Development of a new global rainfall rate model based on ERA40, TRMM, GPCC and GPCP products

This paper presents the study carried out to improve the accuracy of the rainfall rate prediction model in Rec. ITU-R P.837-4. By combining different precipitation data sources, new global maps of the model input parameters have been calculated, which guarantee both higher accuracy and spatial resolution. An optimization of model coefficients has been performed by considering cumulative distribution functions (CDFs) of measured rainfall rate, provided by DBSG3, selected by taking into account their quality, worldwide distribution and statistical stability in terms of years of measurements. The results show that the prediction of the mean annual rainfall rate distribution is improved of ~10% and ~25% in terms of bias and root mean square of the relative estimation error, respectively. An attempt in estimating intra-annual rain rate has also been carried out, showing promising results that call for a larger collection of measured monthly rainfall rate CDFs.