Comprehensive precipitation evaluation of TRMM 3B42 with dense rain gauge networks in a mid-latitude basin, northeast, China

Knowledge on understanding quality of The Tropical Rainfall Measuring Mission (TRMM) 3B42 V7 dataset over mid-high latitudes regions is limited, which restricts its potential application in climate and hydrology fields. This study focuses on giving a detailed evaluation of the accuracy of 3B42 with observation data obtained from a high density rain gauge network over the Hun-Tai Basin in Liaoning Province, northeast China during 1998-2006. Several accuracy statistics are used to evaluate it quantitatively in terms of error of precipitation amount and ability in detecting the occurrence of precipitation events. Comparative results for three timescales (daily, monthly, and annual scale) at the basin scale show that 3B42 is more suitable for analyzing precipitation at large timescale, especially monthly scale (strong correlation of 0.93) due to the use of monthly rain gauge observation for bias correction in producing 3B42. Yet, 3B42 generally overestimates precipitation at all three timescales, especially the most serious degree of overestimation at daily scale with the absolute bias of 123.94 % and light to moderate rain events (1-20 mm). Moreover, the performance is influenced by topography, and 3B42 has a larger error of precipitation amount but has a better detection of the occurrence of precipitation events over high-altitude region than those over low-altitude region. Also, accuracy of 3B42 decreases with precipitation intensity, it suggests that 3B42 is incapable of capturing heavy precipitation events with desirable accuracy for the study on extreme precipitation events. In following works, overestimation characteristic should be weaken by improving satellite-based precipitation estimations algorithms and developing more effective bias correction techniques, it is important for streamflow simulations using 3B42 as forcing data over ungauged regions.