Statistical evaluation of gridded precipitation datasets using rain gauge observations over Iran

摘要

The present study assessed the accuracy of twenty-three available globally and regionally gridded precipitation products over Iran. The datasets were classified into five different categories based on the data sources used: gauge-only, gauge-satellite, gauge-reanalysis, reanalysis-only, and gauge-satellite-reanalysis. By considering monthly rain gauge observation data as reference, the above precipitation products were assessed based on a variety of metrics over the period 1979 to 2013. The results indicated that the time-series of the spatial averaged precipitation of all the products do reasonably well in tracking the month-to-month variations with the majority of the products underestimating precipitation amounts with respect to the reference rain gauge observations. Overall, the best performances were demonstrated by Asfazari and APHRODITE followed by CHIRPS and GPCC. WFDEI-CRU was the worst performer among the twenty-three gridded products in revealing precipitation tempo-spatial patterns. The results demonstrate the high performance for all products in the west of the country (the Zagros Range) and in all seasons except the summer. The results also indicate that there is a better agreement between rain gauge observations and the gridded products during the rainy seasons (winter, spring and autumn) compared to the dry season (summer). This study improved our knowledge of the suitability of different precipitation gridded product estimates to capture the temporal and spatial variability of precipitation. Furthermore, it is applicable for developing precipitation products and provides valuable guidance for selecting alternative precipitation products to in situ observations.