Keywords

land use modelling; Bayesian method; weight of evidence

Location

Session D1: GIS and Environmental Modeling for Decision Support

Start Date

16-6-2014 3:40 PM

End Date

16-6-2014 5:20 PM

Abstract

The challenge of modelling the ever changing land use cover (LUC) of the earth has resulted in the creation of diverse models to depict these variations. Most of the land use cover change (LUCC) models were developed for either specific projects or locations; therefore, they are not generic. Also, most of the current LUCC models have been applied to only one LUC class which is less complex, e.g. urban growth or at most two LUC classes.

This work investigated existing LUCC models with the aim of finding a generic geospatial workflow process approach for modelling complex LUCC so that the influence of geospatial variables in LUCC could be measured and tested before inclusion. For the initial calibration phase of the process the Weight of Evidence (WoE) method was used to measure the influence of spatial variables in LUCC and also to determine the variables significance. A Genetic Algorithm tool was used to enhance the WoE coefficients and to give the best fitness of the coefficients for the model. The model process was then validated using kappa and fuzzy similarity map comparison methods, in order to quantify the similarity between the observed and simulated spatial pattern of LUCC.

The performance of the workflow process was successfully evaluated using the Auckland and Wellington Regions of New Zealand as the study areas. To test the generic nature of the workflow process model the data resolution, area extent and complexity of transitions of the Wellington area was different from that of Auckland.

The performance of the workflow process model was found to be generic and robust in both study areas. In addition the model revealed that the set of variables influencing a specific transitional change was unique.

COinS
 
Jun 16th, 3:40 PM Jun 16th, 5:20 PM

Geospatial workflow process for modelling complex land use cover change

Session D1: GIS and Environmental Modeling for Decision Support

The challenge of modelling the ever changing land use cover (LUC) of the earth has resulted in the creation of diverse models to depict these variations. Most of the land use cover change (LUCC) models were developed for either specific projects or locations; therefore, they are not generic. Also, most of the current LUCC models have been applied to only one LUC class which is less complex, e.g. urban growth or at most two LUC classes.

This work investigated existing LUCC models with the aim of finding a generic geospatial workflow process approach for modelling complex LUCC so that the influence of geospatial variables in LUCC could be measured and tested before inclusion. For the initial calibration phase of the process the Weight of Evidence (WoE) method was used to measure the influence of spatial variables in LUCC and also to determine the variables significance. A Genetic Algorithm tool was used to enhance the WoE coefficients and to give the best fitness of the coefficients for the model. The model process was then validated using kappa and fuzzy similarity map comparison methods, in order to quantify the similarity between the observed and simulated spatial pattern of LUCC.

The performance of the workflow process was successfully evaluated using the Auckland and Wellington Regions of New Zealand as the study areas. To test the generic nature of the workflow process model the data resolution, area extent and complexity of transitions of the Wellington area was different from that of Auckland.

The performance of the workflow process model was found to be generic and robust in both study areas. In addition the model revealed that the set of variables influencing a specific transitional change was unique.