Spatial Evaluation of PERSIANN-CDR model over Fars Province, Iran

High resolution and global coverage of satellite -based precipitation data have been found useful in many climate and hydrological studies, particularly in limited and non-gauged areas. Due to systematic and non-systematic factors, there are always deviation between the ground-based and the satellite-based precipitation. Among many satellite-based precipitation products PERSIANN-CDR (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks-Climate Data Record) is the most advanced one. In this paper, we evaluated the PERSIANN-CDR monthly precipitation over Fars Province at pixel and provincial scales during a 13-year period (2003-2015) using 132 rain-gauges. The performance of PERSIANN-CDR was evaluated by calculating the correlation coefficient (CC) and the coefficient of determination (R2) and the error indices: Root Mean Square Error (RMSE) and Relative Mean Error (RME). The result shows that in both scales, PERSIANN-CDR underestimates the monthly-precipitation in the province. An exception is observed in pixel scale in the low-rainfall regions of the province that the precipitation is overestimated. Values of CC and R2 indicate that the monthly PERSIANN-CDR precipitation from pixel to provincial scales is more linearly correlated to the ground-based precipitation and hence PERSIANN-CDR estimates are more accurate as the spatial scale increases. In other words, the performance of PERSIANN-CDR is the best at the provincial scale. The error indices also resulted in the same conclusion. Finally, an equation is proposed to predict precipitation in Fars Province from the PERSIANN-CDR precipitation data, particularly in non-gauged area.

کلیدواژه ها:

PERSIANN-CDR ، Satellite-based precipitation ، spatial evaluation