Estudo de características climáticas da região de Guarapuava por meio de análise de série temporal e modelagem ARIMA

Abstract

The climate varies continuously with time, and its variables can be modeled as time series. The Enos phases with El Niño and La Niña events influence temperature and precipitation in different parts of the world. The objective of the work was to evaluate the trends of the climatic variables, minimum, maximum and average temperature and rainfall, in the geographical region of Guarapuava in Paraná State, using time series and ARIMA modeling. The study comprised a time series of rainfall and temperature of 43 years. The time series trend was analyzed using the Box-Jenkins models, of integrated auto-regressive moving averages (ARIMA). It was found that the rainfall shows stationary behavior in the period. The average air temperature shows an increasing linear trend, which is a consequence of the increase more related to the increase of the minimum temperature, than of the maximum temperature over time. The average air temperature showed an increase of approximately 1.4°C in the period, rising from 16.8 to 18.2°C.