MICROWAVE RAINFALL RETRIEVALS AND VALIDATIONS

1
MICROWAVE RAINFALL RETRIEVALS AND VALIDATIONS

• R.M. GAIROLA, S. POHREL A.K. VARMA
• OSD/MOG
• SAC/ISRO AHMEDABAD

2
ISSUES AND OBJECTIVES

• Development of algorithms for rain rate
  retrievals over land and oceans
• Validation fine-tuning of algorithm using
  ground based measurements like Doppler Weather
  Radar (DWR) rain gauges etc. over Indian
  region (land ocean)
• Study of some cases of monsoon, flash floods
  cyclone rainfall using the developed algorithms
• Study of hydrological processes to assess the
  accuracy of the retrievals

3
(No Transcript)
4
(No Transcript)
5
Physical Approach Forward Modeling (Radiative
Transfer Simulations)
6
(No Transcript)
7
For Non-Raining Atmospheres
Water vapour versus simulated TBs - TMI Channels
8
CLW versus Simulated TBs TMI Channels
9
(No Transcript)
10
RTR for Raining Atmospheres Eddington
approximation
Assumption- For an isotropic field the ratio of
the second moment of the radiation field to the
mean intensity is everywhere equal to 1/3.
Model used- Kummerow et al., 1993
Discrete Ordinate Approximation
Assumption- Cross-polarization is negligible and
the scattering phase function follow
Henyey-Greenstein equation.
Model used- Liu, 1998
11
(No Transcript)
12
Observations from TRMM-TMI and PR (V-Blue, H-Red)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22

• Examples from Data
• Tropical Rainfall Measuring Mission (TRMM)
• TRMM Microwave Imager (TMI) 85.5 GHz
• Precipitation Radar (PR) 13.8 GHz
• Data Product (Version 6)
• 2A25 for Radar Reflectivity
• 1B11 for Brightness Temperature
• 2A12 for Latent Heat information

23
300 275 250 225 200
Spatial distribution of BT (85.5 GHz, Vertical)
over Eastern India
24
50 30 20 10 0
Spatial distribution of Near Surface PR
Reflectivity (dBZ) over Eastern India
25
(a)
(b)
Number of Pixels

• Spatial variation of BT(85.5 GHz) for convective
  and stratiform region
• and (b) Corresponding vertical cross section of
  PR Reflectivity (dBZ)

26
RETRIEVAL ALGORITHMS FOR OCEAN LAND (MR
Approach) OCEANIC RAINFALL Use of emission (at
low frequency) scattering (at high frequency)
signatures with suitable statistical model R
a0a1.ln(Tb19vTb19h)a2.ln(Tb22v
180)a3.ln(Tb85v-Tb19v) LAND RAINFALL Use of
scattering signatures at high frequencies R
c0 c1.PCT c2.SI PCT1.818TB85v0.818TB8
5h , SIE85v(10v,19v,21v) TB85v E85v(10v,
19v, 21v) b0b1.Tb10vb2.Tb19v b3.TB21v (bi
coeff. derived for non-raining conditions)
27
TRMM-TMI Rainfall from NASA Present Algorithms
NASA TMI Rain (mm/hr)
Present TMI Rain (mm/hr)
AUGUST 2, 2002
28
COMPARISON OF RAINFALL FROM NASA, PR PRESENT
ALGORITHMS
NASA PR RAINRATE
NASA TMI RAINRATE
OCTOBER 10, 2002
TMI RAINRATE (PRESENT ALGORITHM)
29
TRMM-TMI Rainfall from NASA-GPROF and SAC
Algorithm, Aug 2, 2002
30
Output Layer
----- Training ----- Testing
Hidden Layer
Input Layer
Fig. ANN architecture having four inputs as
brightness Temp and one output as radar rainfall
Fig. 3. ANN error distribution for training
31
RETRIEVAL ALGORITHMS FOR OCEAN LAND (ANN
Approach)
Observed Rain Rate
ANN Rain Rate
32
ANN Rain Rate
Observed Rain Rate
Fig. Observed and ANN rain rate for the
geographical area used for testing the
performance of the ANN.
33
VALIDATIONS
34
GROUND-BASED OBSERVATIONS BY DOPPLER WEATHER
RADAR (CHENNAI)
Original Image
35
COMPARISON OF TRMM RAINFALL USING DWR DATA
OCTOBER 17, 2002
NASA-GPROF RAINFALL PRODUCT
36
(No Transcript)
37
TRMM Rain (mm/h)
SHAR-DWR Rain (mm/h) Ele 0.5
1929Z
1845Z
SHAR-DWR Rain (mm/h) Ele 1.5

• Statistics (17-Oct-02)
• With 0.5 degree elevation
• r 0.67
• Total Points 256
• rms diff 4.592406
• 2. With 1.5 degree elevation
• r 0.63
• Total Points 256
• rms diff 4.725613594

38
17-10-2002, 1500hrs MetSAT-5
WV Image
IR Image
39
Convective Activity on 1-6th Dec 2002
40
SHAR-DWR Rain (mm/h) Ele 0.5
TRMM Rain (mm/h)
0915Z
1015Z
SHAR-DWR Rain (mm/h) Ele 1.5

• Statistics (6-Dec-02)
• With 0.5 degree elevation
• r 0.45
• Total Points 256
• rms diff 3.13
• 2. With 1.5 degree elevation
• r 0.42
• Total Points 256
• rms diff 3.22

41
VALIDATION CAMPAIGN (Oct-Nov. 2003)
42
SHAR DWR and Disdrometer Observations
Simultaneous DWR and Disdrometer observations
over SHAR on 17-10-03 (0951 IST)
43
PPI Plot (06 Nov,2003) DWR (SHAR)
44
PR Scan (06 Nov, 2003) Orbit No. 34082
45
(a) DWR (b) PR
Magnified View of near simultaneous Observations
of (a) DWR and (b) PR (TRMM) Reflectivity on 06
Nov, 2003
46
(a)
(b)
Scatter plots (a) PR estimated dBZ versus DWR
estimated dBZ (b) PR estimated Rainfall
intensity versus DWR estimated Rainfall
intensity
47
6 Nov. 2003
DWR-4 km
PR
DWR-25 km
TMI
48
Applications Some Examples
49
TRMM Rainfall Anomalies 1998-2004
50
MODIS Water Vapor Obs. 2001-2004
51
TRMM Seasonal Rainfall (2001-2004)
52
TRMM Rainfall Rates (mm/h) 1998-2004
53
(No Transcript)
54
(No Transcript)
55
(No Transcript)
56
Time Latitude Plot of TPW from MODIS 750-850
57
Time Longitude Plot of TPW from MODIS 50-150
58

• Hydrological Studies from Microwave
  Measurements
• Evaporation Estimates from Microwave Measurements
  examples from TRMM
• Precipitation Estimates from Microwave
  Measurements examples from TRMM
• Fresh water fluxes (E-P) - examples from TRMM

59
E-P for July 1, 2003 (mm/day)
60
E-P for July, 2003 (mm/month)
61

• Conclusions and Concerned Areas
• Beam Filling Problem
• Horizontal and vertical inhomogenity
• Drop size distribution
• Melting layer
• Inversion Techniques

62
THANKS