Evaluation of drought utility assessment of TMPA satellite-remote-sensing-based precipitation product in mainland China

Abstract: Satellite-remote-sensing-based precipitation analysis products with high spatiotemporal resolution provided a new data source for the studies of drought. The aim of this study is to evaluate the drought assessment utility of the TMPA (tropical rainfall measurement mission (TRMM) multi-satellite precipitation analysis) 3B42V7 post-real-time satellite-based precipitation product in mainland China. We evaluate the accuracy of 3B42V7-based drought indices, the capability of the 3B42V7 product to reveal the spatiotemporal variation of drought, and the capability to identify the drought events in 1998－2015 based on 2 widely used meteorological drought indices, i.e. SPI (standardized precipitation index) and SPEI (standardized precipitation evapotranspiration index), with CGDPA (China gauge-based daily precipitation analysis) as the evaluation reference, which is an interpolated gauge-based precipitation dataset. The potential evapotranspiration (PET) used to calculate the SPEI is calculated by meteorological data of 818 meteorological stations and 129 radiation stations among the mainland China and interpolated with the Kriging method. Considering that the sensitivity of the drought indices to the length of base period might affect the evaluation, the CGDPA-based drought indices are calculated in the periods with 2 lengths: the period of 1998－2015 (18 years, CGDPA18a), the same period of the 3B42V7 observation history, and a longer period of 1960－2015 (56 years, CGDPA56a). The results show that: 1) SPI and SPEI based on 3B42V7 product generally show acceptable accuracy over mainland China except the mid-west and south Tibetan Plateau, with high correlation coefficients over 0.8 in most of the east part of mainland China with those of both CGDPA18a and CGDPA56a. 2) Evaluation of the spatial and temporal variation of drought focusing on southwestern China (SWC) and northern China (NC), the 2 typical drought affected regions in China, shows that the time series of the area-averaged drought indices, and the spatial pattern of drought indices derived from 3B42V7 have high consistency with both the CGDPA18a and the CGDPA56a, indicating that 3B42V7 can reflect the evolution process and intensity of drought events, capture the spatial pattern of drought events, and reveal the spatiotemporal evolution of drought reasonably. In addition, the 3B42V7-based SPI is generally comparable to the SPEI in many cases, indicating that the SPI calculated only from the 3B42V7 data has already met the requirement of drought monitoring in the non-arid region of mainland China. 3B42V7 can generally accurately catch the moving path of the centroid of drought from September 2010 to October 2011 in eastern China, which indicates that the 3B42V7 can reasonably reveal the spatiotemporal variation of drought. 3) 3B42V7 product is capable of recognizing the large-scale drought events that took place during 1998－2015 and finding out their spatio and temporal positions truly and reasonably. Although some differences of the droughts affected area and severity of the drought events were found between those calculated from 3B42V7 data and those of CGDPA data, 3B42V7 product can reflect the temporal and spatial evolution of droughts in mainland China. In conclusion, the 3B42V7 product is suitable for monitoring and evaluation of meteorological drought at large scale.