Evaluation of the IMERG Product in the MATOPIBA Region, Brazil

Abstract

Managing water resources around the world requires challenges imposed by natural and anthropogenic processes. Having knowledge about water resources, whether qualitative or quantitative, almost always depends on monitoring rainfall stations for their temporal and spatial understanding. Monitoring dry and wet seasons in a country or region is a difficult problem, especially when the network of meteorological stations is sparse or does not cover the entire area. The Integrated Multisatellite Retrieval of Global Precipitation Measurements (IMERG) produces and provides a source of high-resolution spatial and temporal precipitation estimates obtained by satellites and has been widely used since its launch in 2014, providing global precipitation data since 2000. In the present study, the performance of the GPM (Global Precipitation Measurement) product (IMERG-final run) was evaluated in the MATOPIBA region, located in the northern and northeastern regions of Brazil, from October 2000 to September 2021. The monthly and annual IMERG estimates were compared against the rainfall data from 197 ground stations in the study region belonging to the Agência Nacional das Águas (ANA). The metrics used for this study's analysis of IMERG's performance in detecting rainfall were both quantitative and qualitative. The IMERG estimates overestimated those of the rain gauges by about 13%. However, the performance result was quite satisfactory for the estimation of rainfall events, both monthly and annually, that is, excellent correlation with their spatial and temporal distribution as well as their volume. It was concluded that this product is a satisfactory source of alternative rainfall data, especially in areas without a dense network of rain gauges, such as the study region.