Analysis of climatic extremes in the Parnaíba River Basin, Northeast Brazil, using GPM IMERG-V6 products

Abstract

Satellite products, such as the Integrated Multi-Satellite Retrievals of IMERG from the Global Precipitation Measurement (GPM) mission, have emerged as promising tools to analyze precipitation distribution and extremes, particularly in regions with low rain gauge density and sparse distribution, such as Brazil. However, regional validation of satellite data is crucial. In this context, the validation of GPM (IMERG) for the Parnaíba River Basin in northeastern Brazil is important due to its high hydrological potential and the presence of one of the largest expanding agricultural frontiers in the world. This study evaluates the estimation capacity of IMERG version 6 satellite data, including IMERG Early, Late, and Final products, for extreme precipitation in the Parnaíba River Basin from 2001 to 2020. A pixel point-to-point approach is used to compare the satellite estimates with observed precipitation data measured by rain gauges. Eight indices of extreme precipitation are analyzed, along with statistical measures such as bias, mean square error, root mean square error, probability of detection (POD), false alarm ratio (FAR), and the Kling-Gupta efficiency (KGE) index and its components. The results show that the IMERG Final estimates exhibit better agreement with in situ data at the daily scale compared to the IMERG Early and Late estimates. The lower Parnaíba region shows higher POD values, while the middle Parnaíba region exhibits higher KGE values, particularly in tropical climate areas. The IMERG products demonstrate different capabilities in observing extreme rainfall in the basin, with IMERG Final showing satisfactory results for 50 % of the analyzed indices, performing more robustly in capturing the PRCPTOT index and reasonably for CDD, RX5day, and R95p. We conclude that the IMERG Final product can be used as a data source for analyzing precipitation extremes in the Parnaíba River Basin, with bias adjustment recommended for better performance at the daily scale.