Water Vapour Retrieval from SCIAMACHY: AMC-DOAS

SCIAMACHY H2O Retrieval Results using Air Mass Corrected DOAS

First total water vapour column amounts have been retrieved from SCIAMACHY measurements in the (near-)visible spectral region around 700 nm. The algorithm used is based on a modified DOAS approach which has been already successfully applied to GOME data .
The general features of this method are:

Saturation effects arising from highly structured differential spectral features which are not resolved by the measuring instrument are accounted for.

O2 absorption features are fitted in combination with H2O to determine an air mass correction factor which compensates to some degree for insufficient knowledge of the background atmospheric and topographic characteristics, like surface elevation and clouds.

All calculations have been performed using a cloud free tropical reference atmosphere.
As GOME, the SCIAMACHY instrument can not see through clouds, which limits the applicability of the method. However, because the ground pixel size of SCIAMACHY is typically much smaller than for GOME (30 km x 60 km compared to 40 km x 320 km), the probability for cloud-free scenes is much higher for SCIAMACHY.

SCIAMACHY and GOME H2O total columns over Europe for 27 Jan 2003.
The smaller SCIAMACHY ground pixels are marked by boxes.
(Click to enlarge)

Although no clouds are considered in the reference atmospheric model, the method of air mass correction provides the possibility to retrieve meaningful H2O total columns also for partly cloudy scenes.
The size of the air mass correction factor can be used a criterium for the quality of the H2O data product. In the ideal case, i.e. if the atmospheric conditions used in the model calculations match the real conditions, the air mass correction factor should be 1. In the presence of clouds, only the atmosphere above the clouds can be probed by the instrument, so the effective amount of both O2 and H2O seen by SCIAMACHY is smaller than the amount used in the model calculations. The air mass factor correction would then try to compensate for this. In this case the correction factor would be smaller than 1.
If the air mass correction factor deviates too much from 1, this is an indication that the conditions of the reference atmosphere differ to much from reality. In this case the retrieved H2O columns are considered to be unrealistic.

In practice, it could be shown that data retrieved with air mass correction factors smaller than 0.8 are unreliable. Therefore, these data have been omitted from the following plots. In addition, measurements performed at high solar zenith angles (larger than 88 deg) and backscan pixels have been excluded.

All SCIAMACHY results presented here are based on preliminarily calibrated data and are thus also of preliminary nature.

The SCIAMACHY results have been compared with DMSP F14 SSM/I data provided by the Global Hydrology Research Center (GHRC) at the Global Hydrology and Climate Center, Huntsville, Alabama, and with assimilated global water vapour data provided by the European Centre for Medium-Range Weather Forecast (ECMWF).

Previous retrieval results (Noël et al., 2004a) using the AMC-DOAS algorithm with SCIAMACHY data revealed systematically about 10% lower water vapour columns derived from SCIAMACHY data in comparison with corresponding SSM/I data over ocean.
A similar offset was seen by the WFM-DOAS method.
This offset could be essentially removed by using an updated instrument slit function (see Noël et al., 2005). Therefore, the AMC-DOAS water vapour columns presented here have not been scaled.
For the comparison with correlative measurements, all data have been gridded to 0.5 deg x 0.5 deg.

SCIAMACHY H2O total columns for 27 Jan 2003 (swath data)
(Click to enlarge)

SSM/I H2O total columns for 27 Jan 2003, descending part only
(Click to enlarge)

ECMWF H2O total columns for 27 Jan 2003
(Click to enlarge)

Comparison between SCIAMACHY AMC-DOAS and SSM/I (descending) H2O total columns for 27 Jan 2003; red line = 1:1 correlation; r = linear Pearson's correlation coefficient
(Click to enlarge)

Comparison between SCIAMACHY AMC-DOAS and ECMWF H2O total columns for 27 Jan 2003; red line = 1:1 correlation; r = linear Pearson's correlation coefficient
(Click to enlarge)

The SCIAMACHY results agree within a scatter of about 0.5 g/cm2 with corresponding SSM/I and ECMWF water vapour data. This deviation includes contributions from the temporal and spatial variability of water vapour. In fact, the mean deviation between the SCIAMACHY and the correlative data sets is much smaller: the SCIAMACHY total water vapour columns are typically about 0.2 g/cm2 lower than the SSM/I values and less than 0.1 g/cm2 lower than corresponding ECMWF data.
The SCIAMACHY water vapour results agree well with correlative data not only over ocean but also over land, thus showing the capability of SCIAMACHY to derive water vapour concentrations on the global scale.

If you are interested in more information on SCIAMACHY H2O, please contact Stefan Noël.

Data Access:

Limited by the availability of SCIAMACHY Level 1b data products, AMC-DOAS water vapour total column data have been derived for the year 2003 and later. The analysis is ongoing, have a look at the SCIAMACHY support for ICARTT web site for most recent results.
On request, access to these data is possible under certain conditions; click here for details.

Note:
AMC-DOAS water vapour total columns are also available for the SCIAMACHY validation master set. Access to these data is provided via the SCIAVALIG web site.

References:

S. Noël, M. Buchwitz, H. Bovensmann, R. Hoogen, and J. P. Burrows, Water vapour retrieval from GOME data, Proc. European Symposium on Atmospheric Measurements from Space, vol. 1, 375-381, 1999. (PS file, ca. 425 kB)

S. Noël, M. Buchwitz, H. Bovensmann, R. Hoogen, and J. P. Burrows, Atmospheric Water Vapor Amounts Retrieved from GOME Satellite Data , Geophys. Res. Lett., vol. 26(13), 1841-1844, 1999. (PS file, ca. 217 kB)

S. Noël, H. Bovensmann, J. P. Burrows, Water vapour retrieval from GOME data including cloudy scenes, Proc. ENVISAT/ERS Symposium, Gothenburg, 2000. (PDF file, ca.1.7 MB)

S. Noël, M. Buchwitz, H. Bovensmann, and J. P. Burrows, Retrieval of Total Water Vapour Column Amounts from GOME/ERS-2 Data, Adv. Space Res., 29(11), 1697-1702, 2002. (PS file, ca. 885 kB)

S. Noël, M. Buchwitz, J. P. Burrows, First retrieval of global water vapour column amounts from SCIAMACHY measurements, Atmos. Chem. Phys., 4, 111-125, 2004a. (PDF file at ACP web site, ca. 22.6 MB)

S. Noël, M. Buchwitz, H. Bovensmann, J. P. Burrows, SCIAMACHY water vapour retrieval using AMC-DOAS, to be published in Proc. ENVISAT Symposium, Salzburg, Austria, 6-10 September, 2004b. (PDF file, ca. 1.4 MB)

S. Noël, M. Buchwitz, J. P. Burrows, K.-U. Eichmann, H. Bovensmann, Validation of SCIAMACHY AMC-DOAS water vapour columns, Atmos. Chem. Phys. Discuss., 5, 1925-1942, 2005. (link to corresponding ACPD web site)

SSM/I H2O total columns for 27 Jan 2003, descending part only (Click to enlarge)	ECMWF H2O total columns for 27 Jan 2003 (Click to enlarge)
Comparison between SCIAMACHY AMC-DOAS and SSM/I (descending) H2O total columns for 27 Jan 2003; red line = 1:1 correlation; r = linear Pearson's correlation coefficient (Click to enlarge)	Comparison between SCIAMACHY AMC-DOAS and ECMWF H2O total columns for 27 Jan 2003; red line = 1:1 correlation; r = linear Pearson's correlation coefficient (Click to enlarge)