Modelagem de dados meteorológicos da cidade do Recife utilizando a metodologia da análise de séries temporais

Abstract

Models based on the statistical techniques of time series were used to identification of trends, seasonality and randomness in monthly averages of the meteorological parameters of Recife city. To analyses trends it is possible to adjust linear regression and moving averages models. Were used time series data, covering a period of nineteen years, from January 1989 to December 2009, on the average monthly temperature compensated, maximum temperature, minimum temperature, precipitation and relative humidity. Studies on regional and global climate change have drawn attention not only of the scientific community over the last decade, but also the general public. Advance knowledge of weather conditions can help society avoid wasting human and material resources, so it can adjust the schedule of activities, based on forecast demands that depend on these weather conditions. Compensated average temperature, maximum temperature, minimum temperature and rainfall showed positive trends for characterizing the city of Recife, a mild climate change, most evident in the second decade of the studied period. During the 19 year period chosen for this study, the El Nino weather event occurred 9 times causing much of the scientific community to give this phenomenon the probable liability of climate anomalies observed in recent decades