Land surface response to precipitation events using MODIS and NEXRAD data

We have developed a wavelet-based information theoretic approach to examine the interaction between precipitation (PPT) forcing events and the land surface response. Combining Next Generation Weather Radar (NEXRAD) PPT with Moderate Resolution Imaging Spectroradiometer (MODIS) vegetation (NDVI) and surface temperature (T-s) data over the Missouri Basin in the north-central USA, we are able to address the spatial and temporal fluctuations surrounding the hydrometeorology of grassland ecosystems. Information theory metrics of entropy and mutual information content are combined with a wavelet multi-resolution analysis to examine to what extent the observed PPT signal directly determines the spatial distribution of the land surface temperature and vegetation and how this relationship varies with spatial scale. Results indicate that (1) there is a reduction in the temporal variance of the wavelet coefficients as the signal is transferred from the PPT into the surface temperature and finally the vegetation signal, (2) there are significant correlations as a function of spatial resolution between PPT-NDVI and PPT-T-s signals which generally increase with spatial resolution, while there is little correlation between the NDVI and T-s signals as a function of resolution, and (3) the scale-wise entropy and the mutual information content of the signals increase for all fields as the spatial resolution increases. This provides a methodology for determining the relative impact of regional climatology and local land-atmosphere interactions as a function of spatial scale.