MSG/SEVIRI Sea Surface Temperature data record

Cécile Hernandez's picture
June 2010: on the left, nightime monthly mean SST from the MSG reprocessed data; on the right, monthly mean difference to a climatology of the OSTIA reanalysis from 1985 to 2007.

A few months ago, the OSI SAF released a data record of Sea Surface Temperature (SST) on the full Earth disc seen by Meteosat in 0°.

The data record covers eight years of data, from 2004 to 2012 and is derived from the SEVIRI imager on Meteosat Second Generation (MSG) satellites (Meteosat-8 and Meteosat-9).

Since the beginning of the 2000s, the OSI SAF processes Meteosat data in near-real time and distribute hourly SST products, usually 2 hours after the sensing time. The algorithm and the software used to process these data as soon as possible regularly evolves to take into account scientific and technical progress and also the new satellites.

The aim of the reprocessing is to provide users with a consistent data record computed using the latest scientific advances in SST retrieval.


How was this SST data record generated ?

The infrared radiance measured by the SEVERI imager corresponds to the energy emitted by the sea surface and the atmosphere above. The challenge is to process the data to retrieve only the surface temperature of the oceans. It is done by using two window channels which have different sensitivity to the atmosphere component (mainly to water vapor content). SST is actually retrieved from SEVIRI infrared channels (10.8 and 12.0 µm) using a multispectral “split-window” algorithm. An algorithm correction scheme is used to mitigate the regional and seasonal bias due to the inability of the algorithm to cope with the whole range of atmospheric conditions. It relies on atmospheric profiles of temperature and humidity from a numerical weather prediction model and a radiative transfer model.

State of the art input datasets were gathered: reprocessed satellite brightness temperature from EUMETSAT, reprocessed cloud mask from the Climate SAF, ERA-interim atmospheric profiles from ECMWF, OSTIA Sea Surface Temperature re-analysis from CMEMS.

The subskin SST is retrieved at full spatial and temporal resolution (each 15 minutes). It is aggregated in hourly products and remapped onto a 0.05° regular grid. It is then formatted following the Data Specification (GDS) version 2 from the Group for High Resolution Sea Surface Temperatures (GHRSST).

The quality assessment of this data record was done against drifting buoy, Argo float and moored buoy measurements. It shows coherent and stable results over time. For example, the bias/standard deviation during night-time are -0.06/0.44 K for Meteosat-8 and -0.08/0.48 K for Meteosat-9, and during daytime are 0.01/0.42 K for Meteosat-8 and 0.00/0.46 K for Meteosat-9. Regional biases are low thanks to the algorithm correction scheme. For comparison, the bias and standard deviation during night-time for the operational Meteosat SST product (OSI-206) over 2012 are -0.03 and 0.57K respectively.


Possible uses of this data record

The use of this final data record is granted to every interested user, free of charge and at the condition to show EUMETSAT’s copyright credit on each production. The files are available on OSI SAF FTP server (credentials are provided on request, after registration on OSI SAF web site) and soon at the EUMETSAT data centre. Data can be subsetted using the OpenDAP interface provided with our OSI SAF Thredds server.

This SST data record provides a homogeneous series with a quality adapted, for example, for the study of oceanic surface phenomena with high spatial-temporal variability and ocean-atmosphere interactions.

More information on this product are available on the product web page, including the Product User Manual and Validation Report.

Whether your are interested in this product as a meteorologist, as a climatologist, as an oceanographer, as a researcher for fisheries, as navy... we are much interested in receiving your feedback, and would appreciate your acknowledgement in using and publishing about the data, and like to receive a copy of any publication about the application of the data. Your feedback helps us in maintaining the resources for the OSI SAF services.


Want to see what it looks like ? Go to the quicklooks available for each hour on the OSI SAF web site.


The figure on top shows:
June 2010: on the left, nightime monthly mean SST from the MSG reprocessed data; on the right, monthly mean difference to a climatology of the OSTIA reanalysis from 1985 to 2007.