Keywords
Spatial predictors, Land use modelling, CLUE-s model, Ecuadorian Amazonas
Start Date
15-9-2020 11:40 AM
End Date
15-9-2020 12:00 PM
Abstract
CLUE-s based models are one of the most popular and frequently used land change simulation tools worldwide. These models are drawn for policy advice on various aspects of environmental and sectoral land management from single watersheds to whole countries, and supra-national entities such as the EU. Key elements of CLUE-s allocation procedure are spatially-explicit location factors (=drivers of land cover land use change (LULC)) which determine the suitability of a land use type. Decisions on which location factors should be involved in the simulation are based on a thorough review of the processes important for the spatial allocation of land use, and depend on study area context, available data and eventually modellers choice. But CLUE-s based studies usually lack detailed information on the location factor selection process or include frequently used factors for which spatial data is easy to obtain. To our knowledge, there is no CLUE-s study where model outcomes and potential policy implications are presented based on different sets of calibrated location factors. Given the high policy relevance of CLUE-s-based modelling, it is however crucial to investigate the robustness and reliability of modelling outcomes in order to avoid erroneous conclusions that could lead to false policy recommendations. Hence, the overall objective of the presented study is to simulate three different business-as-usual (BAU) scenarios to assess how the variability of calibrated LULC drivers (n=8, 15, 16) influence the prospected location of future deforestation hotspots for the case of Canton Loreto in the Ecuadorian Amazonas and how related policy implications may differ as a result, (ii) to critically reflect on the CLUE-s-based modelling “cookbook” and what needs to be revised in order to report simulated modelling uncertainty in a transparent fashion to policy makers and environmental managers.
How spatial predictors of location suitability influence land allocation – a neglected issue in CLUE-s based scenario modelling?
CLUE-s based models are one of the most popular and frequently used land change simulation tools worldwide. These models are drawn for policy advice on various aspects of environmental and sectoral land management from single watersheds to whole countries, and supra-national entities such as the EU. Key elements of CLUE-s allocation procedure are spatially-explicit location factors (=drivers of land cover land use change (LULC)) which determine the suitability of a land use type. Decisions on which location factors should be involved in the simulation are based on a thorough review of the processes important for the spatial allocation of land use, and depend on study area context, available data and eventually modellers choice. But CLUE-s based studies usually lack detailed information on the location factor selection process or include frequently used factors for which spatial data is easy to obtain. To our knowledge, there is no CLUE-s study where model outcomes and potential policy implications are presented based on different sets of calibrated location factors. Given the high policy relevance of CLUE-s-based modelling, it is however crucial to investigate the robustness and reliability of modelling outcomes in order to avoid erroneous conclusions that could lead to false policy recommendations. Hence, the overall objective of the presented study is to simulate three different business-as-usual (BAU) scenarios to assess how the variability of calibrated LULC drivers (n=8, 15, 16) influence the prospected location of future deforestation hotspots for the case of Canton Loreto in the Ecuadorian Amazonas and how related policy implications may differ as a result, (ii) to critically reflect on the CLUE-s-based modelling “cookbook” and what needs to be revised in order to report simulated modelling uncertainty in a transparent fashion to policy makers and environmental managers.
Stream and Session
false