Geostatistical modeling of the rainfall patterns and monthly multiscale characterization of drought in the South Coast of the Northeast Brazilian via Standardized Precipitation Index

Abstract

Variations in rainfall patterns in the Northeast region of Brazil (NEB) are high and multiscalar, increasing susceptibility to extreme drought and/or flood events. The objective of this study was to characterize rainfall patterns and monthly dry-wet periods using the Standardized Precipitation Index (SPI) from 1990 to 2019 in the South Coast of NEB, utilizing geostatistical interpolation methods. The study was based on a climatological dataset from the coastal region of the state of Bahia, collected from 112 weather stations. A map projecting aquifer in the study area was established, and SPI was determined. The data were subjected to descriptive, multivariate, and geostatistical statistics. Hydrogeological and hydrochemical maps were prepared. The months of October to April are characterized as rainy months (>300 mm). The coefficient of variation showed low standards due to atmospheric circulation systems. The Gaussian and exponential models presented the best fits (R2 > 0.86) for rainfall and SPI data. The quality of groundwater in the study area ranges from excellent to good, except for the north center part of the study area, where the groundwater quality is poor. An alert is issued for the southern region of the Bahian coast regarding the safety of the local population, including the risk of landslides resulting from rain and floods.