Validação dos dados de precipitação estimada pelo CHIRPS na região da Sealba entre os anos de 1986 e 2012

Abstract

In recent decades, remote sensing techniques for capturing rainfall data have become more common, due to improvements and developments in satellites, increasing the amount of information on rainfall estimates. Therefore, the objective of the present work is to validate the values of rainfall in the daily, monthly and annual time scales, estimated by the product Climate Hazards Group InfraRed Precipitation with Stations (CHIRPS), in the period from 1986 to 2012. Precipitation estimated by six meteorological stations of the National Institute of Meteorology (INMET) inserted in the Sealba region. The statistical indicators used were Pearson’s correlation coefficient (r), mean absolute error (EMA) and root mean square error (REQM). The r values were 0.36, 0.86 and 0.86, on the daily, monthly and annual scale, respectively. For the annual scale, the best indices were obtained with EMA and REQM equal to 204.27 mm (33%) and 261.89 mm (47%), respectively. We can conclude that CHIRPS data can be used in the SEALBA region, preferably on monthly and annual timescales.