Abstract
The aim of this work was to identify climate trends in seasonal time series for precipitation and average air temperature, observed in 243 municipalities distributed over the Brazilian biomes. The data were obtained from National Institute of Meteorology (INMET). The Mann-Kendall and Pettitt tests were applied. The statistical results and the seasonal trend distributions maps allow us to observe that all biomes exhibit non-significant trends in the both meteorological variables. The average temperature was the variable that showed the highest number of positive trends (72 in autumn, 81 in winter, 80 in spring and 92 in summer). Such trends were mostly detected during the decade of 1990 and in the Amazônia and Cerrado biomes. Negative trends were observed in the Caatinga and Cerrado biomes for the seasons of autumn, winter and spring. For the precipitation, negative trends we observed in all the four seasons, during the decades of 1980, 1990 and 2000. Positive trends were observed in the decade of 2000 and during the autumn and winter seasons, especially in the Amazônia and Mata Atlântica biomes. These trends seem to be related to the El Niño and La Niña events.
Keywords:
time series; statistical analysis; climate changes
