Unpacking dasymetric modelling to correct spatial bias in environmental model outputs. (November 2022)

• Record Type: Journal Article
• Authors: Kallio, Marko; Guillaume, Joseph H.A.; Burek, Peter; Tramberend, Sylvia; Smilovic, Mikhail; Horton, Alexander J.; Virrantaus, Kirsi
• Is Part Of: Environmental modelling & software. Volume 157(2022)
• Publication Date: 2022-11
• Subjects: Bias correction -- Dasymetric modelling -- Downscaling -- Areal interpolation -- Environmental modelling -- meta-Modelling; 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.2022.105511
• Languages: English
• ISSNs: 1364-8152

Abstract

Abstract: Complex environmental model outputs used to inform decisions often have systematic errors and are of inappropriate resolution, requiring downscaling and bias correction for local applications. Here we provide a new interpretation of dasymetric modelling (DM) as a spatial bias correction framework useful in environmental modelling. DM is based on areal interpolation where estimates of some variable at target zones are obtained from overlapping source zones using ancillary information. We explore DM by downscaling runoff output from a distributed hydrological model using two meta-models and describe the properties of the methodology in detail. Consistent with properties of linear scaling bias correction, results show that the methodology 1) reduces errors compared to the source data and meta-models, 2) improve the spatial structure of the estimates, and 3) improve the performance of the downscaled estimates, particularly where meta-models perform poorly. The framework is simple and useful in ensuring spatial coherence of downscaled products. Highlights: We introduce dasymetric modelling (DM) as a spatial bias correction method. Relationship between DM and linear scaling bias correction is shown. DM consists of meta-modelling, bias correction, and areal interpolation components. Performance of meta-models are significantly improved by DM. DM corrects the wider spatial autocorrelation patterns in meta-model outputs.