Data Assimilation and Use of EOS Data
in Land Surface Model
National Satellite Meteorological Center, CMA
Lu Qifeng, Yang Zhongdong, Lu Naimeng, Zhang Peng, Tang Shihao
E-mail: luqf@cma.gov.cn, lqifeng@163.com
Impact on the Soil temperature by assimilating LAI
Designing Conception on the Land Satellite Data
Assimilation System of NSMC/CMA
•
the first load REOF展开第一载荷向量RLV
vector of REOF, RLV
•The China Land Data Assimilation System (CLDAS) project aims at
the construction and evaluation of a routine system for soil moisture
and temperature initialization in NWP applications, which is partially
inspired on the NSMC/CMA FY-3 Satellite system Project, containing
assimilation on the following variables:
•Assimilation for preparation of forcing fields: One 3Dvar analysis
system, based on datasets of reanalyzed ECMWF (or NCEP), in situ
observation and satellite retrieved products, was developed to prepare
the forcing fields (radiation, air temperature, precipitation, humidity
and wind, etc.) needed by land surface model CLM.
•Assimilation for predicted variables: Both atmospheric and land
surface variables (soil temperature and moisture) are taken as analysis
variables. The Kalman filter and 3Dvar method are adopted to design
the assimilation scheme.
.05
.04
.36
.02 .19
.13
-.18
.24
.14
-.29
-.29
.16
-.13
-.11
-.05
-.11
.1.1 -.06
-.11
-.02
-.15 -.02
-.15
stands for results of -.07 -.01
为加入动态的LAI模拟结果经
dynamic LAI from REOF, 00 00
REOF旋转后，RLV≤
0的区域
ie, RLV≤0
.11
.12
-.04
-4.168 -6.338
4.112
-4.186-6.351
5.273
5.271
.6
.535
-.722
3.072 4.359
-.343
-2.892
00
3.12
-.379 3.071
-2.924
0
0
0
Time coefficient
.678 -1.823七月份0.37m土壤温度
0.37m soil temp in July 4.132
3.039 4.424
-.31 00
00 0
the time coefficient of first three load vector
-4.42
-4.453
.975
-.13
00 -.14
-.31
Bold itatic number for
黑色粗斜体数字为动态LAI模拟
results of dynamic LAI
黑色字体数字为静态LAI模拟
Bold number for static
of dynamic LAI
the third load
vector of REOF, RLV
REOF展开第三载荷向量RLV
-2.329
.09
-.11
.11
.11
-.09 -.01
-.09
stands for results of .14 .02
为加入动态的LAI模拟结果经
dynamic LAI from REOF, 00 00 0.03
.12.1
ie, RLV≤0
REOF旋转后，RLV≤
0的区域
00
Bold itatic number for
黑色粗斜体数字为动态LAI模拟
00 00
results of dynamic LAI
0
0
Bold number for static of
黑色字体数字为静态LAI模拟
dynamic LAI
-3.722
.09
.36
.27
.27
-.11
-.01
-3.664
.36
0.01
.18
-.1
-.2
-.22
-2.366
.06
.18
.18
-.36
.976
.06
.18
.12
.12
.16
-.35
-.25
-.25
-.13
.29
.29
-.35 -.03
-.36 -.03
.23
.14
.23
.24
-.19
.08
.07
•Assimilation of radiance: The fast radiation transfer model was taken
as observation operators for the direct assimilation of radiance.
Frame of Land Satellite Data Assimilation System
.13
.36
-.34
-.35
0.37m soil temp in July -.15
-.15 -.13 七月份0.37m土壤温度
.2
.01
-.18
-.15 -.18
-.13
.12
0.37m soil temp in July
七月份0.37m土壤温度
.14
-1.844
•Assimilation for non-predicted variables: We designs the common
frame of observation operator for non-predicted variables, adopting
different observation operators according to the variable itself, with its
corresponding tangent linear model and adjoint model constructed
from interpolation operator or some experience retrieval algorithms.
The second load
vector of REOF, RLV
REOF展开第二载荷向量RLV
.1.1
-.02
-.21 -.02
-.745
00
time
-3.831
1.345
-3.875
1.409
Variable
3.084
2.5173.045
2.542
-2.013
为加入动态的LAI模拟结果经
stands for results of
0
dynamic LAI from REOF, 0 00 3.778
REOF旋转后，RLV≤
0的区域
ie, RLV≤0
-.639
-.602
.272
.839
Bold itatic number for
黑色粗斜体数字为动态LAI模拟
results of dynamic LAI
黑色字体数字为静态LAI模拟
Bold number for static of
dynamic LAI
Accumulated variance of first three load
0.37m Soil temperature
1
2
3
STA
0.4521
0.6782
0.8220
DYN
0.4512
0.6756
0.8206
.837 00
00 0
0
0
0
Dynt1-3 for the time coeffient the first three
Impact on the Soil Content Water by assimilating LAI
the first loadREOF展开第一载荷向量RLV
vector of REOF, RLV
REOF展开第二载荷向量RLV
The second load
vector of REOF, RLV
-.17
.09
.07
.25
.1
-.21
0.37m soil water in July
-.04 -.12七月份0.37m土壤水含量
.16
-.31
.21
-.11
-.49
-.01
-.02
-.04
.11
.08
-.06
-.22
.06
.21
-.2
-.02
-.08
.2
.03
-.13
-.06
.15
-.04
.06
-.03
-.06
-.06
.05
-.18
-.27
Bold itatic number for
黑色粗斜体数字为动态LAI模拟
results of dynamic LAI
黑色字体数字为静态LAI模拟
Bold number for static of
dynamic LAI
-.2
.4 .15
.32 .05
.17
.15
.03
.02
.04
.02
-.24
-.26
-.24
-.31
.52
-.1
.56
.05
-.36
-.19
-.18
-.26
.14
.1
.08
.14
.03
-.08
.11
.17 .01
-.23
为加入动态的LAI模拟结果经
stands for results of.19
dynamic LAI from REOF, 00 00 0.43
REOF旋转后，RLV≤
0的区域
ie, RLV≤0
-.12
.08
0.37m soil water in July .13
.18七月份0.37m土壤水含量
-.08
.24
.17
-.18
-.06
.15 -.1
.55
-.04
0
-.03 .07
.02
0-.11
stands for results of .14 .15
为加入动态的LAI模拟结果经
dynamic LAI from REOF, 00 00 0.17
ie, RLV≤0
REOF旋转后，RLV≤
0的区域
Bold itatic number for
黑色粗斜体数字为动态LAI模拟
results of dynamic LAI
Bold number for static
黑色字体数字为静态LAI模拟
of dynamic LAI
.03
.05
.13
-.03
.18 00
00 0
0
0
0
.12
.25
-.08
-.13
-.07
00
0000
00
the time coefficient of first three load vector
the third load
vector of REOF, RLV
REOF展开第三载荷向量RLV
3.789
4.377
-2.469 .178
0.37m soil water in July-3.942
-2.691 .929七月份0.37m土壤水含量
2.176
simu
obs
ana_lst
1
25
moisture
temperature
28
27
26
25
24
23
22
21
20
19
hour
49
0.525
0.52
0.515
0.51
0.505
0.5
0.495
0.49
0.485
0.48
.367
.56
5.841
6.165
simu
obs
ana_lst
moisture
temperture
24
23
22
21
20
19
1
25
hour
49
-1.6
-3.393
-.429 -3.46
time
-1.867
00
1.883
-1.09
Variable
2.315
Accumulated variance of first three load
25
hour
49
73
-3.244
stands for results of1.412 -1.355 -2.983
为加入动态的LAI模拟结果经
dynamic LAI from REOF, 00 00 0
-2.834
REOF旋转后，RLV≤
0的区域
ie, RLV≤0
00
Bold itatic number for
黑色粗斜体数字为动态LAI模拟
0000
results of dynamic LAI
0
0
Bold number for static of
黑色字体数字为静态LAI模拟
dynamic LAI
0.37m soil moisture
1
2
3
STA
0.4521
0.6782
0.8220
DYN
0.4512
0.6756
0.8206
2.594
2.486
-1.022
soil moisture at 0-5cm
simu
obs
ana_moi
25
3.312
-.153
-3.626-1.651
Soil temperature at 0-5cm
27
26
-2.697
3.21 -3.058
2.918 -3.058
-1.437
Effect of assimilating surface soil temperature on the analyzed soil temperature and soil
moisture, on the left for analyzed soil temperature, on the right for analyzed soil moisture
28
-2.467
3.496
00
-.484 -.769
-.657-.456
-.98
-1.124
-2.085
1
73
-3.977
2.07
Soil moisture at 0-5cm
Soil temperature at 0-5cm
Time coefficient
Validation by a Single Site Test
73
0.525
0.52
0.515
0.51
0.505
0.5
0.495
0.49
0.485
0.48
Conclusion
simu
obs
ana_simu
1
25
hour
49
73
Effect of assimilating surface soil moisture on the analyzed soil temperature and soil
moisture, on the left for analyzed soil temperature, on the right for analyzed soil moisture
ITSC-16 May 2008 Brazil
The validation from Yucheng comprehensive experiment site and REOF analysis
from 20-year LAI assimilation indicates that the preliminary results obtained are
still inspiring. There are still many detailed work to do for the routine operation
of LDAS, such as how to get dynamic P in 3dvar, how to select the spacing
interpolation algorithm, etc.