FUEL MOISTURE CONTENT BASED ON MODIS-DERIVED VEGETATION INDICE IN WEST TEXAS HIGH DESERT

In fire behavior analysis and fire danger rating systems, vegetation is an important component. Vegetation density, vegetation type and vegetation moisture are all important aspects of a fire analysis model. Both dead and live fuel moistures are calculated for these models. Currently, live fuel moisture of vegetation is calculated mathematically using data models. The calculation takes information such as temperature, humidity, precipitation, wind speed, barometric pressure, and wind direction from Remote Automated Weather Stations (RAWS). Some stations also have a dowel that tests capacitance to determine dead fuel moisture. Together, this data can be input into a model such as FireFamilyPlus to determine fire behavior components such as live fuel moisture. This data has limitations in that the RAWS stations are located at large distances and the accuracy for the station diminishes over distance, elevation and vegetation changes. Satellite imagery provides the ability for vegetation analysis at a larger spatial scale. Various vegetation indices will be used to analyze the spectral reflectance of vegetation from Moderate Resolution Imaging Spectradiometer (MODIS) imagery. MODIS product MOD9A1 is an 8 day composite to produce atmospherically corrected and nearly cloud free data at 500m resolution. This product was tested to determine its ability to derive live fuel moistures in the high desert of West Texas. Live fuel moisture from a RAWS station located in Fort Davis Mts. will be regressed with various vegetation indices derived from the MODIS (tile H09V05) imagery covering the study area in Fort Davis over the time period of February 2000 to February 2004. Vegetation indices of 1x1, 3x3 and 5x5 pixel size centered at the RAWS station will be analyzed to determine both the sensitivity of RAWS' fuel moisture content and its correlation with vegetation indices at different spatial scales. Then, the ability of MODIS to predict fuel moistures for the 2004 fire season will be tested to determine the ability of these indices to be used as a predictive tool for input to fire prediction models.