Comparison of AMSU-B Brightness Temperature with Simulated
Brightness Temperature using Global Radiosonde Data
V.O. John, S.A. Buehler, and M. Kuvatov
Institute of Environmental Physics, University of Bremen, Otto-Hahn Alle 1,
28359 Bremen, Germany, vojohn@uni-bremen.de
Abstract
We present a comparison of brightness temperature measured by AMSU to radiative transfer model calculations based on radiosonde data. The forward model used is
the stable version of the Atmospheric Radiative Transfer Simulator(ARTS), a general
purpose radiative transfer model which can handle many different remote sensing instruments in the millimeter to infrared spectral region. The atmospheric profiles used
are the Met Office - Global Radiosonde Data taken from the British Atmospheric
Data Center (BADC). As a first step, the comparison is done for Lingenberg, Germany which uses Vaisala RS80 radiosondes and Kem, Russia which uses Goldbeater’s
skin radiosondes.
As the forward model ARTS has already been validated against AMSU brightness
temperatures using high resolution radiosonde data from Lindenberg which is a reference station for German Weather service(DWD), the main aim of this comparison is
to check the quality of the radiosonde data from the different stations.
Introduction
Upper tropospheric humidity (UTH) is a crucial parameter for meteorology and climate
research. There are two global and continuous data sets for this parameter, one from
polar orbiting meteorological sensors, the other from synoptic meteorological radiosondes.
The basic idea of the study is to compare satellite and radiosonde data. The atmospheric
radiative transfer simulator - ARTS [1] is used to generate simulated AMSU measurement
from the radiosonde data.
The aims of the study are, to develop a robust methodology for such a comparison and
to pave the way for a systematic comparison of all stations in the global radiosonde network
to satellite data. This will allow an intercomparison and quality control of the different
radiosonde stations, assuming that the satellite instrument’s properties are stable during
a few orbits. In this article a comparison between two mostly used radiosonde instruments
namely Vaisala RS80 and Goldbeater’s skin radiosondes is demonstrated.
Data Used
Radiosonde data
To develop the methodology for the comparison we used radiosonde data from Lindenberg
(52.22◦ N, 14.12◦ E) which is a reference station of the German Weather Service, DWD.
This station uses RS-80 Vaisala radiosondes for routine operations. The reason for selecting this particular station is that the data are subjected to a number of corrections as
UTH [500 hPa - 200 hPa]
20
<UTHLRC> = 35.506
2002
<UTHBADC> = 35.622
10
<UTHHRC> = 35.259
∆ UTH [ % ]
<UTHHRNC> = 31.905
0
-10
-20
<∆ UTHLH> = 0.247 ± 1.204
LRC
<∆ UTHBH> = 0.363 ± 1.238
HRNC
<∆ UTHHH> = -3.354 ± 2.692
BADC
0
200
400
600
Launches
800
1000
1200
Figure 1: ∆U T H for different versions of the Lindenberg radiosonde data for the year
2002 with HRC as the reference.
described in [3]. Three versions of the data have been received: a) High Resolution Not
Corrected (HRNC), High Resolution Corrected (HRC), and Low Resolution Corrected
(LRC). Another version of the data set can be obtained from British Atmospheric Data
Center (BADC). Figure 1 shows a comparison of upper tropospheric humiditiy (UTH) of
the different versions of the data.
The uncorrected version of the data shows a significant dry bias [6] in UTH but the
other two versions (LRC and BADC) do not show any significant difference in UTH.
AMSU data
Advanced Microwave Sounding Unit - B (AMSU-B) is a cross track scanning microwave
humidity sounder [4]. The footprint size is 20×16 km 2 for the innermost scan positions,
but increases to 64×52 km2 for the outermost positions. Figure 2 shows an example of
AMSU data from Channel 20. This is an overpass over the station Lindenberg. The circle
represents a 50 km area around the station and the pixels in this area are used for the
comparison. If there is a cloud over the station, there will be a large inhomogeneity among
the pixels as shown in the figure 2.
Figure 2: An AMSU overpass over station Lindenberg. The circle drawn around the
station has a radius of 50 km.
Results and Discussion
The methodology of our comparison, the validation of the RT model, and the intercomparison between different satellites can be seen in [2]. Figure 3 shows the comparison for
the two stations Lindenberg, Germany and Kem (64.95 ◦ N, 34.65◦ E), Russia for the year
2001. There are significant differences in channels 18 and 19, the two channels which are
sensitive to the upper tropospheric humidity. The bias, slope and offset were calculated
as explained in [2].
There is a slope in the channel 18 for Lindenberg. The possible explanation for the
slope is that the correction of the data is not sufficient enough in extremely dry conditions.
Therefore the modeled brightness temperatures are higher than the measured ones because
the weighting functions peak lower in a drier atmosphere.
A large bias exists in the case of Kem for channels 18 and 19. The modeled brightness
temperatures overestimate the measured ones because there is a wet bias in the radiosonde
humidity measurements using Goldbeater’s skin radiosondes [5].
In order to translate the bias in brightness temperature to bias in relative humidity we
made a sensitivity chart. This is shown in Figure 4. It can be seen that a bias of about
7 K in channel 18 is equivalent to that of 10–12 % in RH. This result is consistent with the
results of Soden and Lanzante [5].
Kem [Russia]
AMSU 18 [ 183.31 ± 1 GHz ]
AMSU 18 [ 183.31 ± 1 GHz ]
BIAS = -0.59 ± 0.06
CORR = 0.95
N = 220
250
240
230
230
240
250
260
ARTS TB [ K ]
240
220
220
270
AMSU 19 [ 183.31 ± 3 GHz ]
BIAS = -0.27 ± 0.06
CORR = 0.97
N = 220
270
250
240
240
SLOPE = 0.94 ± 0.02
OFFSET = 15.70 ± 4.89
250
260
270
ARTS TB [ K ]
260
250
250
SLOPE = 0.99 ± 0.02
OFFSET = 1.74 ± 5.50
260
270
280
ARTS TB [ K ]
290
260
250
SLOPE = 1.05 ± 0.05
OFFSET = -9.00 ± 13.43
240
280
275
270
240
250
ARTS TB [ K ]
BIAS = 2.75 ± 0.13
CORR = 0.94
N = 36
240
AMSU 20 [ 183.31 ± 7 GHz ]
280
230
260
280
BIAS = -0.86 ± 0.06
CORR = 0.96
N = 220
SLOPE = 0.95 ± 0.06
OFFSET = 18.99 ± 15.01
AMSU 19 [ 183.31 ± 3 GHz ]
270
260
290
250
230
SLOPE = 0.92 ± 0.02
OFFSET = 18.99 ± 4.62
AMSU TB [ K ]
AMSU TB [ K ]
280
AMSU TB [ K ]
AMSU TB [ K ]
260
BIAS = 6.66 ± 0.14
CORR = 0.86
N = 36
260
AMSU TB [ K ]
AMSU TB [ K ]
270
Lindenberg
270
250
260
ARTS TB [ K ]
270
AMSU 20 [ 183.31 ± 7 GHz ]
BIAS = 0.10 ± 0.13
CORR = 0.93
N = 36
265
260
255
250
SLOPE = 1.04 ± 0.06
OFFSET = -10.79 ± 15.11
245
245 250 255 260 265 270 275 280
ARTS TB [ K ]
Figure 3: Comparison between modeled and measured radiance for Lindenberg and Kem.
Year: 2001, Satellite: NOAA-16. The blue curve shows a linear fit of modeled and measured radiance.
∆ TB [ K ]
4
2
0
-2
-4
-6
-8
-10
0
5
Wettest
10
20
30
40
50
20
30
40
50
20
30
∆ RH [ % ]
40
50
Medium
∆ TB [ K ]
0
-5
-10
-15
-20
∆ TB [ K ]
0
10
5
10
Driest
0
-5
-10
-15
-20
0
Ch-16
Ch-17
Ch-18
Ch-19
Ch-20
10
Figure 4: Sensitivity of ∆TB to ∆RH for the AMSU-B channels. There profiles were
selected based on the total water vapor content labelled as wettest, medium, and the
driest.
Summary and Future Work
A robust methodology is developed for the comparison of satellite microwave humidity
data with radiosonde data. This method is sensitive to the dry and wet biases in the
radiosonde humidity measurements taking the satellite measurement as reference.
We plan to apply this method for other stations and types of radiosonde sensors to
find the possible biases in data set.
Acknowledgments
We thank Lisa Neclos, CLASS, NOAA, for providing AMSU data, Ulrich Leiterer and
Horst Dier of Lindenberg weather station, for providing the radiosonde data. Furthermore,
we wish to acknowledge funding through the German AFO 2000 research project UTHMOS (GFS 07ATC04). It is a contribution to COST Action 723 ‘Data Exploitation and
Modeling for the Upper Troposphere and Lower Stratosphere’.
References
[1] S. A. Buehler, P. Eriksson, T. Kuhn, A. von Engeln, and C. Verdes. ARTS, the
Atmospheric Radiative Transfer Simulator. JQSRT, submitted 2003. Preprint at:
http://www.sat.uni-bremen.de.
[2] S. A. Buehler, M. Kuvatov, V. O. John, U. Leiterer, and H. Dier. Comparison of
Microwave Satellite Humidity Data and Radiosonde Profiles: A Case Study. JGR,
Submitted 2004. Preprint at: http://www.sat.uni-bremen.de.
[3] U. Leiterer, H. Dier, and T. Naebert. Improvements in radiosonde humidity profiles
using RS80/RS90 radiosondes of Vaisala. 70(4):319–336, 1997.
[4] R. W. Saunders, T. J. Hewison, S. J. Stringer, and N. C. Atkinson. The radiometric
characterization of AMSU-B. 43(4):760–771, 1995.
[5] B. J. Soden and J. R. Lanzante. An Assessment of Satellite and Radiosonde Climatologies of Upper-Tropospheric Water Vapor. J. Climate., 9:1235–1250, 1995.
[6] J. Wang, H. L. Cole, D. J. Carlson, E. R. Miller, and K. Beierle. Correction of
Humidity Measurement Errors from the Vaisala RS80 Radiosonde - Application to
TOGA COARE Data. J. Atm. Ocean. Tech., 19:981–1002, 2002.