Cloud Microphysical Properties, Processes, And

1
Cloud Microphysical Properties, Processes, And
Rainfall Estimation Opportunities Daniel
Rosenfeld Institute of Earth Sciences The Hebrew
University of Jerusalem Carlton W. Ulbrich
Department of Physics and Astronomy Clemson
University Clemson, SC, USA
Presented at the 2003 AMS Radar Meteorology
Conference
2
In the beginning (1947) Marshall and Palmer
created the MP Drop Size Distribution And MP
Z-R relationship
3
And we said that the ZR was good, and created
more
4
and more
and more
5
Early classification of Z-R by rainfall
types Fujiwara, 1965
Thunderstorms
Z ARb
A
Rain Showers
Continuous Rain
b
6
Early classification of Z-R by geographical
variation
Stout and Muller, 1968
A and b of Z-R relations with geographical
variation. Muller and Simms, 1966
7
However, lack of physically based classification
created a tower of Babel of Z-R relations
The building of the tower of Babel by Pieter
Bruegel, 1563 Oil on oak panel, Kunsthistorisches
Museum Wien, Vienna
8
Forms of the raindrop size distribution (RDSD)
?Exponential N(D)N0exp(-3.67D/D0) Marshall-Pa
lmer (1948), Laws-Parsons (1943), Best (1950)
? Log-Normal N(D)N0D-1exp-cln2(D/Dg) Levin
(1954) ? Gamma N(D)N0D?exp-(3.67?)D/D0 De
irmendjian (1969), Willis (1984), Ulbrich (1983)
Interdependence of RDSD parameters In
all of the above the RDSD parameters are NOT
independent. Attempts to resolve this problem.
? Normalized forms do not remove
interdependence Willis (1984), Testud et al.
(2001), Chandrasekar and Bringi (1987) ?
New RDSD parameters Haddad et al. (1996)
attempt to establish independent
parameters. DDmR-0.155 ssmDm-0.2R0.031exp(
0.017R0.74) ? Gamma distribution parameters
N0 and ? strong correlation. Chandrasekar and
Bringi (1987) In this work the Gamma
distribution is used together with Z-R relations
to determine RDSD parameters and integral
rainfall parameters defined in terms of the RDSD
parameters (in spite of lack of independence).
9
Rain Parameter Diagram of Atlas Chmela (1957)
Exponential RDSD. Isopleths of W, Do, No
10
Atlas-Chmela RAPAD with sample Z-R relations ?
Equilibrium raindrop size distribution
11
Method of using the Gamma distribution with the
Z-R law parameters. Any two integral rainfall
parameters P and Q Eliminating D0
yields Inverting
gives With PZ and QR
in ZARb, can find ? and N0 from A and b. These
can then be used to find coefficients and
exponents in theoretical D0-R, NT-R, W-R
relations.
12
Z-R relations
used in this work Source
No. Z-Rs Description Foote (1966)
2 Mountain thunderstorm in AZ,
Filter paper meas. Sims (1964)
1 Thundershowers. Illinois. Joss and
Waldvogel (1970) 1 Thunderstorm. 25
days total, Locarno, disdrometer) Petrocchi and
Banis (1980) 1 Thunderstorm, Norman, OK
(disdrometer) Sauvageot (1994) 1
Tropical squall line. Congo Ulbrich et al.
(1999) 1 Ave. of 7 thunderstorms in
Arecibo, PR.(disdrometer) Tokay et al. (1995)
2 Tropical. Disdrometer data. Darwin,
Australia. Maki et al. (2001) 6
15 squall lines. Disdrometer, Darwin,
Australia. Tokay and Short (1996) 2
TOGA COARE data. Kapingamarangi atoll,
disdrometer Stout and Mueller (1968) 3
Trade wind showers Marshall Islands Jorgensen
and Willis (1982) 2 Hurricane Frederic,
9/12/79 Ulbrich and Atlas (2001) 2
Tropical storms. PMM analysis, TOGA COARE 2DP
data. Fujiwara and Yanase (1968) 3
Filter paper meas. Orographic, slopes of Mt.
Fuji. Atlas et al. (1999) 2 TOGA
COARE, Kapingamarangi atoll, disdrometer. Blanchar
d (1953) 1 Orographic rain,
windward slopes on Mauna Loa. Only those
relations used which involve direct measurements
of drop size spectra and for which the
environmental conditions are well known. No
combinations of instruments (radar-raingages or
radar-disdrometer, etc.) were considered.
13
Rain water content, W g m-3 as a function of
rain drop median mass diameter D0 cm and drop
concentration NT m-3 for R10 and R30 mm h-1,
For all Z-Rs
14
Processes determining the Rain Drop Size
Distribution Wilson and Brandes, 1979
15
Processes determining the Rain Drop Size
Distribution
Coalescence
Modification of the DSD by coalescence alone
decreases the numbers of small diameter drops and
increases those of the larger drops.
Consequently D0 must increase and the total
number concentration of drops NT must decrease.
The process probably also increases ?, but not by
much. The result is a decrease in N0 and a
consequent increase in A and small decrease in b.
Because of the small change in ?, we may
therefore use the same RAPAD before and after
modification. The result is an approximate
parallel shift in the Z-R relation on the RAPAD
upward and to the left, perpendicular to the D0
isopleths.
16
Processes determining the Rain Drop Size
Distribution
Breakup
Modification of the DSD by break-up alone
increases the numbers of small diameter drops and
decreases the numbers of large diameter drops.
There must be a consequent decrease in D0 and an
increase in NT. Accordingly, N0 must increase.
There is probably a small change in ? with a
tendency toward a decrease. The end result is a
decrease in A and a small increase (or perhaps no
change) in b. Because of the small change in ?
we may use the same RAPAD before and after
modification. There is therefore an approximate
shift in the Z-R relation on the RAPAD downward
and to the right, approximately parallel to the
D0 isopleths.
17
Processes determining the Rain Drop Size
Distribution
Coalescence And Break-up Combined
Break-up is more important at the larger sizes,
coalescence more important at small sizes
(insofar as numbers are concerned). Both
processes acting together increase ?
substantially. This requires using a different
RAPAD before and after modification, i.e., the
isopleths will shift. The apparent increase in ?
will decrease b. What will happen to A depends
on which of the two processes in predominant.
18
Processes determining the Rain Drop Size
Distribution
Accretion
Since accretion of cloud particles by raindrops
acts to increase the sizes of all particles
without increasing their numbers, then NT must
remain unchanged. There is probably a shift of
the DSD parallel to itself to larger diameters
with a consequent increase in D0. Since NT
remains constant N0 must decrease. The result is
an increase in A and probably little change in b.
Since ? is unchanged the same RAPAD before and
after modification can be used. Therefore, the
Z-R law will be shifted parallel to itself upward
on the RAPAD.
19
Processes determining the Rain Drop Size
Distribution
Evaporation
The presence of evaporation will result in a
greater loss of the numbers of small diameter
particles than large drops. Consequently, NT is
not constant and must decrease. There must also
be a substantial change in the shape of the DSD
so that ? increases. Also, D0 must increase.
The result is a decrease in N0 and an increase in
A. In addition, since ? increases, b must
decrease. Since there is a change in ? it is
necessary to use a different RAPAD before and
after modification.
20
Processes determining the Rain Drop Size
Distribution
Updraft
The presence of an updraft eliminates the
smallest particles from the DSD at the lower
levels. The effect on the DSD is therefore the
same as evaporation. .
21
Processes determining the Rain Drop Size
Distribution
Downdraft
Here we assume that the downdraft is
accelerating downward. There will therefore be
an increased flux of small particles thereby
decreasing ? which means b must increase. Also,
NT must increase and D0 decrease. Hence N0 must
increase which results in a decrease in A.
Because of the change in ? it is necessary to use
a different RAPAD before and after modification.
22
Processes determining the Rain Drop Size
Distribution
Size-sorting
Size sorting tends to make the DSD much narrower
which means ? must increase substantially.
Therefore b must decrease. NT must obviously
decrease, but what happens to D0 depends on which
segment of the precipitation streamer is being
observed. So A would either increase or decrease
depending on what happens to D0. It is likely
that the decrease in NT will dominate the change
in D0 so that A would increase, but this is not
certain. Because of the dramatic change in ? a
different RAPAD would definitely have to be used
before and after modification.
23
Processes determining the Rain Drop Size
Distribution
Equilibrium DSD Z 600 R D0 1.75 mm
Hu and Srivastava (1995)
24
Impact of Cloud DSD on the evolution of Rain DSD

• Maritime Cloud DSD
• Cloud drop coalescence?Drizzle
• Drizzle coalescence ?Raindrops
• More coalescence ?larger raindrops
• ?breakup and equilibrium DSD
• Approaching D0e from below

• Continental Cloud DSD
• Cloud drop accretion?graupel
• ? hail?large raindrops
• ?breakup ?
• equilibrium DSD
• Approaching D0e from above

25
Equilibrium
Continental
Maritime
Trends of D0 for convective maritime and
continental clouds. Rain water content, W g m-3
as a function of rain drop median mass diameter
D0 cm and drop concentration NT m-3 for R30
mm h-1, For all Z-Rs.
26
The classification scheme of convective clouds
into microphysical zones according to the shape
of the temperature effective radius relations
Note that in extremely continental clouds re at
cloud base is very small, the coalescence zone
vanishes, mixed phase zone starts at Tthe glaciation can occur at the most extreme
situation at the height of homogeneous freezing
temperature of 39oC. In contrast, maritime
clouds start with large re at their base,
crossing the precipitation threshold of 14 mm
short distance above the base. The deep rainout
zone is indicative of fully developed warm rain
processes in the maritime clouds. The large
droplets freeze at relatively high temperatures,
resulting in a shallow mixed phase zone and a
glaciation temperature reached near 10oC
Rosenfeld and Lensky, 1998
27
Why is Continental - Maritime classification so
fundamental?
Annual average lightning density flashes
km-2 Lightning prevail mostly over land, whereas
rainfall is similar over land and ocean,
indicates fundamental differences between
continental and maritime rainfall.
28
Differences in rain DSD forming processes
between Maritime and Continental Clouds In
Maritime clouds there are More Coalescence
?Rainout ?D0 ? Smaller D0 ? Smaller R(Z) Less evaporation ?
Smaller D0 ? Smaller R(Z) In Continental
clouds there are Less Coalescence ?No Drizzle ?
No small rain drops ? Hydrometeors start as
graupel and hail ? D0 D0e ? Larger R(Z) Larger
updrafts ? Larger D0 ? Larger R(Z) More
evaporation ? Larger D0 ? Larger R(Z) .
29
Disdrometer measured DSD of continental and
maritime rainfall, as micophysically classified
by VIRS overpass. The DSD is averaged for the
rainfall during - 18 hours of the overpass
time. The disdrometers are in Florida (Teflun B),
Amazon (LBA), India (Madras) and
Kwajalein. Application of TRMM Z-R shows a near
unity bias in maritime clouds, but overestimates
by a factor of 2 rainfall from continental
clouds.
Rosenfeld and Tokay, 2002
30
Rain Obs in Asia for PR Algorithm Tuning
Study items DSD, BB model, BB RR comp. with
PR. One of the Major Findings DSD properties
Significant seasonal dependence in India Large
D0 in SW monsoon, low D0 in NE. Not much
seasonal dependence in Singapore.
By T. Kozu
Kozu
31
(No Transcript)
32
Z-R relations for rainfall from maritime and
continental convective clouds. The rain
intensities for 40 and 50 dBZ are plotted in the
figure. Note the systematic increase of R for a
given Z for the transition from continental to
maritime clouds.
1. Arizona mountain thunderstorms (Foote
1966) 646 1.46 3. Swiss Locarno
thunderstorms, continental (Joss and
Waldvogel , 1970) 830 1.50 4. Illinois
thunderstorms, continental (Sims,
1964) 446 1.43 5. Oklahoma thunderstorms,
moderate continental (Petrocchi and Banis,
1980) 316 1.36 6. Congo Squall line. Tropical
continental (Sauvageot, 1994) 425 1.29 7.
PurtoRico thunderstorms. Coastal, moderate
maritime (Ulbrich et al., 1999). 261 1.43 8.
Darwin Squalls. Coastal, tropical maritime
(Maki et al., 2001) 232 1.38 9. Darwin
Convective DSD. Coastal, tropical maritime
(Tokay et al., 1995) 175 1.37 10. COARE
Convective DSD. Equatorial maritime (Tokay
and Short, 1996). 139 1.43 11. Marshall Trade
wind cumulus. Warm rain maritime (Stout and
Mueller, 1968) 126 1.47 12. Marshall Showers.
Equatorial maritime. (Stout and Mueller,
1968) 146 1.42 E. Equilibrium DSD. 600 1.00
33
The variation of D0 with the liquid water content
(W) and total drop concentration (NT) for R10
and 30 mm hr-1 of convective rainfall in maritime
and continental regimes. Note that D0 decreases
from continental to maritime clouds.
34
Convective Stratiform classification
In Maritime Convective clouds More Coalescence
?Rainout ?D0 evaporation ? Smaller D0 ? Smaller R(Z) In
Stratiform clouds Aggregation of ice ? Larger
D0 ? Larger R(Z) More evaporation ? Larger D0 ?
Larger R(Z) .
35
Tropical Convective Stratiform received much
attention because of different vertical profiles
of latent heating, to which the global
circulation is quite sensitive. Convective R(Z)
almost double than Stratiform R(Z). Tokay and
Short (1995, 1996) Atlas et al. (2000) Ulbrich
and Atlas (2001) C/S classification already
applied to TRMM Z-R
36
Z-R relations for rainfall from tropical
convective and stratiform clouds. The rain
intensities for 30 and 40 dBZ are plotted in the
figure. Note the systematic decrease of R for a
given Z for the transition from convective to
stratiform clouds.
1. Darwin Convective DSD. Coastal, tropical
maritime (Tokay et al., 1995) 175 1.37 a.
Darwin Stratiform DSD. Coastal, tropical maritime
(Tokay et al., 1995) 335 1.37 2. Marshall
Showers. Equatorial maritime. (Stout and
Mueller, 1968) 146 1.42 b. Marshall continuous.
Equatorial maritime. (Stout and Mueller,
1968) 226 1.46 3. COARE Convective DSD.
Equatorial maritime (Tokay and Short,
1996). 139 1.43 c. COARE Stratiform DSD.
Equatorial maritime (Tokay and Short,
1996). 367 1.30 4. COARE Convective aloft, by
updraft (Ulbrich and Atlas, 2001) 120 1.43 d.
COARE Stratiform aloft, by updraft (Ulbrich
and Atlas, 2001) 203 1.46 5. Hurricane eye wall.
(Jorgensen and Willis, 1982) 287 1.27 e.
Hurricane rain bands. (Jorgensen and Willis,
1982) 301 1.38 E. Equilibrium DSD. 600 1.00
37
The variation of D0 with the liquid water content
(W) and total drop concentration (NT) for R10
and R30 mm hr-1 of convective rainfall in
maritime and continental regimes. Note that D0
increases from convective to stratiform
clouds. Note that Hurricane has near equilibrium
D0
38
Z-R of Orographic Rainfall Enhancement
39
In orographic clouds with active
coalescence Early fall to the mountain slope of
small hydrometeors ? Highly immature DSD ?D0 D0e ? Much Smaller R(Z) .
40
Z-R relations for rainfall from maritime
orographic clouds. The rain intensities for 30
and 40 dBZ are plotted in the figure. Note the
systematic increase of R for a given Z for
greater orographic uplifting.
1. Mount Fuji, at height of 1300 m. (Fujiwara
and Yanase, 1968) 240 1.48 2. Mount Fuji, at
height of 2100 m. (Fujiwara and Yanase,
1968) 88 1.28 3. Mount Fuji, at height of 3400 m.
(Fujiwara and Yanase, 1968) 48 1.11 a.
Marshall Trade wind cumulus, 0 m. (Stout and
Mueller, 1968) 126 1.47 b. Hawaii, Mauna Loa, 800
m. (Blanchard., 1953) 31 1.71 E.
Equilibrium DSD. 600 1.00
41
The variation of D0 with the liquid water content
(W) and total drop concentration (NT) for R10
and R30 mm hr-1 of orographic rainfall in
maritime clouds. Note that D0 decreases
substantially with the orographic uplifting.
42
Summary All Z-R classifications combined
The variation of D0 with the liquid water content
(W) and total drop concentration (NT) for R10
and 30 mm hr-1 of convective rainfall in maritime
and continental regimes.
43
Summary
A review of Z-R relations based on cloud physics
RDSD forming processes revealed that Z-R behave
systematically, Producing larger R for the same Z
when going from Continental ? Maritime
(X4) Maritime ? Orographic (X4) Stratiform ?
Convective Maritime (X2) Opportunities Classific
ation criteria can be detected by Satellite
(cloud drop effective radius for
continentality) Radar 3-D structure Dynamics of
orographic lifting. Potential for dynamic Z-R for
space and ground based radars, accounting for
systematic biases by factors of 2 to 4.