Evaluation of 3B42V7 and IMERG Daily-Precipitation Products for a Very High-Precipitation Region in Northwestern South America


The availability of water is critical in determining the distribution of species by favoring or limiting their development, and leading to the formation of different ecosystems. Thus, analyzing the trends and fluctuations of the precipitation is a key factor to understanding our planet’s biodiversity. Different physical conditions exist in northwestern South America that cause extreme climate conditions, especially on the west coast of the continent. This region, known as Biogeographic Chocó, has high annual precipitation caused by interactions between the humid currents of the Pacific Ocean and the Andes Mountains. A limited network of hydro-meteorological stations are available in the region to monitor precipitation. Satellite precipitation products can provide valuable information in the absence of field stations, complementing the network of field stations in remote areas, and completing time series for stations that have stopped working. However, remote sensing data must be validated before being used. The goal of this study is thus to evaluate the accuracy of the products of the Tropical Rainfall Measuring Mission 3B42V7 and the Integrated Multi-satellite Retrievals for Global Precipitation Measurement (IMERG), under very high precipitation conditions, and evaluate their strengths and shortcomings. These programs provide daily-precipitation information with a spatial resolution of 0.25° for the former and 0.1° for the latter. The validation was done by using a time series of daily-precipitation data obtained from 185 hydro-meteorological stations distributed over the Biogeographic Choc.́ Different statistic metrics were used in the evaluation and comparison: error metrics (mean difference MD, relative mean difference RMD, and root mean square error RMSE), a correlation metric (Pearson correlation CP), contingency metrics (probability of detection POD, false-alarm ratio FAR, and critical success index CSI). We also evaluate the grid and areal scale. The results show that (i) the 3B42V7 and IMERG daily-precipitation products represent well the spatial and temporal distribution of mean daily precipitation over the Biographic Choc’ ́and both products are accurate for detecting precipitation events. (ii) Mean daily precipitation tends to be overestimated in areas with relative low precipitation and medium-to-high altitude whereas, on the contrary, mean daily precipitation tends to be underestimated in areas with very high precipitation and medium-to-low altitude. (iii) Finally, copious precipitation (i.e., an annual accumulated precipitation over 5000 mm, which is common for over 55% of the study area) strongly affects the accuracy of the satellite products, leading to significant errors in estimates of daily precipitation for some regions. This study constitutes one of the first exhaustive validation of the IMERG daily precipitation product over the Biogeographic Choc’oánd the results provide important information about the potential for using this product in the study area and over regions with high precipitation.

Atmospheric Research