Evaluating the utility of the ANSA blended snow cover product in the mountains of eastern Turkey
Abstract
Snow-covered area depletion curves represent a key input for snow run-off melting models, e.g. the snowmelt run-off model (SRM). SRM is a degree-day-based model for daily run-off simulations and forecasts in mountainous areas in which snowmelt is the major run-off contributor. Satellite images and aerial photographs are valuable sources for retrieving snow-covered area. The accuracy of snow cover mapping studies in the optical wavebands is highly dependent upon the algorithm's ability to detect clouds. On very cloudy days it is not possible to map snow cover using only optical sensors; however, microwave sensors can be used to obtain snow information on cloudy days. The snow-water equivalent (SWE) of a dry snowpack can be estimated with passive-microwave sensors such as Special Sensor Microwave/Imager (SSM/I) and Advanced Microwave Scanning Radiometer for EOS (AMSR-E). Development of snow cover products based on multi-sensor data sources is needed for continuous regional and global snow cover mapping for climate, hydrological and weather applications. A preliminary blended snow product has been developed jointly by the US Air Force Weather Agency (AFWA) and NASA/Goddard Space Flight Center. The AFWA-NASA Snow Algorithm, or ANSA, blended snow product is an all-weather product that utilizes both visible and near-infrared (Moderate Resolution Imaging Spectroradiometer, MODIS) and microwave (Advanced Microwave Scanning Radiometer-Earth, AMSR-E) data. In this study the validation of the ANSA blended snow cover product, having 25 and 5 km resolution, respectively, was performed for the eastern part of Turkey for five months in the winter of 2007-2008. This is the first time that the ANSA snow cover product has been evaluated in a mountainous area, where the elevation ranges between 850 and 3000 m. Daily snow data collected at 36 meteorological stations were used in the analysis. Use of the ANSA snow products was found to improve the mapping of snow cover extent relative to using either MODIS or AMSR-E products alone, for the 2007-2008 winter in the eastern part of Turkey. 91% agreement was obtained between the ANSA snow maps and in situ observations for February. The lowest agreement percentage of 68% was obtained for March due to shallow snow depth and wetness of the snow. Change in the spatial resolution of the ANSA product from 25 km to 5 km increased the agreement percentages from 68% to 74% for March. ANSA prototype maps of 5 km resolution from February and March 2008 were used to derive snow cover depletion curves for the upper Euphrates basin located in the eastern part of Turkey. The results were compared with the curves obtained from MODIS daily snow products, and found to provide an improvement over using MODIS daily maps alone. This is because the ability of the microwave sensors to map snow through clouds provides snow cover information on cloudy days when the MODIS maps cannot, though at a coarser spatial resolution than can be obtained using MODIS.