ANÁLISE DE SÉRIES TEMPORAIS DE PRECIPITAÇÃO E TEMPERATURA NO MUNICÍPIO DE AREIA-PB UTILIZANDO MODELOS ESTATÍSTICOS ARIMA E ETS

This research analyzed the time series of monthly average precipitation and temperature in Areia-PB, covering the period from 1996 to 2023, employing ARIMA (Autoregressive Integrated Moving Average) and ETS (Exponential Smoothing State Space Models). The methodology included time series analysis focusing on identifying error, trend, and seasonality characteristics of the data. A noticeable trend was identified, characterized by a decline in precipitation along with a rise in temperature after 2010. The ARIMA model demonstrated greater efficacy, forecasting the persistence of these trends in 2023. The time series decomposition analysis revealed strong seasonal fluctuations and enduring trends. The results showed a significantly lower RMSE (Root Mean Square Error) for the ARIMA model compared to the ETS model. This investigation highlights the critical need for continuous climate monitoring and provides significant insights for water resource management and agricultural strategies in the region, emphasizing the imperative for adaptive methodologies in light of the detected climate changes.