Areal transformation of geographical data and its principles in new issue of Geografia Slovaca

Sládeková Madajová, M. Hurbánek, P.
AREAL TRANSFORMATION OF GEOGRAPHICAL DATA: PRINCIPLES, METHODS AND APPLICATION.

Geografický ústav, 2016, 112 s.
Fulltext

 

Geographers face several obstacles in the analysis of developments and processes in individual regions which make impossible a correct analysis, direct comparison or a combination of data from different sources. The territorial units, data are available for, may vary over time or in many cases it is necessary to analyse the data in other areas than currently available administrative / statistical units. A problem of spatial discrepancy is frequent especially in interdisciplinary research, where quality integration of data from various sources (demographic, economic, social, environmental, etc.) using different spatial divisions of geographical space is a common task. The transfer of data from one geography into another is among the needy in Slovak literature however still underdeveloped themes. The paper therefore aims to introduce the issue of areal transformation of geographic data more detailed, to explain its principles, summarize, describe and classify different approaches and methods enabling data transfer and estimation from one system of geographical units to another (modified or completely different) zonal system, to apply the selected interpolation methods and to compare their accuracy in Slovak conditions. The performance of traditional areal interpolation method – areal weighting – was compared with the approaches, which use auxiliary information to improve the results of interpolation process. The testing suggests that the accuracy of areal interpolation is influenced by the shape/size of the source and target zones. The results confirm that the methods using ancillary information are able to make more accurate estimation but only to the extent, to which these additional data reflect (or correlate with) the spatial distribution of the estimated variable.