Estimating daily precipitation climatology by postprocessing high‐resolution reanalysis data

Spatial information of climatological frequency distribution daily precipitation is highly valuable for a wide range applications. Accurate estimation climatology can be made gauged locations where quality and lengthy observations are available. For ungauged or poorly locations, however, indirect needed. One approach to use gridded dataset derived from interpolating observations. However, data subject large errors when gauge density low. In addition, most interpolation methods tend smooth the extreme values increase low ones, leading unrealistic statistical properties therefore poor climatology. Another first derive at then interpolate locations. While this likely more robust than approach, still cause significant especially in areas complex terrain. study, we develop method that postprocesses spatially consistent rich reanalysis using accurate At an location, amounts bias-corrected quantile-mapping guided by distributions nearby location (reference location). The used estimate location. This eliminates need its adverse effects. Special care taken extrapolating beyond reference We evaluate 50 Australia, Bureau Meteorology Atmospheric high-resolution Regional Reanalysis Australia (BARRA) observation network across Australia. These chosen represent different climate regions have validate postprocessed precipitation. Results show with observations, terms distribution, high quantiles, probabilities wet dry days their transitions.