Diurnal Variability of Global Precipitation: Insights from Hourly Satellite and Reanalysis Datasets

Accurate estimation of precipitation at the global scale is of utmost importance. Even though satellite and reanalysis products are capable of providing high spatial-temporal resolution estimations at the global level, their uncertainties vary with regional characteristics, scales, and so on. The uncertainties among the estimates, in general, are much higher at the sub-daily scale compared to daily, monthly and annual scales. Therefore, quantifying these sub-daily estimations is of specific importance. In this context, this study seeks to explore the diurnal cycle of precipitation using all the currently available space-borne and reanalysis-based precipitation products with at least hourly resolution (IMERG, GSMaP, CMORPH, PERSIANN, ERA5) at the quasi-global scale (60^◦N - 60^◦S). The diurnal variability of precipitation is estimated using three parameters, namely, the precipitation amount, frequency, and intensity, all remapped at a common resolution of 0.25^◦ and 1 h. All the estimates well represent the spatio-temporal variation across the globe. Nevertheless, considerable uncertainties exist in the estimates regarding the peak precipitation hour, as well as the diurnal mean precipitation amount, frequency, and intensity. In terms of diurnal mean precipitation, PERSIANN shows the lowest estimates compared to the other datasets, with the largest difference observed over the ocean rather than over land. As for diurnal frequency, ERA5 exhibits the highest disparity among the estimates, with a frequency twice as high as that of the other estimates. Furthermore, as expected being based on model reanalysis, ERA5 shows an early diurnal peak and the highest variability compared to the other datasets. Moreover, among the satellite estimates, IMERG, GSMaP, and CMORPH exhibit a similar pattern with a late afternoon peak over land and an early morning peak over the ocean.