concurrent CG flash for individual storms
range from about 0.43 104 to 4.9 104 m3/CG flash,
with a mean and standard deviation of 1.9 104 ± 1.7 104 m3/CG flash. When the total rain volume per CG flash
is considered without a lag, the values range from about
0.70 104 to 6.4 104 m3/CG flash with a mean of 2.6 104 ± 2.1 104 m3 per CG flash. For the (unlagged) rain
volume that is concurrent with CG flashes, the values range
from 0.11 104 to 4.9 104 m3/CG flash with a mean of
2.1 104 ± 2.0 104 m3 per CG flash. A linear fit to the
lagged rain volume vs. the concurrent CG flashes (Figure 4)
suggests that the optimum PV/CG flash in Florida storms is
about 1.9 104 m3/CG flash.
[32] These results clearly support the suggestions of Gosz
et al. [1995], Chèze and Sauvageot [1997], Holle and
Bennett [1997], Soula et al. [1998], and Tapia et al.
[1998] that CG lightning can be a useful tool for estimating
the location and intensity of convective rainfall and for
generating near real-time warnings of flash floods, especially when weather radar signals are attenuated by heavy
precipitation or blocked by intervening topography. Our
data also support the suggestions of Buechler et al. [1994],
Tapia et al. [1998], and Alexander et al. [1999] that
lightning-inferred rainfall rates and the associated latent
heat transport might provide useful information for the
initialization of numerical forecast models. Since relationships between lightning and rainfall in other seasons and in
other geographic locations could well be different from
those in warm-season storms in Florida, we believe such
relationships merit further study.
Appendix A
[33] Figures A1– A9 show the regressions between the
precipitation volume at the optimum lag time, Dt, versus the
number of CG flashes in consecutive 5-min bins for each
storm in our data set.
10 of 15
D19203
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
Figure A1. Linear regression showing the volume of surface precipitation (5-min bins) at the optimum
lag time, Dt, versus the number of CG flashes for storm 980805A.
Figure A2. Same as Figure A1 except for storm 980805B.
Figure A3. Same as Figure A1 except for storm 980810B.
11 of 15
D19203
D19203
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
Figure A4. Same as Figure A1 except for storm 990428.
Figure A5. Same as Figure A1 except for storm 990509.
Figure A6. Same as Figure A1 except for storm 990521.
12 of 15
D19203
D19203
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
Figure A7. Same as Figure A1 except for storm 990610A.
Figure A8. Same as Figure A1 except for storm 990611A.
Figure A9. Same as Figure A1 except for storm 990611B.
13 of 15
D19203
D19203
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
[34] Acknowledgments. One of the authors (BG) has submitted this
research in partial fulfillment of the requirements for a M.S. degree in
atmospheric sciences at University of Arizona. This work was supported in
part by NASA under NRA-97-MTPE-03 (grants NAG5-6461 and NAG59208) and the NASA Kennedy Space Center (grant NAG10-302). The
authors would like to thank Scott Handel for his assistance in collecting the
data and for several helpful discussions.
References
Adler, R. F., and A. J. Negri (1988), A satellite infrared technique to
estimate tropical convective and stratiform rainfall, J. Appl. Meteorol.,
27, 30 – 51.
Alexander, G. D., J. A. Weinman, V. M. Karyampudi, W. S. Olson, and
A. C. L. Lee (1999), The effect of assimilating rain rates derived from
satellites and lightning on forecasts of the 1993 Superstorm, Mon.
Weather Rev., 127, 1433 – 1457.
Baker, M. B., H. J. Christian, and J. Latham (1995), A computational study
of the relationships linking lightning frequency and other thundercloud
parameters, Q. J. R. Meteorol. Soc., 121, 1525 – 1548.
Baker, M. B., A. M. Blyth, H. J. Christian, J. Latham, K. Miller, and A. M.
Gadian (1999), Relationships between lightning activity and various
thundercloud parameters: Stellite and modelling studies, Atmos. Res.,
51, 221 – 236.
Battan, L. J. (1965), Some factors governing precipitation and lightning
from convective clouds, J. Atmos. Sci., 22, 79 – 84.
Bauman, W. H., III, M. L. Kaplan, and S. Businger (1997), Nowcasting
convective activity for space shuttle landings during Easterly flow regimes, Weather Forecasting, 12, 78 – 107.
Blyth, A. M., H. J. Christian, K. Driscoll, A. M. Gadian, and J. Latham
(2001), Determination of ice precipitation rates and thunderstorm anvil
ice contents from satellite observations of lightning, Atmos. Res., 59 – 60,
217 – 229.
Boccippio, D. J., S. J. Goodman, and S. Heckman (2000), Regional differences in tropical lightning distributions, J. Appl. Meteorol., 39, 2231 –
2248.
Boccippio, D. J., K. L. Cummins, H. J. Christian, and S. J. Goodman
(2001), Combined satellite- and surface-based estimation of the intracloud-cloud-to-ground lightning ratio over the Continental United States,
Mon. Weather Rev., 129, 108 – 122.
Boyd, B. F., W. P. Roeder, D. L. Hajek, and M. B. Wilson (2005), Installation, upgrade, and evaluation of a short baseline cloud-to-ground lightning surveillance system in support of space launch operations, paper
presented at Conference on Meteorological Applications of Lightning
Data, Am. Meteorol. Soc., San Diego, Calif.
Brandes, E. A., and J. W. Wilson (1988), Measuring storm rainfall by radar
and rain gage, in Instruments and Techniques for Thunderstorm Observation and Analysis, Thunderstorms: A Social, Scientific, and Technological Documentary, vol. 3, 2nd ed., edited by E. Kessler, pp. 171 – 186,
Univ. of Okla. Press, Norman, Okla.
Brandes, E. A., J. Vivekanandan, and J. W. Wilson (1999), A comparison of
radar reflectivity estimates of rainfall from collocated radars, J. Atmos.
Oceanic Technol., 16, 1264 – 1272.
Bringi, V. N., K. Knupp, A. Detwiler, L. Liu, I. J. Caylor, and R. A. Black
(1997), Evolution of a Florida thunderstorm during the Convection and
Precipitation/Electrification experiment: The case of 9 August 1991,
Mon. Weather Rev., 125, 2131 – 2160.
Brown, R. A., C. A. Kaufman, and D. R. MacGorman (2002), Cloudto-ground lightning associated with the evolution of a multicell storm,
J. Geophys. Res., 107(D19), 4397, doi:10.1029/2001JD000968.
Buechler, D. E., P. D. Wright, and S. J. Goodman (1990), Lightning/rainfall
relationships during COHMEX, paper presented at 16th Conference on
Severe Local Storms, Am. Meteorol. Soc., Kananaskis Park, Alberta,
Canada.
Buechler, , H. J. Christian, and S. J. Goodman (1994), Rainfall estimation
using lightning data, paper presented at 7th Conference on Satellite Meteorology and Oceanography, Am. Meteorol. Soc., Monterey, Calif.
Camp, J. P., A. I. Watson, and H. E. Fuelberg (1998), The diurnal distribution of lightning over North Florida and its relation to the prevailing lowlevel flow, Weather Forecasting, 13, 729 – 739.
Carey, L. D., and S. A. Rutledge (2003), Characteristics of cloud-to-ground
lightning in severe and nonsevere storms over the central United States
from 1989 – 1998, J. Geophys. Res., 108(D15), 4483, doi:10.1029/
2002JD002951.
Carte, A. E., and R. E. Kidder (1977), Lightning in relation to precipitation,
J. Atmos. Terr. Phys., 39, 139 – 148.
Chèze, J.-L., and H. Sauvageot (1997), Area-average rainfall and lightning
activity, J. Geophys. Res., 102(D2), 1707 – 1715.
CSR (Computer Sciences Raytheon) (2005), Cloud-to-ground lightning
surveillance system, in Eastern Range Instrumentation Handbook,
45RMS/RMSS, Air Force Space Command, Patrick Air Force Base, Fla.
D19203
Cummins, K. L., M. J. Murphy, E. A. Bardo, W. L. Hiscox, R. B. Pyle, and
A. E. Pifer (1998a), A combined TOA/MDF technology upgrade of the
U.S. National Lightning Detection Network, J. Geophys. Res., 103,
9035 – 9044.
Cummins, K. L., E. P. Krider, and M. D. Malone (1998b), The U.S. National Lightning Detection Network2 and applications of cloud-toground lightning data by electric power utilities, IEEE Trans. Electromagn. Comput., 40(4), 465 – 480.
Ezcurra, A., J. Areito, and I. Herrero (2002), Relationships between cloudto-ground lightning and surface rainfall during 1992 – 1996 in the Spanish
Basque Country area, Atmos. Res., 61, 239 – 250.
Goodman, S. J., and D. E. Buechler (1990), Lightning-rainfall relationships,
paper presented at 16th Conference on Severe Local Storms, Am. Meteorol. Soc., Kananaskis Park, Alberta, Canada.
Goodman, S. J., D. E. Buechler, P. D. Wright, and W. D. Rust (1988),
Lightning and precipitation history of a microburst producing storm,
Geophys. Res. Lett., 15, 1185 – 1188.
Gosz, J. R., D. I. Moore, G. A. Shore, H. D. Grover, W. Rison, and C. Rison
(1995), Lightning estimates of precipitation location and quantity on the
Sevilleta LTER, New Mexico, Ecol. Appl., 5, 1141 – 1150.
Gremillion, M. S., and R. E. Orville (1999), Thunderstorm characteristics of
cloud-to-ground lightning at the Kennedy Space Center, Florida: A study
of lightning initiation signatures as indicated by the WSR-88D, Weather
Forecasting, 14, 640 – 649.
Grosh, R. C. (1978), Lightning-precipitation—The life history of isolated
thunderstorms, paper presented at Conference on Cloud Physics and
Atmospheric Electricity, Am. Meteorol. Soc., Issaquah, Wash.
Habib, E., and W. F. Krajewski (2002), Uncertainty analysis of the TRMM
ground-validation radar-rainfall products: Application to the TEFLUN-B
field campaign, J. Appl. Meteorol., 41, 558 – 572.
Habib, E., W. F. Krajewski, and G. J. Ciach (2001a), Estimation of rainfall
interstation correlation, J. Hydrometeorol., 2, 621 – 629.
Habib, E., W. F. Krajewski, and A. Kruger (2001b), Sampling errors of
tipping-bucket rain gauge measurements, J. Hydrol. Eng., 6, 159 – 166.
Hodanish, S., D. Sharp, W. Collins, C. Paxton, and R. E. Orville (1997), A
10-year monthly lightning climatology of Florida: 1986 – 95, Weather
Forecasting, 12, 439 – 448.
Holle, R. L., and S. P. Bennett (1997), Lightning ground flashes associated
with summer 1990 flash floods and stream flow in Tucson, Arizona: An
exploratory study, Mon. Weather Rev., 125, 1526 – 1536.
Holle, R. L., A. I. Watson, R. E. Lopez, D. R. MacGorman, and R. Ortiz
(1994), The life cycle of lightning and severe weather in a 3 – 4 June 1985
PRE-STORM mesoscale convective system, Mon. Weather Rev., 122,
1798 – 1808.
Huff, F. A. (1970), Spatial distribution of rainfall rates, Water Resources
Res., 6, 254 – 260.
Huff, F. A., and W. L. Shipp (1969), Spatial correlations of storm, monthly
and seasonal precipitation, J. Appl. Meteorol., 8, 542 – 550.
Jameson, A. R., M. J. Murphy, and E. P. Krider (1996), Multiple-parameter
radar observations of isolated Florida thunderstorms during the onset of
electrification, J. Appl. Meteorol., 35, 343 – 354.
Joss, J. (1990), Precipitation measurement and hydrology, in Radar in
Meteorology: Battan Memorial and 40th Anniversary Radar Meteorology
Conference, edited by D. Atlas, pp. 577–606, Am. Meteorol. Soc., Boston.
Kane, R. J. (1992), Lightning-rainfall relationships in an isolated midAtlantic thunderstorm, paper presented at 3rd National Heavy Precipitation Workshop, Natl. Weather Serv., Pittsburgh, Penn.
Kempf, N. M., and E. P. Krider (2003), Cloud-to-ground lightning and
surface rainfall during the Great Flood of 1993, Mon. Weather Rev.,
131, 1140 – 1147.
Kinzer, G. (1974), Cloud-to-ground lightning versus radar reflectivity in
Oklahoma thunderstorms, J. Atmos. Sci., 31, 787 – 799.
Krider, E. P., R. C. Noggle, and M. A. Uman (1976), A gated, wideband
magnetic direction finder for lightning return strokes, J. Appl. Meteorol.,
15, 301 – 306.
Krider, E. P., R. C. Noggle, A. E. Pifer, and D. L. Vance (1980), Lightning
direction-finding systems for forest fire detection, Bull. Am. Meteorol.
Soc., 61, 980 – 986.
Laird, M. P., D. A. R. Kristovich, R. M. Rauber, H. T. Ochs III, and L. J.
Miller (1995), The Cape Canaveral sea and river breezes: Kinematic
structure and convective initiation, Mon. Weather Rev., 123, 2942 –
2956.
Lang, T. J., and S. A. Rutledge (2002), Relationships between convective
storm kinematics, precipitation and lightning, Mon. Weather Rev., 130,
2492 – 2506.
Lericos, T. P., H. E. Fuelberg, A. I. Watson, and R. L. Holle (2002), Warm
season lightning distributions over the Florida peninsula as related to
synoptic patterns, Weather Forecasting, 17, 83 – 98.
Lopez, R. E., R. Ortiz, W. D. Otto, and L. Holle (1991), The lightning
activity and precipitation yield of convective systems in central Florida,
14 of 15
D19203
GUNGLE AND KRIDER: LIGHTNING AND RAINFALL
paper presented at 25th Conference on Radar Meteorology, Am. Meteorol.
Soc., Paris, France.
MacGorman, D. R., and W. D. Rust (1998), The Electrical Nature of
Storms, 422 pp., Oxford Univ. Press, New York.
Maier, M. W. (1991), Evaluation of 1990 bias error corrections for the
ESMC cloud-to-ground lightning surveillance system, Rep. CSR-3220003, Eastern Space and Missile Cent., Computer Sciences/Raytheon,
Patrick Air Force Base, Fla.
Maier, M. W., and M. B. Wilson (1996), Accuracy of the NLDN real-time
data service at Cape Canaveral, Florida, paper presented at International
Lightning Detection Conference, Global Atmospherics, Inc., Tucson,
Ariz.
Maier, M. W., A. G. Boulanger, and J. Sariat (1978), Cloud-to-ground
lightning frequency over south Florida, paper presented at Conference
on Cloud Physics and Atmospheric Electricity, Am. Meteorol. Soc.,
Issaquah, Wash.
Mazany, R. A., S. Businger, S. I. Gutman, and W. Roeder (2002), A lightning prediction index that utilizes GPS integrated precipitable water vapor, Weather Forecasting, 17, 1034 – 1047.
McCaul, E. W., Jr., D. E. Buechler, S. Hodanish, and S. J. Goodman (2002),
The Almena, Kansas, tonadic storm of 3 June 1999: A long-lived supercell with very little cloud-to-ground lightning, Mon. Weather Rev., 130,
407 – 415.
Molinie, G., S. Soula, and S. Chauzy (1999), Cloud-to-ground lightning
activity and radar observations of storms in the Pyrenees range area, Q. J. R.
Meteorol. Soc., 125, 3103 – 3122.
Nielsen, K. E., S. J. Goodman, and D. E. Buechler (1990), Cloud-to-ground
lightning and rainfall volumes in mesoscale convective systems, paper
presented at 16th Conference on Severe Local Storms and Conference on
Atmospheric Electricity, Am. Meteorol. Soc., Kananaskis, Canada.
Petersen, W. A., and S. A. Rutledge (1998), On the relationship between
cloud-to-ground lightning and convective rainfall, J. Geophys. Res., 103,
14,025 – 14,040.
Piepgrass, M. V., E. P. Krider, and C. B. Moore (1982), Lightning and
surface rainfall during Florida thunderstorms, J. Geophys. Res., 87,
11,193 – 11,201.
Reap, R. (1994), Analysis and prediction of lightning strike distributions
associated with synoptic map types over Florida, Mon. Weather Rev., 122,
1698 – 1715.
Rutledge, S. A., and D. R. MacGorman (1988), Cloud-to-ground lightning
activity in the 10 – 11 June 1985 mesoscale convective system observed
during the Oklahoma-Kansas PRE_STORM project, Mon. Weather Rev.,
116, 1393 – 1408.
Rutledge, S. A., E. R. Williams, and T. D. Keenan (1992), The Down Under
Doppler and Electricity Experiment (DUNDEE): Overview and preliminary results, Bull. Am. Meteor. Soc., 73, 3 – 16.
Seity, Y., S. Soula, and H. Sauvageot (2001), Lightning and precipitation
relationship in coastal thunderstorms, J. Geophys. Res., 106, 22,801 –
22,816.
Seity, Y., S. Soula, P. Tabary, and G. Scialom (2003), The convective storm
system during IOP 2a of MAP: Cloud-to-ground lightning flash production in relation to dynamics and microphysics, Q. J. R. Meteorol. Soc.,
129, 523 – 542.
Senesi, S., P. Bougeault, J. Chaze, P. Cosentino, and R. Thepenier (1996),
The Vaison-LaRomaine flash flood: Mesoscale analysis and predictability
issues, Weather Forecasting, 11, 417 – 442.
Shackford, C. R. (1958), The relation between precipitation and electrical
activity in New England thunderstorms, paper presented at 7th Weather
Radar Conference, Am. Meteorol. Soc., Miami Beach, Fla.
Shackford, C. R. (1960), Radar indications of a precipitation-lightning
relationship in New England thunderstorms, J. Meteorol., 17, 15 – 19.
D19203
Sheridan, S. C., J. F. Griffiths, and R. E. Orville (1997), Warm season
cloud-to-ground lightning – precipitation relationships in the south-central
United States, Weather Forecasting, 12, 449 – 458.
Soriano, L. R., F. de Pablo, and E. Garcia Diez (2001), Relationship between convective precipitation and cloud-to-ground lightning in the Iberian Peninsula, Mon. Weather Rev., 129, 2998 – 3003.
Soula, S., and S. Chauzy (2001), Some aspects of the correlation between lightning and rain activities in thunderstorms, Atmos. Res., 56,
355 – 373.
Soula, S., H. Sauvageot, G. Moline, F. Mesnard, and S. Chauzy (1998), The
CG lightning activity of a storm causing a flash-flood, Geophys. Res.
Lett., 25, 1181 – 1184.
Tapia, A., J. A. Smith, and M. Dixon (1998), Estimation of convective
rainfall from lightning observations, J. Appl. Meteorol., 37, 1497 –
1509.
Toracinta, E. R., D. J. Cecil, E. J. Zipser, and S. W. Nesbitt (2002), Radar,
passive microwave, and lightning characteristics of precipitating systems
in the tropics, Mon. Weather Rev., 130, 802 – 824.
Watson, A. I., R. E. Lopez, R. L. Holle, and J. R. Daugherty (1987), The
relationship of lightning to surface convergence at Kennedy Space Center, Weather Forecasting, 2, 140 – 157.
Watson, A. I., R. L. Holle, R. E. Lopez, R. Ortiz, and J. R. Nicholson
(1991), Surface wind convergence as a short-term predictor of cloudto-ground lightning at Kennedy Space Center, Weather Forecasting, 6,
49 – 64.
Watson, A. I., R. E. Lopez, and R. L. Holle (1994), Diurnal cloud-to-ground
lightning patterns in Arizona during the southwest monsoon, Mon.
Weather Rev., 122, 1716 – 1725.
Weather Bureau (1947), The reliability of areal rainfall determination, in
Hydrometeorological Rep. 5: Thunderstorm Rainfall, pp. 234 – 259,
Waterways Experiment Station, Vicksburg, Miss.
Williams, E. R., M. E. Weber, and R. E. Orville (1989), The relationship
between lightning type and convective state of thunderclouds, J. Geophys. Res., 94, 13,213 – 13,220.
Williams, E. R., S. A. Rutledge, S. G. Geotis, N. Renno, E. Rasmussen, and
T. Rickenbach (1992), A radar and electrical study of tropical ‘‘hot
towers’’, J. Atmos. Sci., 49, 1386 – 1395.
Williams, E. R., et al. (2002), Contrasting convective regimes over the
Amazon: Implications for cloud electrification, J. Geophys. Res.,
107(D20), 8082, doi:10.1029/2001JD000380.
Wilson, J. W., and D. L. Megenhardt (1997), Thunderstorm initiation,
organization, and lifetime associated with Florida boundary layer convergence lines, Mon. Weather Rev., 125, 1507 – 1525.
Woodley, W. L., A. R. Olsen, A. Herndon, and V. Wiggert (1975), Comparison of gage and radar methods of convective rain measurement,
J. Appl. Meteor., 14, 909 – 928.
Workman, E. J., and S. E. Reynolds (1949), Electrical activity as related to
thunderstorm cell growth, Bull. Am. Meteorol. Soc., 30, 142 – 144.
Zhou, Y., X. Qie, and S. Soula (2002), A study of the relationship between
cloud-to-ground lightning and precipitation in the convective weather
system in China, Ann. Geophys., 20, 107 – 113.
Zipser, E. J. (1994), Deep cumulonimbus cloud systems in the tropics with
and without lightning, Mon. Weather Rev., 122, 1837 – 1851.
B. Gungle, U. S. Geological Survey, Water Resources Discipline, Tucson,
AZ 85719, USA.
E. P. Krider, Institute of Atmospheric Physics, University of Arizona,
Tucson, AZ 85721-0081, USA. (krider@atmo.arizona.edu)
15 of 15