Roberts, DA, Dennison, PE, Peterson, S, Sweeney, S, Rechel, J (2006). Evaluation of airborne visible/infrared imaging spectrometer (AVIRIS) and moderate resolution imaging spectrometer (MODIS) measures of live fuel moisture and fuel condition in a shrubland ecosystem in southern California. JOURNAL OF GEOPHYSICAL RESEARCH-BIOGEOSCIENCES, 111(G4), G01S02.
Dynamic changes in live fuel moisture (LFM) and fuel condition modify fire danger in shrublands. We investigated the empirical relationship between field-measured LFM and remotely sensed greenness and moisture measures from the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) and the Moderate Resolution Imaging Spectrometer (MODIS). Key goals were to assess the nature of these relationships as they varied between sensors, across sites, and across years. Most AVIRIS-derived measures were highly correlated with LFM. Visible atmospherically resistant index (VARI) and visible green index (VIg) outperformed all moisture measures. The water index (WI) and normalized difference water index (NDWI) had the highest correlations of the moisture measures. All relationships were nonlinear, and a linear relationship only applied above a 60% LFM. Changes in the fraction of green vegetation (GV) and nonphotosynthetic vegetation (NPV) were good indicators of changes in fuels below the 60% LFM threshold. AVIRIS-and MODIS-derived measures were highly correlated but lacked a 1: 1 relationship. MODIS-derived greenness and moisture measures were also highly correlated to LFM but generally had lower correlations than AVIRIS and varied between sites. LFM relationships improved when data were pooled by functional type. LFM interannual variability impacted relationships, producing higher correlations in wetter years, with VARI and VIg showing the highest correlations across years. Lowest correlations were observed for sites that included two different functional types or multiple land cover classes (i.e., urban and roads) within a MODIS footprint. Higher correlations for uniform sites and improved relationships for functional types suggest that MODIS can map LFM effectively in shrublands.