Geostatistical Analysis of Time Series: An Exploratory Analysis of Climatologic and Ecological Data using a Spatial Instrument

Alexandru PETRISOR
Sorin CHEVAL
Victor ZINEVICI
Laura PARPALA

Keywords

exploratory data analysis, geostatistics, kriging, GIS, time series

Abstract

The extensive use of the Geographical Information Systems, but also of their extensions, in fields that do not necessary connect to geography, but allow for the implementation of methodologies based on the analysis of spatially referenced data, had resulted into new applications and methodological approaches to answer specific research questions. This paper proposes such an extension from the real space define by longitude and latitude to a virtual space defined by two time parameters, month and year, in order to reveal specific temporal patterns of climatologic and ecological phenomena. The methodology is applied in two studies, one aiming to analyze temperature and precipitation data in Romania over a long period, and the other to analyze the effects of anthropic impact and ecological restoration on the density of specific groups of aquatic organisms. In the first study, the method was able to detect climate hazards, while in the second the results had shown different behaviors of analyzed groups, relevant to their ecology. In both cases, the proposed methodology proves its potential for being used as an exploratory research tool in climatology, ecology, and other fields as well.