Análise de agrupamentos de séries climatológicas de temperatura e amplitude térmica em Cuiabá

Abstract

The objective of this work was to use statistical analysis tools to evaluate possible annual trends of the series of average daily temperature and mean daily thermal amplitude, as well as to classify the months of the year according to the distributions of these variables. The historical series of 12/1986 to 12/2016 average surface air temperature was used from the meteorological station of the Brazilian Institute of Meteorology in Cuiabá, Mato Grosso. The trend analysis was evaluated by the non-parametric Mann-Kendall test, not detecting a significant tendency for the average monthly average temperature, as well as for the medians and minimum and maximum month-to-month confidence intervals. A non-hierarchical K-means cluster analysis was developed for the months from the monthly averages and percentiles of the analyzed variables, obtaining two main stations and two of smaller amplitudes, corresponding to the transitions.