You are here:
Predicting Plausible Impacts of Sets of Climate and Land Use Change Scenario on Water Resources
Citation:
Tong, S. T., Y. Sun, T. Ranatunga, J. He, AND Y. J. YANG. Predicting Plausible Impacts of Sets of Climate and Land Use Change Scenario on Water Resources. Applied Geography. Elsevier BV, AMSTERDAM, Netherlands, 32(2):477-489, (2012).
Impact/Purpose:
To inform the public.
Description:
As the new decade ushers in, there will be new challenges. The world’s population is increasing and the land use patterns are changing. Inevitably with these global changes, there will be various environmental consequences. For example, our water resources, both in terms of quantity and quality, may be affected. In an attempt to provide basic information on the future conditions, this study used the East Fork Little Miami River (EFLMR) watershed as a case study to postulate the plausible changes in population and land use in the year 2050. The purpose of this study was to develop a research protocol for generating a future land use scenario, providing information which may be useful in city and urban planning, as well as in other areas of research, such as in predicting the hydrologic consequences of future land use changes in tandem with climate alterations. In this study, the cellular automata and Markov Chain modules (CA-Markov) in IDRISI (IDRISI 2008) were employed to develop a land use model for the EFLMR watershed. Moreover, since population growth is often the driving force for urbanization and suburban sprawl, this study considered its effects in future land use development. The population for the horizon year 2050 was estimated using the geometric growth model, and the population density for each census block group in the study area was calculated. These results were then incorporated into a suitability map for the “Urban area” through the Multi-Criterion Evaluation (MCE) method to produce a population filter, a surrogate depicting the impacts of population growth on land use changes. The population filter was used with cellular automata of the CA-Markov model to predict the future land use patterns. To assess the predictability of the population-coupled land use model, the land use pattern for the year 2001 was first generated from the 1976 and 1992 land use maps. The suitability map for “Urban area” was compared with the actual urban land area of that year to generate a ROC statistic. From the comparison, it was apparent that the suitability map with a population filter provided a higher ROC (ROC = 0.971) value than the one without a population filter (ROC = 0.956), indicating that he population filter can improve model predictability. Moreover, the predicted land use patterns from CA-Markov were compared to the actual land use imagery of that year to get a Kappa index of agreement. The Kappa statistic was 0.9248, showing a very good agreement. Based on these findings the land use scenario for the 2050 horizon year was generated for the EFLMR watershed. The projected land use map of 2050 shows significant changes with much urban expansion in the downstream portion of the basin.
