|Gomez-Landesa, Enrique - NEW MEXICO STATE UNIV|
Submitted to: American Geophysical Union
Publication Type: Abstract Only
Publication Acceptance Date: November 19, 2002
Publication Date: December 6, 2002
Citation: GOMEZ-LANDESA, E., RANGO, A. IMPROVEMENTS IN SNOW MAPPING OF THE RIO GRANDE BASIN USING MODIS DATA. AMERICAN GEOPHYSICAL UNION. 2002. V. 83(47). ABSTRACT P. F540. Technical Abstract: Snow-covered area data is valuable for assessing watershed condition. Additionally, it is being used as an input for the Snowmelt Runoff Model (SRM) which provides simulations and forecasts of a basin's daily streamflow. A visible and a near infrared channels of the Moderate Resolution Imaging Spectroradiometer (MODIS) are being used to generate snow maps, replacing the traditional Advanced Very High Resolution Radiometer (AVHRR) data and, therefore, improving the spatial resolution of the snow maps from 1 km to 250 m. In order to implement the MODIS data, a geometric correction algorithm was developed to remove the so-called Bowtie Effect from the MODIS scenes. Radiances from pixels located on shaded slopes of the basin are strongly diminished because the two channels used in this method receive radiation from the solar reflective region of the spectrum. To correct this problem, an algorithm was developed to simulate the shaded areas on a Digital Elevation Model (DEM). This effect was found to be dependent on the altitude of the pixel, so an empirical, second-order polynomial was applied. In order to make the snow mapping system as automated as possible, new methods were studied to retrieve snow albedo using bidrectional reflectance distribution functions (BRDF). Albedo retrieval can also be used as a criterium to infer the start of the snowmelt season, due to the decrease in reflectance of the snowpack as it ages.