Towards a generic theoretical framework for pattern-based LUCC modeling: An accurate and powerful calibration–estimation method based on kernel density estimation. (December 2022)

• Record Type: Journal Article
• Authors: Mazy, François-Rémi; Longaretti, Pierre-Yves
• Is Part Of: Environmental modelling & software. Volume 158(2022)
• Publication Date: 2022-12
• Subjects: Land use change -- Land cover change -- LUCC -- Model development -- Model evaluation -- Model accuracy -- Calibration -- Probability -- Transition -- Kernel density estimation; Environmental monitoring -- Computer programs -- Periodicals; Ecology -- Computer simulation -- Periodicals; Digital computer simulation -- Periodicals; Computer software -- Periodicals; Environmental Monitoring -- Periodicals; Computer Simulation -- Periodicals; Environnement -- Surveillance -- Logiciels -- Périodiques; Écologie -- Simulation, Méthodes de -- Périodiques; Simulation par ordinateur -- Périodiques; Logiciels -- Périodiques; Computer software; Digital computer simulation; Ecology -- Computer simulation; Environmental monitoring -- Computer programs; Periodicals; Electronic journals; 363.70015118
• DOI: 10.1016/j.envsoft.2022.105551
• Languages: English
• ISSNs: 1364-8152

Abstract

Abstract: Several modeling strategies have been proposed to study Land Use and Land Cover changes (LUCC). However, substantial discrepancies have been noted between different models for the same problem, questioning their overall reliability and reproducibility. To address this challenge, we elaborate a generic, formally correct, theoretical framework for pattern-based LUCC modeling, which is implemented in our own software, CLUMPY (Comprehensive Land Use [and cover] Modeling in PYthon). The present work focuses on calibration. We devise a kernel density calibration–estimation method (Bayes-eKDE) that is shown on synthetic artificial data to be both accurate and algorithmically efficient. We also introduce a generic evaluation method that allows us to compare the calibration efficiency of existing models. The gain in precision and computational time of our calibration method is precisely quantified in this way. Highlights: Accurate calibration–estimation methods improve LUCC models reliability. A method based on Kernel Density Estimation has been developed (Bayes-eKDE). Our method is substantially more accurate and efficient compared to existing ones. The method is implemented in our own software CLUMPY. Validation must be performed in explanatory variable space, not from map comparison.