Ulating Hydrologic Impacts of Urban Growth Using SLEUTH, Multi Criteria Evaluation and Runoff Modeling

Abstract

Research on the sustainability of future urban environments has shown not only the value of simulation modeling, but also the strength of coupling their forecasts together into loosely integrated models. In this research we propose a way to improve modeling of urban futures by using enhanced land use change modeling as an input to a hydrological model. Our forecasts of hydrologic impacts are based on the L-THIA model, but this model is sensitive in turn to the land uses associated with urbanization, especially impervious surfaces and other urban uses. In order to improve the performance of the SLEUTH land use/cover change model, we modified one of that models inputs using multi-criteria evaluation (MCE). The data basis for the modeling used land cover classes mapped for the years 1987, 1992, 2000 and 2005 in the Gorgan Township of northeastern Iran using Landsat TM and ETM+ remotely sensed images. A hybrid unsupervised, supervised and on-screen classification method was used to generate land use/cover maps of the area. To test the value of enhanced modeling, we compared the results from three model forecasts of the year 2040, an original SLEUTH forecast, an urban suitability-informed SLEUTH forecast, and a hydrologic plus urban suitability-informed SLEUTH forecast. As a control, we also used standard MCE to forecast urban growth for the year 2040. The integrated model comparison was made based on two future urban landscape descriptors: (1) potential surface runoff volume; and (2) landscape metrics. Results indicate that the integrated modeling approach produces a land use arrangement through SLEUTH that can be better directed towards higher land suitability and more favorable hydrologic conditions. The results for the combined hydrologic plus urban suitability-informed SLEUTH modeling were superior to the other three methods in terms of lower potential for runoff volume generation and more sustainable landscape metrics of shape, size and proximity. We demonstrate that the self-organizing cellular automata method behind SLEUTH is capable of better informed land use planning when compared to the simple MCE method. The approach can also be improved by using suitability layers for other land use practices to achieve a more comprehensive land use plan.