STUDY OF THE MAXIMUM DAILY RAINS IN THE CITY OF BARREIRAS/BA THROUGH TIME SERIES METHODOLOGY

Abstract

Natural disasters are increasingly present in everyday society. In Brazil the most common natural disasters are landslides and floods which are phenomena directly related to hydrological variables such as rainfall. The study and statistical monitoring of the meteorological regime of a given region, in particular, daily maximum rainfall data can provide important information to guide public disaster prevention policies, as well as reduce the human vulnerability of the local population. The main objective of this paper was to analyze a historical series related to the maximum daily rainfall of the city of Barreiras/BA from january 1970 to may 2019 through time series analysis. Results of applied tests indicated the presence of seasonality and also that there is no trend in the series. Given this, the SARIMA model class was considered the most suitable for modeling and the adjusted models presented good predictions allowing the identification of patterns in the series.