EVALUATION OF THE TEMPORAL ACCURACY OF DIFFERENT SATELLITE-BASED ESTIMATES OF PRECIPITATION IN ANAHEIM, CALIFORNIA

Satellite weather data play a significant role in hydrological studies and provide an important source of continuous precipitation data that can be used to study regions without ground-based precipitation measurements. The high temporal resolution, comprehensive spatial coverage, and availability of satellite data are significant advantages of satellite-based precipitation estimates. However, there is still significant uncertainly about the accuracy of these data. In this research, we analyze the accuracy of different satellite-based precipitation methods, including Global Precipitation Measurement (GPM), Multi-satellite Precipitation Analysis (TRMM 3B42RT and TRMM 3B42), Global Satellite Mapping of Precipitation (GSMaP), Climate Forecast System Reanalysis (CFSR), Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN, PERSIANN-CCS and PERSIANN-CDR) over a time period ranging from 2013 to 2016. The study area is located in Anaheim, California, and has ten ground-based weather stations. Statistical analyses between ground-based and satellite precipitation data show that daily correlations between the two data types are typically poor. While no satellite data are accurate on the daily scale, GSMaP and CFSR show the best correlation with ground-based data. Satellite data on a monthly scale have much better correlation with ground-based measurements, but have an obvious lack of temporal resolution. Analysis of data on a weekly and bi-weekly basis is ongoing; these analyses will provide a better understanding of the trade-off between temporal resolution and accuracy for satellite-based precipitation estimates.