Rainfall mapping in a Region of Northeastern Brazil using geostatistical techniques

Abstract

In the northeast of Brazil, the history of droughts and the high variability of rainfall generate significant economic losses, which leads to a demand for climate studies that can subsidize actions that minimize impacts caused by intense rainfall events. The objective of this study was to investigate the spatial dynamics of precipitation in the state of Alagoas, which stands out as a major producer of sugarcane in the northeast region. For that, variogram models were used: Exponential, Spherical and Gaussian, in addition to the kriging technique. From 20 pluviometric stations, the average index was above 120 mm for the months of April to August for a period of five years (2014-2018), moreover, these same months presented a coefficient of variation greater than 65%, indicating high spatial variability of the studied variable. The mesoregion that shows the highest number of precipitated water is the East Alagoano, while the Sertão Alagoano shows the lowest precipitation rate throughout the year. Through the map of the prediction error, it was observed that, even the Sertão Alagoano representing a mesoregion called drought, it is possible to have a high rate of precipitation, in the same way the East Alagoano can experience droughts, based on the accumulated value annual precipitation. The techniques used proved to be useful for mapping rainfall and describing the variable of interest in its main aspects and among the models considered, the one that best fitted the data was the Gaussian.