STATISTICAL CHARACTERIZATION OF CLIMATE TRENDS IN ANNUAL SERIES OF AVERAGE TEMPERATURE AND PRECIPITATION TO THE STATE CAPITAL OF BRAZIL
The objective of this work was to use statistical methods to identify the existence of possible climate
trends in annual data of average temperatures and precipitation in the capitals of the Brazilian
states. Considering this aim, statistical tests of linear regression and non-parametric tests of Mann-
Kendall and Pettitt were employed on the data that is available at the National Institute of
meteorology. The information analyzed include time intervals with varying periods depending on
the locality studied, although, the data records are mostly distributed between 1961 and 2011. The
results indicate that from the 25 capitals considered, there was a 40% positive trend in average
temperature, and 88% of the rainfall indexes did not accuse a trend.
PaperID: p 001-015
Author's Name: Júlio César Penereiro, Denise Helena Lombardo Ferreira and Vanessa Fernanda Vick Garcia
Volume: Volume 3
Issues: No 4
Keywords: Trends, precipitation, average temperature, statistical tests.