Global daily 1km land surface precipitation based on cloud cover-informed downscaling

TL;DR

This paper introduces a semi-mechanistic downscaling method that uses satellite-derived cloud cover and topographic predictors to generate high-resolution daily precipitation data, improving accuracy over existing products.

Contribution

The study develops a novel downscaling approach integrating cloud cover, orographic factors, and bias correction to produce detailed daily precipitation maps at 1km resolution.

Findings

Improved spatio-temporal accuracy over ERA5

Better performance in complex terrain

Enhanced representation of precipitation patterns

Abstract

High-resolution climatic data are essential to many applications in environmental research. Here we develop a new semi-mechanistic downscaling approach for daily precipitation that incorporates high resolution (30 arc sec) satellite-derived cloud frequency. The downscaling algorithm incorporates orographic predictors such as wind fields, valley exposition, and boundary layer height, with a subsequent bias correction. We apply the method to the ERA5 precipitation archive and MODIS monthly cloud cover frequency to develop a daily gridded precipitation time series in 1km resolution for the years 2003 onward. Comparison of the predictions with existing gridded products and station data indicates an improvement in the spatio-temporal performance of the downscaled data in predicting precipitation. Regional scrutiny of the cloud cover correction from a topographically highly heterogeneous area further confirms that CHELSA-EarthEnv performs well in comparison to other precipitation products such as numerical weather models. The presented CHELSA-EarthEnv daily precipitation product improves the temporal accuracy compared to ERA5 with an additional improved in spatial accuracy and much better representation of precipitation in complex terrain