Map Viewer
Data & Products
Missions & Sensors
AATSR
ACE
AIRS
AMSR
AMSU
ATMOS
ATOVS/TOVS
AVHRR
AVHRR3
CALIPSO
CRISTA
GOME
GOME-2
GOMOS
HALOE
HIRDLS
IASI
MAS
MERIS
MHS
MIPAS
MISR
MLS
MODIS
MOPITT
OMI
OSIRIS
POAM
SABER
SAGE
SBUV
SCIAMACHY
SEVIRI
TES
TMI
TOMS
VIRS
WINDII
About WDC
User Account
Contact

NDMC
WDC Home









Local Weather
Oberpfaffenhofen

Daily cumulative snow cover product over Europe and Africa from MSG/SEVIRI

Daily cumulative snow cover from SEVIRI at WDC-RSAT

Description:
A daily snow cover map is derived in the SEVIRI pixel resolution (3 km nadir, 5-6 km in Europe) for the full field of view.

Product examples:
Snow Cover Europe, March 01, 2006 Snow Cover Europe, March 05, 2006 Snow Cover Europe, March 10, 2006 Snow Cover Europe, March 15, 2006
Snow Cover Europe, March 20, 2006 Snow Cover Europe, March 25, 2006 Snow Cover Europe,
March 30, 2006
 

Fig. 1: Cumulative snow cover maps for Europe in steps of 5 days for 1, 5, 10, 15, 20, 25, 30 March 2006 (daily data products are available).

 

Methodology:


In a first step, cloud masking is performed using the APOLLO/SEVIRI approach. APOLLO/SEVIRI (AVHRR processing scheme over clouds, land and ocean) is an APOLLO version specifically adapted to MSG SEVIRI data.

For each cloud free pixel, a snow indicator is calculated from reflectances in the 0.6 and 1.6 µm and from brightness temperatures in the 12 µm channels. The snow indicator is an arbitrary measure taking high reflectances in the 0.6 µm channel, low reflectances in the 1.6 µm channel, and brightness temperatures close to 273 K as an indicator for a high snow probability.

Snow cover in a pixel is assumed if this snow indicator is above a dynamical threshold. This threshold is set individually for each MSG slot based on the histogramm distribution of the snow indicator. This dynamic approach helps to identify snow cover also in non-typical weather situations (e.g. extreme cold in late spring) where thresholds based on climatological values fail.

In the next step, a daily composite is derived from all MSG slots between 10:00 and 15:00 UTC. On the composite basis, a temporal and spatial filtering is performed following ideas published by De Wildt et al. (2007). A continuity of snow cover over 3 consecutive MSG slots and the existence of a snow covered neighbouring pixel are required.

Finally, a composite product is generated

  • taking new information into account if a pixel is available on the current day
  • keeping information of the preceding day for all pixels invisible (e.g. due to cloudiness or missing satellite data).

Validation:

A comparison of snow cover for January to April 2006 (Wirth et al., 2007) based on German and Swiss meteorological stations revealed a positive hit rate of 70 % for the proper detection of existing and non-existing snow cover.

A separate analysis for both classes shows that 63% of the snow covered cases were detected correctly, while 73% of the cases without snow cover were retrieved properly.

Due to the cumulative approach only up to 1% of the days analyzed are affected by missing data due to clouds.

Users and relevance:

Daily cumulative snow cover can be used for hydrological applications as well as for energy management in hydropower and solar energy. This experimental product was created as a response to a request from the solar energy community. In solar energy plant monitoring, snow cover can be misinterpreted as cloud cover or can create a false alarm assuming a breakdown of the power plant (Wirth et al., 2007). Therefore, auxiliary information on snow cover is helpful in a satellite based plant monitoring scheme as e.g. developed in the EU FP5 project PVSAT-2 by users of WDC data.

Data access:

A major reprocessing is currently ongoing starting from February 2004 to today. Further Data is available on request.

Archive: Daily Images (GIF) and Data (HDF)

Filenames follow the convention snow_MSG_yyyymmdd_12345678_L00013712E00013712.hdf.gz

12345678 stands for a product based on all 8 MSG SEVIRI segments (default).
L00013712E00013712 marks a product which is calculated for the full SEVIRI disk from line 1 to 3712 and element 1 to 3712 (default).

Contact:

For questions and more information please contact the SEVIRI team at DLR

Publications:

De Wildt, M. de Ruyter; G. Seiz, A. GrĂ¼n, Operational snow mapping using multitemporal Meteosat SEVIRI imagery, Remote Sensing of Environment, 109, 29-41, 2007

G. Wirth, M. Schroedter-Homscheidt, M. Zehner, G. Becker, Satellite-based snow identification and its impact on monitoring photovoltaic systems, accepted by Solar Energy, 2009-11-16