Author(s): Anthimos Spyridis; Alexandros Konstantinidis; Vasiliki Koutalou; Konstantinos Perakis; Aris Psilovikos
Linked Author(s):
Keywords: No Keywords
Abstract: Natural phenomena, as observed and recorded in nature, can be determined by a spatial distribution of certain characteristic measurable quantities that describe them and are called Regionalized Variables. The geostatistic approach of such physical problems is based on the spatial structure presented by all these regional variables. The correct mathematical study of such variables is difficult and shows significant deviations because spatial variability is usually extremely unstable, with all kinds of discontinuity and anisotropies. The numeric values of a regional variable z (x) and z (xi) should not be interpreted as independent realizations of a random function Z (x), because this view does not consider the spatial autocorrelation between two adjacent values z (x+h) and z (x). In the science of geostatistics the two independent aspects, randomness and spatial structure, that characterize peripheral variables, are a random function. The present paper examines a new regional variable of the average hyper-annual rainfall, of five years for the years 2015-2020 for an area of the Region of Central Macedonia (Fig. 1.), which was approached through geostatistical analysis. For this purpose, rainfall data from 14 Meteorological Stations of the region were used, supported by the National Observatory of Athens (free data from meteo. gr).
Year: 2022