Comparison of land cover spatial trend model and real land cover changes: case study of Slovak Republic

Abstract

Land change models offer an essential resource for predicting future land use and land cover (LULC) change. Model accuracy and proper validation are essential for supporting the decision process related to landscape planning. The paper's main aim is to test and analyse the accuracy of the LULC trend model based on the land cover changes from 1990 to 2012. The Land Change Modeler (TerrSet) and Corine Land Cover data for the area of Slovakia was used. The obtained model was compared with the real changes captured in the next period (2012 to 2018). Different model reliability was observed in different types of land cover. From a relatively low level of reliability to a level of more than 80% concordance. Results pointed on significance and limits of tools for predictive modelling in a specific type of landscape (included very heterogeneous natural conditions) after period of strong institutional changes impact on LULC.