An assessment of combination strategies for passive microwave/infrared data integration

Research into the combination of passive microwave and infrared data is currently receiving much attention: The ability of the passive microwave to delineate and estimate instantaneous rainfall has been shown by numerous intercomparison projects. Infrared data benefits from better temporal and spatial sampling, and therefore a combination of these two data sources holds the promise of much improved temporal and spatial rainfall estimates. This paper presents results of research into the strategies that could be used in combining passive microwave and infrared data. The results are based upon data used in the third Algorithm Intercomparison Project (AIP-3) over the TOGA-COARE region of the Western Pacific. Temporal and spatial combination techniques are applied to the infrared data using rainfall estimates derived from three passive microwave algorithms and shipborne radar data, the latter providing the best-case scenario. The calibrated infrared data is then used to generate results at differing spatial and temporal resolutions. Results derived from the Global Precipitation Index (GPI) are also included for comparison. Validation of these results indicate that significant variations occur in the temporal thresholds of the infrared data with these threshold rarely applicable beyond 18 hours suggesting a diurnal influence.

These results are further investigated by using data from the Tropical Rainfall Measuring Mission (TRMM): Data from the TRMM Microwave Imager (TMI) and from the Visible and Infrared Scanner (VIRS) are analysed for 12 months over Africa to investigate the variations of the infrared temperatures with rainfall intensities. This data set is augmented by the analysis of data over Amazonia using GOES-E, VIRS and PR data. Initial results indicate a significant variation both spatially and temporally in the infrared threshold, suggesting that any long-term set thresholds and rainfall intensities would be inappropriate to better temporal and spatial resolution global rainfall products.