WATER VAPOUR COLUMN DENSITY PRODUCT FROM GOME-2:
VALIDATION WITH INDEPENDENT SATELLITE OBSERVATIONS
Margherita Grossi 1, Pieter Valks 1, Sander Slijkhuis 1, Diego Loyola 1, Bernd Aberle 1, Steffen
Beirle 2, Kornelia Mies 2, Thomas Wagner 2, Hans Gleisner 3, Kent Baekgaard Lauritsen 3
(1) DLR-IMF Deutsches Zentrum für Luft- und Raumfahrt/ Institut für Methodik der Fernerkundung,
Oberpfaffenhofen, Germany (2) MPI Max-Planck Institut für Chemie, Mainz, Germany (3) DMI Danish
Meteorological, Copenhagen, Denmark.
Abstract
Total column water vapour (TCWV) from measurements of the GOME-2 instruments aboard
EUMESAT's Met-Op-A satellite has proved to be a valuable input quantity in climate monitoring, and
could be useful for numerical weather prediction. However, the absolute accuracy estimate on the H2O
column needs further investigation. The intercomparison of H2O observations is a well established
method to asses the reliability of the data product. This contribution focuses on the validation of the
operational GOME-2 level 2 H2O data against independent satellite observations.
The operational GOME-2 H2O column used in this study are developed in the framework of O3M-SAF
and generated by DLR using the GOME Data Processor (GDP) version 4.5. The retrieval algorithm is
based on a classical DOAS method (Wagner et al., 2003; 2006) to obtain the trace gas slant column.
Subsequently, the vertical column is derived, making use of the simultaneously measured O2 absorption
and radiative transfer calculations. In this study we considered monthly averages of the global H2O VCD
observations computed from the level-2 data, for the years 2007 and 2008, and screen for cloud cover
conditions. We performed the validation using comparisons against global positioning system radio
occultation (GPS RO) measurements from CHAllenging Minisatellite Payload (CHAMP) as well as from
the Constellation Observing System for Meteorology, Ionosphere and Climate (COSMIC4) mission.
Total column water vapour estimates from GOME-2 satellite are then collocated and compared with
SSM/I measurements processed by the latest HOAPS 3.2 algorithm (oven ocean only) and with a
combined SSM/I + MERIS dataset (as developed in the framework of the ESA DUE GlobVapour
project).
Within our analysis, we found biases as small as +/- 1 kg/m2 between GOME-2 and all other datasets in
the period January 2007- December 2008. This confirms the results of previous validation studies which
showed small bias against SSM/I measurements (Slijkhuis et al., 2009; Kalakoski et al., 2010) and
obtained differences within 10% over ocean. We observe that the combined SSM/I-MERIS sample is
typically drier than the GOME-2 retrievals. For the comparison with GPS RO monthly mean, a mean
bias as small as 0.22 kg/m2 and -0.02 kg/m2 is found in the period 2007/2008 for CHAMP and COSMIC
measurements, respectively. We also observe that in general the annual variability over land and
coastal area is low, but over sea a clear seasonal cycle with dry bias over the northern hemispheric
winter and wet bias during the northern hemispheric summer is observed. These differences can mainly
be related to the impact of clouds on the accuracy of the GOME-2 observations and to the different
sampling statistics of the instruments.
GOME-2 SENSOR
The GOME-2 sensor is the follow up of the Global Monitoring Experiment (GOME) launched in 1995 on
ERS-2 (Burrows at al., 1999), and the SCIAMACHY sensor launched 1995 on ENVISAT (Bovensmann
et al., 1999). GOME-2 is a nadir viewing scanning spectrometer with a ground pixel size of 80 × 40 km2
and a scan width of 1920 km that provides an almost daily global coverage at the equator. The
instrument covers the same spectral range as GOME, i.e., from 240 to 790 nm, with spectral resolution
of about 0.5 nm in the visible spectral region. The German Aerospace Centre (DLR) plays a major role
in the design, implementation and operation of the GOME-2 ground segment for trace gas column
products, including total column water vapour (TCWV).
GOME-2 is mounted on the flight direction side of the MetOP-A satellite, which follows a Sunsynchronous orbit with a mean altitude of 817 km. The overpass local time at the equator is 09:30 LT
with a repeat cycle of 29 days. MetOp-A was launched on 19 October 2006 and GOME-2 products are
available from the 1st January 2007 onwards. A second GOME-2 type sensor, on board of the MetOp-B
satellite was launched on the 17th September 2012 and will operate in tandem with MetOp-A,
increasing the wealth of the data even further. The third and final satellite of the EUMETSAT Polar
System series, MetOp-C, is planned to be launched in 2017, and it is foreseen to guarantee the
continuous delivery of high-quality data until at least 2020.
WATER VAPOUR RETRIEVAL FROM GOME-2
A variety of methods for the retrieval of the Total Column Water Vapour (TCWV) from space-born
spectrometers operating in the visible region has been developed (AMC-DOAS: Noël et al., 1999; ERA:
Casadio et al., 2000; OCM: Maurellis et al., 2000; IGAM: Lang et al. 2003; Classical DOAS: Wagner et
al. 2003). The algorithm we used for the retrieval of the total column water vapour is based on a
classical Differential Optical Absorption Spectroscopy (DOAS) performed in the wavelength interval
614-683 nm. It consists of three basic steps (described in details by Wagner et al., 2003, 2006; Valks et
al. 2012).
In the first step, the spectral DOAS fitting is carried out, taking into account the cross sections of O2 and
O4, in addition to that of water vapour. To improve the broadband filtering, 3 types of vegetation spectra
are included in the fit, together with a synthetic Ring spectrum and, finally, an inverse solar spectrum to
correct for possible offsets, e.g. caused by instrumental stray light. In the second step, the water vapour
slant column density (SCD: the concentration integrated along the light path), is corrected for the nonlinearities arising from the fact that the fine structure water vapour absorption lines are not spectrally
resolved by the GOME instrument. This saturation effect can become especially important for large H2O
SCDs. In the last step, the corrected water vapour slant column density is divided by a ''measured" Air
Mass Factor (AMF) which is derived from the simultaneously retrieved O2 and it is defined as the ratio
between the measured SCD of O2 and the known VCD of O2 for a standard atmosphere. This simple
approach has the advantage that it corrects in first order for the effect of varying albedo, aerosol load
and cloud cover without the use of additional independent information.
However, because the vertical profile of WV is much more peaked in the troposphere with respect to
that of O2, we have to introduce an additional correction in the AMF computation. The correction factor
is strongly dependent on the solar zenith angle, on the exact vertical profile of H2O in the troposphere,
on the surface albedo and on cloud cover. In this analysis, we use two cloud indicators to screen for
cloudy measurements. We require that the product of cloud fraction and cloud top albedo does not
exceed 0.6 and we consider only satellite observations for which at least 80% of the O2 SCD is present.
It is also important to remark that, in contrast to most other algorithms, our WV analysis from GOME-2
does not rely on additional information, except for the use of an albedo database (based on surface
reflectivity data from Grzegorski 2009, and Koelemeijer et al., 2002) for the AMF correction. This serves
the aim to derive a climatologically relevant time series of TCWV measurements (Wagner et al., 2006;
Lang et al., 2007; Noël et al., 2008).
Figure 1: Monthly mean maps of TCWV from GOME-2 for June 2007 (on the left) and January 2008 (on the right).
Figure 1 shows the mean distribution of GOME-2 TCWV data for the months June 2007 and January
2008. In contrast to other satellite datasets, the GOME-2 product has the advantage that it covers the
entire Earth, including both sea and continents, leading to a more consistent picture of the global
distribution of the atmospheric humidity. The retrieval is performed in the visible spectral range and it is
very sensitive to WV in the lower troposphere, which contributes the major fraction of the total
atmospheric column. From Figure 1 we can observe that for both months there is high humidity in the
tropics and low humidity at higher latitudes. Also the movement of the inner tropical convergence zone
with seasons is clearly visible from the shift of the high water vapour column in the tropics between
June 2007 and January 2008. In both hemispheres, the total column precipitable water vapour
distribution follows the seasonal cycle of the near surface temperature and so the tropical total column
precipitable water has a maximum during the northern hemisphere (NH) summer, and a minimum in
winter.
VALIDATION OF THE GOME-2 H2O PRODUCT
We compare the GOME-2 WV column product with four independent satellites datasets. The first
dataset is based on passive microwave observations from SSM/I orbits processed with the latest
version of the HOAPS algorithm (HOAPS-3.2 dataset: Andersson et al, 2010; Fenning et al., 2012). It is
available over ocean only, needs model input and independent calibration (adjusted against radiosonde
data: Wentz et al., 1997). However it includes also TCWV measurements for cloudy scenes, both day
and night overpasses and spans a very large time range (SSM/I has been in orbit since 1987).
The second dataset consists of combined SSM/I + MERIS TCWV Level 3 data (Schroeder et al., 2010).
This dataset is based on radiance measurements in the microwave range taken by SSM/I over ocean
and TCWV retrievals from measurements of the visible and near infrared by MERIS (over land and
costal regions). MERIS provides water vapour column amounts for cloud free scenes for day time
overpasses over land with a very good spatial resolution. However, like for GOME-2, the quality of the
retrievals is affected by the presence of clouds.
Finally, we compare GOME-2 data with Global Positioning System Radio Occultation (GPS RO)
soundings from COSMIC (Rocken et al., 2000) and CHAMP (Wickert et al., 2001). We used the
COSMIC RO data collected during the period from January 2007 to December 2008 from the FM4
micro-satellite. Despite the relatively small amount of data available, GPS RO measurements are
valuable because their information can complement that provided by satellite radiance measurements.
GPS RO measurements have good vertical resolution and very high temporal resolution, are available
over all weather conditions over both land and sea.
Figure 2: Monthly mean TCWV bias between GOME-2 and four independent satellite datasets for the period January
2007 - December 2008.
VALIDATION RESULTS
We performed comparisons between GOME-2 and the four different datasets described above for the
time period January 2007- December 2008. Figure 2 shows the time series of the globally averaged
bias derived from the monthly mean TCWV distribution. The agreement between GOME-2 TCWV data
and the other independent satellite measurements is very good for all comparisons: the mean biases for
the full time series are close to 0, while the Root Mean Square Error (RMSE) varies between 3.5 and 7
kg/m2 (see Table 1). In this contest, the validation and intercomparison has been performed in such a
way that positive and negative bias imply respectively larger (and lower) GOME-2 data.
2
2
Data
Bias (kg/m )
RMSE (kg/m )
GOME-2 - SSM/I
0.23 +/- 0.64
3.5 +/- 0.45
GOME-2 - SSM/I+MERIS
-0.22 +/-0.27
4.0 +/- 0.46
GOME-2 - COSMIC4
-0.02 +/-0.63
7.0 +/- 0.24
GOME-2 - CHAMP
0.22 +/-0.63
6.9 +/- 0.31
Table 1: Statistics of the time period January 2007- December 2008. Bias and RMSE refer to the average difference
GOME2-DATA.
Further analysis of the GOME-2 and SSM/I HOAPS intercomparison (green line and points in Figure 2)
shows a positive bias in NH summer and a negative bias in NH winter, with the averaged TCWV for
SSM/I HOAPS being higher than GOME-2 (0.23 kg/m2,see Table 1). The monthly averaged bias ranges
from -0.62 kg/m2 in January 2007 to 1.25 in July 2007. We can infer a seasonal cycle in the geographic
distribution of the bias between clouds affected measurements (GOME-2) and all sky (HOAPS) data
sets, which is probably related, among other reasons, to the seasonability of cloud properties in some
locations, as well as the variability of geographic distribution of major cloud structure as the ITCZ. We
observe a similar seasonal variation in the bias between GOME-2 and GPS RO measurements.
However, in this case some of the differences in WVTC are a result of the limited number of CHAMP
and COSMIC measurements. Also the RMSE is larger for both the GPS RO observation due to the low
number of collocations with GOME-2. Finally, for the SSM/I + MERIS dataset, we can see that the
seasonal behaviour is not as evident as for the other comparisons, as a result of the different biases
over land (MERIS) and sea (SSM/I). In general the MERIS measurements present a wet bias with
respect to all other datasets, which might be partly caused by spectroscopic uncertainties, such as the
description of the WV continuum (Lindstrot at al., 2011). Interpreting these results, we should also have
in mind the limitations of GOME-2 retrieval. Although, as discussed before, a specific advantage of the
visible spectral region is that it is sensitive to the water vapour concentration close to the surface and
that it has the same sensitivity over land and ocean and can thus yield a consistent global picture, the
accuracy of an individual observation is, in general, reduced for cloudy sky observations. In addition,
since GOME-2 observations are made at 9:30 LT, especially in regions with a pronounced diurnal cycle,
they might not be representative for the daily, and therefore monthly, average TCWV.
Figure 3: Geographical distribution of the differences between GOME-2 TCWV product and SSM/I HOAPS measurements
in June 2007.
SSM/I HOAPS
In Figure 3 we show the geographical distribution of the differences GOME-2 - SSM/I HOAPS TCWV
data for the month June 2007, which presents an average positive bias of about 0.94 kg/m2. While both
dataset show similar WV distribution, it is clearly visible a higher bias in regions at high latitudes, in
particular in the northern areas of the Atlantic and Pacific. Some residual differences in the southern
hemisphere are probably due to the low number of remaining collocations for southern high latitudes,
which is due to the drift of the SSM/I orbit in the first part of 2007. The map also exhibits a number of
dry spots along the inner tropical convergence zone and the Pacific warm pool regions, most of which
can be attributed to unsatisfactory removal of clouds and the reduced number of measurements after
cloud screening. An orthogonal regression analysis of GOME-2 TCWV against SSM/I HOAPS shows a
very good correlation between the two datasets with a slope very close to 1 (0.997) and a positive
offset of about 1.01, which is compatible with the positive mean bias.
Figure 4: Geographical distribution of the differences between GOME-2 TCWV product and the combined SSM/I+MERIS
dataset in January 2008.
SSM/I+MERIS
The cross-comparison of the GOME-2 product versus the combined SSM/I+MERIS (GlobVapour)
dataset for January 2008 is shown in Figure 4. The MERIS TCWV algorithm is used above land and
costal areas and these measurements are combined with SSM/I derived TCWV data over ocean to
provide a global coverage. As can be seen in Figure 1, a systematic variation of the bias is observed
between winter and summer months also for this dataset, with lower GOME-2 values in the NH winter
season and higher data in summer. For the month January 2008 as shown in Figure 4, we found an
average bias of -0.5483.
Figure 4 shows that the agreement between GlobVapour data and GOME2 data over land seems to be
somewhat better than over ocean. The differences over ocean areas in January 2008 are quite noisy in
the difference plots and the GOME-2 data over ocean tend to be lower than the corresponding
SSMI+MERIS monthly mean. This is in line with the results we obtain from the comparison with SSM/I
data from the HOAPS algorithm for the same month. Over the continents the agreement between both
datasets is generally very good. Extended regions with very small biases, close to zero, are evident
especially in Asia and Africa. Exceptions are found in some specific small regions where GOME-2
columns are higher than the MERIS values. This over-estimation of water vapour content by GOME-2
(or underestimation by SSM/I + MERIS) seems to occur preferably over Europe and the western part of
North America. Major differences are located in coastal areas, where neither SSM/I, nor MERIS provide
accurate estimates. For MERIS, this is due to the weak reflectance of the ocean in the near infrared and
on the resulting uncertainties introduced by the unknown contribution of aerosol scattering and
absorption, while SSM/I measurements cannot be used in case of relative large footprint contaminated
by land. Finally, a negative bias is present in some specific regions like the Western Pacific Warm Pool,
South Africa and South America, and over regions of major orografic features (such as Himalaya and
Andes), mainly due to low statistics and uncertainties in cloud detections by GOME-2. A regression
analysis of GOME-2 total column against SSM/I+MERIS measurements show a good correlation
between both datasets (slope 0.96) and a small offset (only 0.17).
Figure 5: Zonal mean total water vapour vertical column from GOME-2 (mean values in red), GPS RO measurements
from CHAMP (left, green points with mean and standard deviation) and COSMIC FM4 (right, blue points with mean and
standard deviation) for the winter season (December-January-February) 2007 and 2008.
GPS RO
In Figure 5 we show results of the validation of GOME-2 with RO measurements. We perform an
analysis of multi year seasonal water vapour means for the winter months December-January-February
2007 and 2008 and for three datasets. The points in the plots denote the zonal mean water vapour
measurements as a function of latitude and the comparison is performed between GOME-2 data (red
points) and the GPRO measurements from CHAMP and COSMIC (green and blue points, respectively).
We can see that there is in general a good agreement: at all latitudes the amplitude of CHAMP and
COSMIC TCWV resembles the one of GOME-2 data. Some residual discrepancies are located
prevalently in the northern sub-tropical and mid-latitude zone, where GOME-2 shows a dry bias
compared to both COSMIC4 and CHAMP data, but the differences are within the uncertainty estimates
of the GPS-RO measurements. On the other hand, in the summer season (not shown here), at these
locations the GPS RO TCVW are slightly smaller then the GOME-2 and we can observe a positive bias,
with some residuals in the Southern sub-tropic zone in summer and spring. For both seasons, the bias
is larger in the northern hemisphere then in the southern one.
The validation between GOME-2 and both GPS RO measurements shows very similar results, but
wider deviations for CHAMP data than for COSMIC, which is likely to be caused by the limited number
of CHAMP data points. Results support also the expectation that the properties of different RO
observations are consistent.
SUMMARY
The operational GOME-2 H2O data used in this study are developed in the framework of O3M-SAF and
generated by DLR using the GOME Data Processor (GDP) version 4.5. The retrieval algorithm is based
on a classical DOAS method (Wagner et al., 2003; 2006) to obtain the trace gas slant column.
Subsequently the vertical column is derived, making use of simultaneously measured O2 absorption and
RT calculations. The algorithm is almost independent of external sources, provide nearly global
coverage, although is representative for cloud-free situations only, and shows the potential for
climatological studies.
In summary, the agreement between GOME-2 and the 4 independent datasets analyzed in this study is
very good, we found a mean bias within +/- 1 kg/m2 for the time interval January 2007- December 2008.
The annual variability of the bias over land (and coastal area) is low, but over sea a clear seasonal
cycle with highest values during Northern Hemisphere summer is observed for both SSM/I and GPS RO
measurements. We observe negative values for the bias in the Northern Hemisphere winter. The
GOME-2 data are typically drier than the COSMIC and CHAMP measurements in the northern
hemisphere winter, both considering the monthly and seasonal mean water vapour total columns and
higher in NH summer season, with minor differences located prevalently in the northern sub-tropical
area. Based on the results of the validation with independent satellites data, we can conclude that there
is a good agreement with GOME-2 H2O measurements for all seasons and latitude.
GOME-2 Level 2 total column and cloud products are available from the DLR ftp server in HDF5 format.
Web access is possible for NRT and offline products via http://atmos.caf.dlr.de/gome2. In the same
website are also available documents, reports, as well as quick look maps and links to related server. In
addition, GOME-2 products can be also ordered at the Help desk of the O3M-SAF hosted by the
Finnish Meteorological Institute (FMI) at [email protected].
ACKNOLEDGEMENTS
The development of the GOME-2 water vapour product and its validation has been funded by
EUMETSAT through the O3M-SAF Continuous Development and Operation Project (CDOP) and by
national contributions.
REFERENCES
Andersson, A., Fennig, K., Klepp, C., Bakan, S., Grassl, H., and Schulz, J. (2010): The Hamburg Ocean
Atmosphere Parameters and Fluxes from Satellite Data - HOAPS-3, Earth Syst. Sci. Data, 2, 215-234,
doi:10.5194/essd-2-215-2010
Bovensmann, H., Burrows, J.P., Buchwitz, M., Frerick, J., Noël, S., Rozanov, V. V., Chance K. V., and
Goede, A. H. P. (1999) SCIAMACHY - Mission objectives and measurement modes, J. Atmos. Sci., 56,
(2), 127-150
Burrows, J. P., Weber, M., Buchwitz, M., Rozanov, V., Ladstätter-Weißenmayer, A., Richter, A., de
Beek, R., Hoogen, R., Bramstedt, K., Eichmann, K.-U., Eisinger, M., and Perner, D. (1999) The Global
Ozone Monitoring Experiment (GOME): Mission Concept and First Scientific Results, J. Atmos. Sci.,
56, pp 151–175
Casadio, S., Zehner, C., Piscane, G., and Putz, E. (2000) Emperical Retrieval of Atmospheric Airmass
factor (ERA) for the Measurement of Water Vapour Vertical Content using GOME Data, Geophys.Res.
Lett. 27, pp 1483-1486
DUE GLOBVAPOUR Technical Specification Document (TSD), issue 1, revision 0, 16 April 2010
Fennig, K., Andersson, A., Bakan, S., Klepp, C., Schroeder, M., (2012): Hamburg Ocean Atmosphere
Parameters and Fluxes from Satellite Data - HOAPS 3.2 - Monthly Means / 6-Hourly Composites.
Satellite Application Facility on Climate Monitoring. doi:10.5676/EUM_SAF_CM/HOAPS/V001
Grzegorski, M. (2009) Cloud retrieval from UV/VIS satellite instruments (SCIAMACHY and GOME),
PhD thesis, University of Heidelberg
Koelemeijer, R.B.A., de Haan, J.F., Stammes, P.A (2002) Database of spectral surface reflectivity in
the range 335-772 nm derived from 5.5 years of GOME observations J. Geophys. Res. (Atm.), 108,
Issue D2, pp. ACH 8-1, DOI 10.1029/2002JD002429
Lang, R., Williams, J.R., Maurellis, A.N., and Van der Zande, W.J. (2003) Application of the Spectral
Structure Parametrization Technique: Retrieval of Total Water Vapour Columns from GOME, Atmos.
Chem. Phys. 3, pp 145-160
Lindstrot, L., Preusker, R., Diedrich, H., Doppler, L., Bennartz, R., and Fischer, J.(2011): 1-D-Var
retrieval of daytime total columnar water vapour from MERIS measurements, Atmos. Meas. Tech.
Discuss., 4, 6811-6844, doi:10.5194/amtd-4-6811-2011
Lang, R., Casadio S., Maurellis, A.N., and Lawrence M.G., (2007) Evaluation of the GOME Water
Vapor Climatology 1995-2002, J. Geophys. Res. 112, D12110, doi:10.1029/2006JD008246
Kalakoski, N., T. Wagner, K. Mies, S. Beirle, S.Slijkhuis, D. Loyola (2010) O3M SAF Validation Report,
Offline Total Water Vapour, SAF/O3M/FMI/VR/H2O/I02 in preparation
Maurellis, A.N., Lang, R., Van der Zande, W.J., Ubachs, W, and Aben, I (2000) Precipitable Water
Column Retrieval from GOME, Geophys. Res. Lett. 27, 903-906
Noël, S., Buchwitz, M., Bovensmann, H., Hoogen, R., and Burrows, J. P. (1999) Atmospheric Water
Vapor Amounts Retrieved from GOME Satellite Data, Geophys. Res. Lett. 26, p.1841
Noël, S., Mieruch S., Bovensmann, H., and Burrows, J. P. (2008) Preliminary results of GOME-2 water
vapour retrieval and first applications in polar regions. Atmos. Chem. Phys. 8, pp 1519-1529
Rocken, C., Y.-H. Kuo, W.Schreiner, D.Hunt, S. Sokolovskiy, and C. McCormick (2000), COSMIC
system description, Terr. Atmos. Oceanic Sci., 11(1), 21-52
Slijkhuis, S., S. Beirle, N. Kalakoski, K. Mies, T. Wagner, S. Noël, J. Schulz (2009) Comparison of H2O
retrievals from GOME and GOME-2, Proc. EUMETSAT P.55, http://www.eumetsat.int/Home/Main/
AboutEUMETSAT/Publications/ConferenceandWorkshopProceedings/index.htm
Valks, P.,Loyola, D., N.Hao, M.Rix and S.Slijkhuis (2012) Algorithm Theoretical Basis Document for
GOME-2 Total Column Products of Ozone, Minor Trace Gases and Cloud Properties (GDP 4.5 for
O3M-SAF OTO and NTO), DLR/GOME-2/ATBD/01, Iss./Rev.: 2/GG
http://atmos.caf.dlr.de/gome2/docs/DLR_GOME-2_ATBD.pdf
Wagner, T., Heland, J., Z¨oger, M., and Platt, U.(2003): A fast H2O total column density product from
GOME – Validation with in-situ aircraft measurements, Atmos. Chem. Phys., 3, 651–663
Wagner, T., Beirle, S., Grzegorski, M., and Platt, U (2006) Global trends (1996–2003) of total column
precipitable water observed by Global Ozone Monitoring Experiment (GOME) on ERS-2 and their
relation to near-surface temperature, J. Geophys. Res. 111, D12102, doi:10.1029/2005JD006523
Wentz, F.J. (1997) A well-calibrated ocean algorithm for SSM/I, J.Geophys.Res. 102, No. C4, pp.
8703-8718
Wickert Wickert J., Reigber, C., Beyerle, G., K¨onig, R., Marquardt, C., Schmidt, T., Grunwaldt, L.,
Galas, R., Meehan, T. K., Melbourne, W. G., and Hocke, K.(2001): Atmosphere sounding by GPS radio
occultation: First results from CHAMP, Geophys. Res. Lett., 28, 3263–3266