Geostatistics applied to the geoprospective GP-SET. Krige method theory and application

Abstract

Big Data provides the ability to describe many living environment dimensions, from administrative data. Territorial scales allow decision-making but are not sufficiently accurate to establish effective policies tailored to the needs of sustainable development. Geographic Information Systems (GIS) can integrate any kinds of data. Unfortunately GIS are still used without exactly knowing the methods implemented, or having geographic knowledge of the phenomena studied. Very different results can be obtained with a same dataset. The inappropriate use of GIS is damaging for prospective studies derived from spatial analysis. The goal of geoprospective is to develop robust methods to address these challenges. GP-SET. krige makes two spatiotemporal indicators that accurately model the spatial spread of human phenomena and their uncertainty. It is based on univariable geostatistics. Applied to census data, GP-SET. krige precisely models the potential to have a demographic growth in the next ten years.