Transcript Correlation coefficient to flash rate in precipitation features

What kind of clouds have
lightning?
Observing storms from space
Why there is a relationship between radar reflectivity and lightning flash?
Riming electric charge separation
Keys :
Graupel
Temperature
Super-cooled liquid water
Takahashi and Miyawaki, 2002, JAS
Define precipitation features using TRMM
Precipitation Radar
PR echo top height (km)
Radar Precipitation Features (RPFs)
Contiguous area with rainfall
Properties of precipitation features
Flash rate
Temperature at 20, 30, 40 dBZ
echo tops
Maximum reflectivity at temperatures
Area of 20, 30, 40 dBZ at
temperatures
Volume of 20, 30 40 dBZ at -5oC - 35oC
IWC of 20, 30 40 dBZ at -5oC - -35oC
Population of RPFs and those with lightning
• 1998-2010
Land
Coast
Open ocean
All 36S-36N
RPF
3.7 million
6.4m million
11.8 million
22.2 million
With flash
11.5%
2.6%
0.5%
3%
2-D histogram of RPFs and probability with lightnin
Temperature of 20 dBZ echo top vs. Temperature of 30 dBZ echo top
Regional variation of lightning probability in precipitation systems
Correlation coefficient to flash rate in precipitation features
maximum reflectivity at altitudes vs. Flash rate
Correlation coefficient to flash rate in precipitation features
Echo top temperature vs. Flash rate
Ocean
Coast
Land
All 36S-36N
Maximum 20 dBZ echo top
Temperature
-0.12
-0.18
-0.27
-0.24
Maximum 30 dBZ echo top
Temperature
-0.30
-0.36
-0.40
-0.38
Maximum 40 dBZ echo top
Temperature
-0.49
-0.50
-0.49
-0.50
Maximum reflectivity does not have good correlation
with Flash rate neither ( < 0.5 )
Correlation between radar reflectivity vs. flash rate in precipitation features
Correlation coefficient to flash rate in precipitation features
area of reflectivity at temperatures vs. Flash rate
Solid : land
Dotted: ocean
area of reflectivity at temperatures vs. Flash rate
auto-correlation among different part of storms removed
Solid : land
Dotted: ocean
Regional variation of correlations (I)
Correlation coefficients
Regional variation of correlations (II)
Slope of linear relationship
Summary
•
Flash rate is best correlated with the volume of the high reflectivity (>
30 dBZ) in the mixed phase region, confirming the importance of the
presence of large particles in the charging process.
•
More flashes are generated over land than over ocean given the
same volume of 30-35 dBZ in the mixed-phase region.
• There are large regional differences in the correlations between radar
reflectivity properties vs. flash rate.
• Even with the similar radar reflectivity profiles to the oceanic systems,
it is still easier to have lightning flashes over Amazon. Other factors must
play important roles besides the graupel at mixed phase region in the
lightning generation.
Relative Contributions of Electrified Shower
Clouds and Thunderstorms to the Global Circuit:
Can 10 Years of TRMM Data Help Solve an Old
Puzzle?
Fair Weather Charge
• In fair weather there is a natural separation
of charge in the atmosphere.
+
-
+
-
+
-
+
-
+
-
Upper troposphere is
positively charged.
Ground is negatively
charged.
The atmosphere normally has a
voltage gradient of 100 volts/meter…
… which may
sound like a
lot, but what
happens when
you stand one
meter from a
110 volt outlet
?
Old Puzzle / Hypothesis
• 90 years ago
“A thundercloud or shower-cloud is the seat of the
electromotive force which must cause a current to
flow through the cloud between the Earth’s surface
and the upper atmosphere... In shower-clouds in
which the potentials fall short of what is required to
produce lightning discharges, there is no reason to
suppose that the vertical currents are of an
altogether different order of magnitude.”
------ Wilson (1920)
Carnegie Curve vs. thunder days
80 years ago
Diurnal variation of electric field seems reproduced by the thunder days
(after add an arbitrary uniform oceanic storms to bring the amplitude down)
Thunderstorms observed by TRMM
Americas
Africa
Asian
Thunderstorms over different regions have different lightning flash rates
Carnegie curve vs. flash count
Good correlation in phase, but much higher amplitude
Same as pointed out by Bailey et al. 2007
Carnegie curve vs. rainfall in
thunderstorms
Good correlation in both phase and amplitude
Diurnal variation of flashes
Asian
Africa
Americas
Diurnal variation of rainfall in
thunderstorms
Asian
Africa
Americas
Vostok electric field vs. rainfall in thunderstorms
DJF Vostok electric field
might be contaminated
by weathers
Good correlation between
Thunderstorm rain vs. electric field
in different seasons
End of story?
• It seems that the diurnal variation of rainfall in
thunderstorms has a very good correlation
with the Carnegie curve both in the phase and
the amplitude.
• However, this is not the end of the story. What
about the shower clouds without lightning as
mentioned by Wilson 90 years ago?
ER-2 overflight of Emily
In-situ storm current observations
Storm current from 850 ER-2 overflights
Courtesy of Mach et al. 2010
A VERY rough way to identify the
electrified shower clouds
30 dBZ echo top colder than -10oC over land, -17oC over ocean
Rainfall from thunderstorms,
electrified show clouds and others
Carnegie vs. electrified shower
clouds
Thunderstorm + electrified
shower?
A different approach
define convective cells
Three “cell” definitions:
•Convective pixels
(red color fill)
•40 dBZ pixels in any
levels (black line)
•30 dBZ pixels at 6 km
(white line)
Global distribution of convective cells
(Solid line) Rain in cells defined by 30 dBZ at 6 km has the best correlation
Electric field vs. Rainfall from convective cells defined by area
of 30 dBZ at 6km
Summary of the Carnegie curve
• With more observations available today, we have better tools
to play the same game as Whipple (1929) played 80 years ago:
relating the diurnal variation of thunderstorms to that of the
electric field.
• Diurnal variation of rainfall from thunderstorms has a good
correlation with the Carnegie curve both in the phase and the
amplitude.
• The role of electrified shower clouds is still hard to describe
due to the difficulty of identifying them and quantifying the
electric field that they contribute.
• A different approach of adding convective cells gives more
hints to this puzzle.
Let’s talk about the final project
• Written reports have to be submitted to me in e-mail before
end of Tuesday Feb 26. Late report is not accepted. If you
insist to give me a paper report, you have to hand the report
to me on Monday Dec 25.
• We have 8 students
Each one have about 10 minutes of presentation and a few
minutes for questions.
So we need 2 class time. We do a lottery on who present first.
If you have presentation on the day, come to classroom 5
minutes early to upload your slides onto my laptop.