MULTIVARIATE STATISTICAL ANALYSIS OF FOG PHENOMENON IN NORTHWESTERN ROMANIA

Adina-Eliza CROITORU, Iulian-Horia HOLOBÂCĂ, Ioan RUS, Tatiana MUREŞAN

ABSTRACT: In the present study a statistical analysis of annual, seasonal and monthly data series with number of fog days was made in order to identify the main factors, which decide the spatial variability of the phenomenon. In order to investigate the behavior of fog phenomenon in the Northwestern part of Romania, several statistical techniques were applied to time series. Hierarchical cluster analyses and principal components analyses (PCA) were used to explain the spatial distribution. The datasets spread over 46 years in 8 weather stations from Northwestern Romania, located in the plain and hilly areas. Thus, in the area, there are two types of variability, and among the local factors, the vicinity of water resources is the most important one. There are three principal components identified which explain more than 92 % of the variance for the spatial distribution of fog phenomenon in the area. Among them, seasonal numbers of fog days are associated to general climatic conditions of the analyzed region and distance from water resources, altitude and relief unit type are considered local factors which influence the spatial distribution of the fog phenomenon.

Keywords: Fog, cluster analysis, principal components analysis, local factors, Northwestern Romania .

Full article here