Land use and land cover change has been recognized to have significant environmental impacts in a watershed, such as regulation of water quality. However, the identification of potential regions that are sensitive to land change activities for the protection of water quality poses a grand challenge particularly in a large watershed. These potential regions are o en associated with critical thresholds in terms of, for example, water quality. In this study, we developed an agent-based land change

more »

... model to investigate the relationship between land development activities and water quality in eight North Carolina counties that cover the lower High Rock Lake Watershed area. This agent-based model, which is empirically calibrated, is used to identify space-time locations of those regions at critical thresholds of water quality in this study area. Our experimental results suggest that land development as a form of system stress is of pivotal importance in a ecting water quality at sub watershed level and the state transition of water quality. The agent-based model developed in this study provides solid support for investigations on the impact of land development under alternative scenarios in a large watershed. identify those regions within the threshold (Haidary et al. ; Schueler et al. ). The identification of these potential regions in a watershed is very important for the regulation of water quality, watershed management practices, and landscape conservation. Once these regions are degraded (transition to a di erent equilibrium), the restoration process may be di icult (or even infeasible), time-and cost-consuming particularly for those regions that are beyond the critical threshold (e.g., densely urbanized or heavily deforested; o en irreversible) (Schueler ). However, the identification of where and when those potential regions in a watershed may experience threshold due to, for example, future land development or alternative policy intervention poses a grand challenge. The processes of LULCC per se and the way that LUCC a ect water quality in a watershed are complex, nonlinear, and uncertain. For example, a time lag e ect may exhibit in terms of the impact of LULCC on watershed-level water quality (Meals et al. ): the impact of LULCC on water quality may not be immediately observed. It needs a period of time (depending on indicators of interest; see Meals et al. for water quality to respond to the land development or management practices. Further, the processes associated with LULCC may not be stationary over time. Complex dynamics in these processes further complicate the identification of space-time locations of those potential regions within thresholds. Study Area and Data . Our study region (see Figure ) includes eight counties in North Carolina, USA, which cover the lower High Rock Lake Watershed (HRLW) area. The total area of our study region is , km with a total population of . million in the year of . This study region is located within the Piedmont area of North Carolina, with elevation varying from m to m. Percentage of urban land cover in the study region is . %, . %, and . % in year , , and . This study region experienced . % of increase in percent urban cover from to , and this rate becomes . % from to (due to the impact of economic recession). JASSS, ( ) ,