Study and Comparison of Simulation of
Satellite Microwave Observations in
Cloudy and Rainy Areas using RTTOV
and CRTM
HAN Wei* and DONG Peiming**
hanwei@cma.gov.cn
*
* *Chinese
CAMS/CMA
Numerical Prediction Center, CMA
Academy of Meteorological Sciences,CMA
Outline
 Background
 Clear Radiance assimilation in GRAPES
 Cloudy Radiance assimilation: onging Research
 Comparisons of RTTOV and CRTM for Cloudy Radiance
 Input from 24h forecast
 RTTOV: cloud frac.,water,ice,rain,snow
 CRTM:effective radius,water,ice,rain,snow,graupel,hail
 Jacobians
 Discussion
 Training of the fast RT model for cloudy radiance
 L2 norm or H1 norm
CAMS/CMA
 Control variables choices
Background: Road map of GRAPES
1999-2000
Framework Design
2001.-2005
2006.7
GRAPES_SDM
sand storm
GRAPES_RUC
CAMS/CMA


Other Apps.:
GRAPES_TCM
typhoon track
Finish development
GRAPES dynamical core
Implement WRF physics 
GRAPES_Meso
GRAPES_VAR
2007.7-present
GRAPES_Meso
operation
Regional 4-DVAR
development
Setup of GRAPES_GFS
& pre-operation
 Global 4-DVAR
development

Satellite data assimilation in GRAPES
 Observation Operator for Radiance Assimilation
 RTTOV: RTTOV6->RTTOV7->RTTOV9
 CRTM: CRTM1.2->CRTM2.1
 Satellite Observations in GRAPES
 ATOVS(NOAA,METOP,FY3)
 GPS Reflectivity (COSMIC)
 IASI and AIRS
 Bias correction
 Harris and Kelly(2001)
 VarBC
 Constrained VarBC :considering of mobel bias
CAMS/CMA
 Observation Error tuning
Background
 GRAPES_VAR
 RT model: RTTOV(9.3) and CRTM(1.2)
 Clear radiance assimilation in preoperational mode
 Cloudy and rain affected radiance: ongoing research
More than 70% of data are rejected due
to cloud and rain
CAMS/CMA
Impact of AMSUs and AMVs in GRAPES
Baseline+AMVs: positive
Control+AMV_HA: positive
Control+AMSU_ch4: positive
Resolution:1
CAMS/CMA degree,33Level
Baseline:Sonde+Airep+Synop+ships+COSMIC
Control: Baseline+AMVs+AMSUs
Verification Hours
The impact of water content: CRTM
CAMS/CMA
Water contents
No water contents
Cloud Scattering and Emission in CRTM vs. RTTOV
Radiative Transfer Solver
Scattering Properties
(look-up table approach)
CRTM
RTTOV (RTTOV-SCATT)
Advanced Adding and Doubling
Delta-Eddington
(ADA) scheme
Approximation
Precalculated using Mie theory
Precalculated using Mie
and tabulated as a function of
theory and tabulated as a
frequency, temperature, radii,
function of frequency,
and hydrometeor type
temperature, and
anddensity
hydrometeor type and
density.
Cloud types
Water, ice, rain, snow, graupel
Water, ice, rain, and snow
and hail
Cloud cover
CAMS/CMA
Not handle yet
Cloud fraction profile
From Yong Chen et al,2011
Comparison of Input for hydrometeors
 RTTOV
 cloud liquid water, cloud ice water, rain flux and solid
RH − RH 0 b
precip. flux
N =(
)
1.0 − RH 0
 effective cloud fraction
RTTOV-SCATT
 CRTM
 water, ice, rain, snow, graupel and hail
 effective radius: const.
∞
re =
∫ n( r ) r
0
∞
dr
2
n
(
r
)
r
dr
∫
0
CAMS/CMA
3
24h Forecast (Initial:18Z 2nd Oct. 2007)
Area(10°-30°N，120°-140°E) water content vertical integrate, Unit：kg m-2
cloud
ice
rain
Vertical
integral
Vertical
Section
Along 20N, Vertical Section,Unit：kg/kg
CAMS/CMA
18km,Lin scheme
snow
Input of RT model from background
cloud fraction
mean
profile
cloud
ice
rain
Along 20N, Effective Radius,
CAMS/CMA
Unit：um
snow
RTTOV AMSUA CH1: diff. cloud type
clear
cloud
rain
snow
ice
observation
CAMS/CMA
4 water type
CRTM AMSUA CH1: diff. cloud type
clear
cloud
rain
snow
ice
observation
CAMS/CMA
4 water type
RTTOV AMSUB CH1: diff. cloud type
clear
cloud
rain
snow
ice
observation
CAMS/CMA
4 water type
CRTM AMSUB CH1: diff. cloud type
clear
cloud
rain
snow
ice
observation
CAMS/CMA
4 water type
Impact of water content on O-B
RTTOV 和CRTM正向模式在晴空和加水物质条件下AMSUA/B各通道的
AMSUA
AMSUB
模拟亮温相对于卫星实际观测亮温的偏差
15
35
RTTOV_amsua_dtb
10
RTTOV_amsub_dtb
30
5
clear
cloud
ice
rain
snow
4wat
25
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
-10
clear
cloud
ice
rain
snow
4wat
-15
-20
-25
20
bias
RTTOV
bias
-5
10
5
0
-30
1
chan
-35
15
15
2
-5
35
CRTM_amsua_dtb
3
4
5
chan
CRTM_amsub_dtb
10
30
clear
cloud
ice
rain
snow
4wat
5
25
0
bias
-5
1
2
-10
-15
-20
-25
3
4
5
6
7
8
clear
cloud
ice
rain
snow
4wat
CAMS/CMA
10
11
12
13
14
15
20
15
10
5
0
-30
-35
9
bias
CRTM
chan
1
2
3
chan
4
5
AMSUA: water vapor Jacobian
CAMS/CMA
RTTOV Jacobian:T,Qv,Qc,Qi,Qr,Qs
AMSUA
RTTOV Jacobian模式输出的温度、湿度和各水物质AMSUA15个通道响应函数
CAMS/CMA
CRTM Jacobian:T,Qv,Qc,Qi,Qr,Qs
AMSUA
CRTM Jacobian模式输出的温度、湿度和各水物质AMSUA 15个通道响应函数
CAMS/CMA
RTTOV Jacobian:T,Qv,Qc,Qi,Qr,Qs
AMSUB
RTTOV Jacobian模式输出的温度、湿度和各水物质AMSUB 5个通道响应函数
CAMS/CMA
CRTM Jacobian:T,Qv,Qc,Qi,Qr,Qs
CAMS/CMA
AMSUB
Impact of cloud fraction on simulted Tb:RTTOV
云量取0.8时AMSUA/B各通道亮温相对于标准云量下的偏差增量：
RTTOVma_deta0.8_piancha
1
2
3
4
5
6
7
8
9
10
RTTOVma_deta0.8_piancha
11
12
13
14
15
1
0.2
3
4
5
6
7
8
9
10
11
12
13
14
15
0.2
0
0
-0.2
-0.2
-0.4
-0.4
-0.6
-0.6
cloud
ice*10
rain
snow
4wat
-0.8
-1
-1.2
bias
bias
2
-1
-1.2
-1.4
0.8
cloud
ice*10
rain
snow
4wat
-0.8
-1.4
-1.6
-1.6
-1.8
-1.8
chan
chan
云量取0.5时AMSUA/B各通道亮温相对于标准云量下的偏差增量：
RTTOVmz_deta0.5_piancha
RTTOVmb_deta0.5_piancha
1
6
0
2
3
4
5
6
7
8
10
11
12
13
14
-5
bias
cloud
ice*10
rain
snow
4wat
-3
3
0.5
2
1
0
-6
-7
cloud
ice
rain
snow
4wat
5
15
4
-4
CAMS/CMA
9
-2
bias
台风中心区域云量的垂
直廓线（标准云量）
1
-1
chan
1
-1
2
3
chan
4
5
Impact of effective radius on simulted Tb: CRTM
云水、冰水、雨水和雪有效粒子半径分别取15、30、200、200（单位：μm）
cloud:15, ice:30,rain:200, snow:200 um
时CRTM模拟各通道亮温相对于标准情况下的偏差增量：
CRTM_cnstma_piancha
8
1
bias
2
2
3
4
5
6
7
8
9
10
11
12
13
14
15
-6
cloud
ice
rain
snow
4wat
-2
-8
-4
-10
-6
chan
-8
5
-4
0
1
4
-2
bias
4
CRTM_cnstmb_piancha
3
0
cloud
ice
rain
snow
4wat
6
2
chan
-12
云水、冰水、雨水和雪有效粒子半径分别取8、240、500、1400（单位：μm）
cloud:8, ice:240,rain:500, snow:1400 um
时CRTM模拟各通道亮温相对于标准情况下的偏差增量：
0.1
CRTM_cnstma_piancha
1
CRTM_cnstmb_piancha
0
0.5
1
2
3
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
bias
cloud
ice
rain
snow
4wat
-1
-1.5
-0.2
-0.3
cloud
ice
rain
snow
4wat
-0.4
-0.5
-2
chan
-2.5
5
15
bias
-0.5
CAMS/CMA
4
-0.1
-0.6
chan
Summary
 Background
 Clear Radiance assimilation in GRAPES
 Cloudy Radiance assimilation: onging Research
 Comparisons of RTTOV and CRTM for Cloudy Radiance
 Input from 24h forecast
 Jacobians
 Discussion
 Jacobian uncertainties for water content in RT models
 Inter comparison studies?
 OSSE: unknowns and obs. information content for cloudy
radiance(IR+MW). Could it be conducted by ITWG?
CAMS/CMA
Garand et al,2001:Radiance and Jacobian intercomparison of radiative transfer
models applied to HIRS and AMSU channels, JGR,106, 24017-24031
CRTM Jacobian Calculations Compared with RTTOV
 RTTOV is another fast
radiative transfer model used
by NWP community for
satellite data assimilation
 Radiance Jacobians at 6.2
and 7.2 micron water vapor
channels (GOES-R ABI and
MSG SEVIRI) are derived from
CRTM & RTTOV
 Both models produce
Jacobian profiles peaked at
the same altitude
 But the magnitudes are
slightly different between two
fast
models
CAMS/CMA
Assumption: surface emissivity = 0.98,
local zenith angle = 0 deg., and skin
temperature = 300 K
From Fengzhung Weng,2011