Monitoring Areal Snow Cover Using NASA Satellite Imagery
- Created on Thursday, 01 September 2011
The objective of this project is to develop products and tools to assist in the hydrologic modeling process, including tools to help prepare inputs for hydrologic models and improved methods for the visualization of stream-flow forecasts. In addition, this project will facilitate the use of NASA satellite imagery (primarily snow cover imagery) by other federal and state agencies with operational streamflow forecasting responsibilities.
A GIS software toolkit for monitoring areal snow cover extent and producing streamflow forecasts is being developed. This toolkit will be packaged as multiple extensions for ArcGIS 9.x and an opensource GIS software package. The toolkit will provide users with a means for ingesting NASA EOS satellite imagery (snow cover analysis), preparing hydrologic model inputs, and visualizing streamflow forecasts. Primary products include a software tool for predicting the presence of snow under clouds in satellite images; a software tool for producing gridded temperature and precipitation forecasts; and a suite of tools for visualizing hydrologic model forecasting results. The toolkit will be an expert system designed for operational users that need to generate accurate streamflow forecasts in a timely manner.
The Remote Sensing of Snow Cover Toolbar will ingest snow cover imagery from multiple sources, including the MODIS Operational Snowcover Data and convert them to gridded datasets that can be readily used. Statistical techniques will then be applied to the gridded snow cover data to predict the presence of snow under cloud cover. The toolbar has the ability to ingest both binary and fractional snow cover data. Binary mapping techniques use a set of thresholds to determine whether a pixel contains snow or no snow. Fractional mapping techniques provide information regarding the percentage of each pixel that is covered with snow. After the imagery has been ingested, physiographic data is attached to each cell in the snow cover image. This data can be obtained from a digital elevation model (DEM) for the area of interest. If the snow cover image contains cloud cover, regression tree analysis is used to predict the presence of snow cover under clouds.
The Gridded Temperature and Precipitation Forecast Toolbar will ingest forecasts from numerical weather prediction models and produce gridded forecasts that can be used as input for distributed hydrologic models. This toolbar will enable users to easily produce gridded fields of temperature and precipitation from location-specific forecasts, which is needed since a majority of hydrologic models are run on a distributed basis. This is completed using temperature data, and will be expanded in the future to include precipitation data.
The Streamflow Forecast Visualization Toolbar will generate visualizations of streamflow forecasts. Outputs include a variety of tables, charts, and figures depicting streamflow forecasts in formats that can be easily interpreted by the general public.
The interpolation process entails: (1) obtaining a DEM for the watershed (basin) of interest; (2) obtaining temperature (forecasted or observed) and elevation values for an individual weather station (base station) located within the watershed; and (3) applying the monthly temperature lapse rates to create gridded values. After a DEM is selected for the area of interest, the GIS tools essentially complete the interpolation process for any specified day automatically. Tools are included to assist in the validation of the forecast grids.
This work was done by Brian J. Harshburger, Troy Blandford, and Brandon Moore of Aniuk Consulting, LLC, for Goddard Space Flight Center. GSC-15791-1