Separation of the time periods with dominant climatologic parameters by multivariate statistical analysis

摘要

Climatology is one of the areas that rely on collecting huge quantities of data. The longer a time period is observed, the better; the more parameters are included, the better. But the human mind cannot easily extract useful information from the abundance of data; thus, many valuable facts may be overlooked. Having that in mind, the authors of this paper have focused on the data condensation with the goal of gathering more information about the underlying trends of the main climatologic parameters change to show climatic variability. The data from the Belgrade Meteorological Observatory are analysed using a number of different methods of multivariate statistical analysis. Separation of the years and time periods with similar weather pattern characteristics was successful and indicates that there is a trend of temperature increase, as well as a trend of the temperature range decrease.