How robust are future projections of forest landscape dynamics? Insights from a systematic comparison of four forest landscape models. (December 2020)

• Record Type: Journal Article
• Authors: Petter, Gunnar; Mairota, Paola; Albrich, Katharina; Bebi, Peter; Brůna, Josef; Bugmann, Harald; Haffenden, Austin; Scheller, Robert M.; Schmatz, Dirk R.; Seidl, Rupert; Speich, Matthias; Vacchiano, Giorgio; Lischke, Heike
• Is Part Of: Environmental modelling & software. Volume 134(2020)
• Publication Date: 2020-12
• Subjects: Forest landscape models -- Model comparison -- Variance partitioning -- Disturbances -- Dispersal -- Future projections; 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.2020.104844
• Languages: English
• ISSNs: 1364-8152

Abstract

Abstract: Projections of landscape dynamics are uncertain, partly due to uncertainties in model formulations. However, quantitative comparative analyses of forest landscape models are lacking. We conducted a systematic comparison of all forest landscape models currently applied in temperate European forests (LandClim, TreeMig, LANDIS-II, iLand). We examined the uncertainty of model projections under several future climate, disturbance, and dispersal scenarios, and quantified uncertainties by variance partitioning. While projections under past climate conditions were in good agreement with observations, uncertainty under future climate conditions was high, with between-model biomass differences of up to 200 t ha −1 . Disturbances strongly influenced landscape dynamics and contributed substantially to uncertainty in model projections (~25–40% of observed variance). Overall, model differences were the main source of uncertainty, explaining at least 50% of observed variance. We advocate a more rigorous and systematic model evaluation and calibration, and a broader use of ensemble projections to quantify uncertainties in future landscape dynamics. Highlights: The first systematic comparison of forest landscape models is presented. Variance of model projections under several future scenarios is substantial. Model differences explain most of the simulated variance.