Author(s): Xiang Zhang; Yu Song

Linked Author(s):

Keywords: Fusion; Ground data; Random forest

Abstract: This paper proposed a multi-source precipitation data fusion and downscaling method named Generate high Resolution, Accurate, Seamless data using Point-Surface fusion (GRASPS). The advantages of currently several satellite/model data were fully integrated to generate a more accurate precipitation dataset at daily and 1 km scale covering the Wuhan Urban Agglomeration, including the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG) from the Global Precipitation Measurement (GPM) mission, Tropical Rainfall Measuring Mission (TRMM) Multi-satellite Precipitation Analysis (TMPA), Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks - Climate Data Record (PERSIANN-CDR). The Pearson Correlation Coefficient reached 0.77, while Root Mean Squared Error, Mean Absolute Error and Bias were reduced to 6.08 mm, 2.20 mm and -0.13 mm, respectively, under the validation of precipitation at 36 ground gauges. Compared to previous studies, this research has successfully improved the spatial resolution of precipitation dataset to 1 km and more importantly, the accuracy of extreme precipitation was specifically corrected, resulting in an accuracy increase from 76.92% to 91.67%.

DOI: https://doi.org/10.3850/iahr-hic2483430201-357

Year: 2024