Inter-Research >  > Prepress Abstract

CR prepress abstract   -  DOI: https://doi.org/10.3354/cr01610

Accuracy analysis of GPM IMERG and CMORPH precipitation data over North China

Lingling Shen*, Runsheng Lin, Li Lu, Chong Xu, Yu Liu

*Corresponding author:

ABSTRACT: Large-scale agricultural production in North China makes the study of precipitation in this area vital. The performance of IMERG and CMORPH precipitation products for 2015 was evaluated against daily precipitation data from 404 rain gauges in North China. Relative errors, correlation coefficients, Pearson’s chi-squared test values, and root mean square errors, as well as the probability of detection (POD), false alarm ratio, and critical success index, were used to analyze the accuracy of both IMERG and CMORPH precipitation products on daily, monthly, and seasonal timescales. The probability density function (PDF) was also considered. Overall, both products overestimated ground precipitation, especially in summer. Positive correlation coefficients between satellite-derived and rain-gauge monthly precipitation data were higher over plains and coastal areas, compared with plateau regions. The PODs of both IMERG and CMORPH data were highest in summer. The PODs of IMERG data were much higher than for CMORPH data in autumn. The PODs over coastal regions, plains, and plateaus at lower latitudes also were considerably better than over inland and plateau areas at higher latitudes. The precipitation products performed best over coastal areas, plains, and areas with high rainfall. Both CMORPH and IMERG products were prone to identify non-rainy days as rainy days. They also overestimated the light (0.1–9.9mm/day) and moderate (10–24.9 mm/day) precipitation events, although the IMERG product was more sensitive to precipitation. Accordingly, we believe that both these satellite-derived precipitation products require further modification to enable them to substitute for gauge precipitation data in North China.