资源环境科技发展态势分析平台

Understanding the performance of remotely-sensed precipitation with rain gauge as reference is essential before their application in estimating multiple processes. In current study, the Global Precipitation Measurements from Integrated Multi-satellite Retrievals (GPM-IMERG) and Tropical Rainfall Measuring Mission (TRMM-3B42) precipitation products were evaluated, taking the rain gauge data as a reference for the period of 2004 to 2018 over Pakistan. The data at daily, monthly, annual and seasonal timescales were evaluated using multiple statistical (three continuous and four categorical) error metrics. The Satellite Precipitation Products (SPPs) showed consistent precipitation climatology with some regional differences. All SPPs overall performed well in the central and southern regions compared to the northern region of study area. Evaluation revealed that the performance of SPPs was influenced by topography, while the error characteristics of SPPs were dependent on the precipitation intensity and climatic condition of the region. The SPPs were able to capture the extreme precipitation events (95th percentile) across different climatic regions. The highest values for probability of detection (POD), critical success index (CSI), and accuracy (ACC) at the rate of 0.7, 0.4, and 0.9 were observed respectively, while the lowest value for false alarm ratio (FAR) is 0.4. The IMERG final run harmonizes rain gauges very closely in plain and medium elevation regions, while root mean square difference greater than 20 mm/month were observed for high altitude. The spatial distribution of correlation coefficients with the values greater than 0.7 suggest that selected SPPs well captured the seasonality and mutual comparison shows that IMERG final run performed well, followed by TRMM-3B42, whereas in winter and pre-monsoon, the IMERG early run and IMERG late run captured higher precipitation values in the northeast of the study area. The results can help the researchers to select reliable SPPs as an alternate source of rain gauge in the country and other adjoining regions and the data evaluation information can help algorithm developers to rectify the errors and improve the selected SPPs.