Similarities and improvements of GPM IMERG upon TRMM 3B42 precipitation product under complex topographic and climatic conditions over Hexi region, Northeastern Tibetan Plateau

Abstract

Building upon the success of TRMM launched in 1997, NASA and JAXA successfully deployed the Global Precipitation Measurement (GPM) mission on February 28, 2014. GPM started to release the new-generation global precipitation products Integrated Multi-satellite Retrievals for GPM (IMERG) since April 2014. Critical data validation over diverse topographic and climatic regions could effectively help users and algorithm developers maximize the accuracy, strengths and weaknesses of the new product. To this end, this study comprehensively evaluated and compared this newly released precipitation product (IMERG V05B) and its predecessor TRMM 3B42V7 based upon the ground-based observations under complex topographic and climatic conditions over the Hexi Region in the northwest arid region of China. The evaluation was conducted at daily, monthly, seasonal and annual scales from April 2014 to September 2017. Results indicated that: (1) compared to ground-based observations, both IMERG and 3B42V7 showed good performance with slightly overestimation (with RB values of 4.58% and 2.27%, respectively) at basin scale. While at the point-to-pixel scale, both products showed overestimation in low-elevation regions and underestimation in high-altitude mountainous areas; (2) on multiple temporal and spatial scales, IMERG correlated better with ground-based observations than 3B42V7, whereas its bias were larger than 3B42V7; (3) both IMERG and 3B42V7 showed better performance in warm seasons (summer and autumn) than in cold seasons (spring and winter); (4) Compared to 3B42V7, IMERG represented both larger hit rate (POD) and false alarm ratio (FAR), but similar correct detection rate (CSI); (5) IMERG showed poorer detection capabilities in light precipitation events (0–2 mm/day), while represented a certain advantage in detecting moderate (5–10 mm/day) and heavy (>25 mm/day) rain events. Most of precipitation estimates over Hexi region are unreliable on daily and monthly time scales. Results suggest that IMERG do not show significant improvement compared to its predecessor 3B42V7, and performed even worse in estimating precipitation amounts. Further data bias-correction is needed before IMERG can be used for climatic and hydrologic research in the Hexi region.