Applying a Conditional Probabilistic Approach to the Rural Growth Model

Abstract

The Rural Urban Growth (RUG) model predicts future urban growth across Europe. To understand the uncertainties associated with the model’s predictions, this study investigates the individual and the combined influence of the model’s parameters and attempts to reduce the uncertainty of future by apply a conditional probabilistic approach. This approach combines the common techniques of scenario based and probability based modeling. The approach has the advantage of enabling the probability of an outcome with in a scenario to be identified, the relative likelihood of a scenario occurring to be established and the ability to re-define unconditional parameter values to more accurately represent the assumptions of the different scenarios. This is the first time the conditional probability approach has been applied to an urban growth model, and expanded to visualise and analyse the spatial patterns of uncertainty associated with the different scenarios.

The results reveal that the RUG model is most susceptible to the accuracy of the growth values projected by the NEMISIS model. Additionally, the combined influence of the spatial parameters can generate over-estimated development values, and greatly impacts the reliability of the model’s results for regions with a low growth value. The conditional probability approach revealed that there is less uncertainty associated with the B1 scenario projection and that the spatial pattern of uncertainty is strongly correlated to areas of high development and projected growth. Based on equally probable simulations, the results also show that the B2 scenario is more likely to occur. This research also illustrates the limitations of the conditional probabilistic methodology and addresses potential improvements.