Les images de satellite et les mesures de la couche de neige

Abstract

This partly bibliographical study starts with a review of satellites transmitting useful data for snow surveys and draws attention to difficulties experienced in converting raw data into usable information (e.g. extent of snow-covered areas). Examples are then described of reports published by mostly North American authors, in which satellite data are related to hydrological variables. Most of the information originates from a NASA seminar at South Lake Tahoe in August 1975 and a WMO seminar in Geneva in October 1976, on the subject of surveys of snow cover for hydrological forecasting. This study ends with a discussion of present use of ground-based measurements by de for forecasting seasonal precipitation supplies and the types of additional service teledetection could provide. It seems unlikely that teledetection from satellites will ever replace ground-based measurements as a means of evaluating the water equivalent of the snow burden in the Alps or Pyrenees while it is accumulating in winter, i.e. calculating probable snow melt inflows sufficiently far ahead. Satellite pictures of snow-covered areas do not become representative of the actual storage volume until several weeks after the snow starts to melt, i.e. not before the 15th May in the Alps, whereas Electricite de France forescasts from ground-based measurements become effective as from the 1st February. The economic importance of such advance notice for water power storage management is the reason why such considerable efforts are being made to automate daily snow depth telemeasurements (the remote-indicating snow gauge with a moving horizontal beam provides a complete decimetre-by-decimetre density profile). In future, and on the assumption that the frequency and accuracy of satellite pictures will substantially improve, teledetection should provide an effective means of improving short-term forecasts by analysis in the visible and infra-red ranges of the fringe areas of the seasonal snow cover while it is melting : contours, albedo, surface temperature. Hyperfrequency teledetection at wavelengths from 1 mm to 300 mm also seems to offer reasonably hopeful prospects; such frequencies pass through clouds and can provide information on the depth of dry snow, but are reflected by wet snow. With existing satellites the only application of teledetection capable of providing useful information, for observation of the short-term behaviour of floods originating from snow and/or rain, is the daily picture of snow-covered areas without pronounced relief features between altitudes of 800 m and 1200 m in the Massif Central and Jura mountains from NOAA satellites on a polar orbit, and perhaps also METEOSAT.