| Abstract: |
Land use and changes in land use can have a profound effect on water quality and water quantity. Urban growth, sea level rise, and other drivers of change are anticipated on the Delaware/Maryland/Virginia (DelMarVa) peninsula over the coming decades. Here we present a modeling approach that links demographic models, spatially explicit models of land use change, and models of nutrient loading and runoff to ultimately provide estimates of future hydrologic change on the DelMarVa peninsula. A challenge for our research team has been the coordination of information flow between individual models and team members in our efforts to make these estimates.
The models involved in this project are called SLEUTH (version 3r), GAMe, and GISHydro. SLEUTH is a spatially explicit land use change model that uses inputs of historic urban land cover change, a transportation network, slope, and a weighted map indicating areas that are more or less likely to experience development. SLEUTH ultimately produces spatial and tabular outputs of probable future urban land cover change. GAMe consists of a demographic model that estimates the number of houses (county scale), which need to be constructed at horizon years, and a suite of statistical models that estimate how the county scale housing and job growth will be distributed to minor civil divisions (MCDs, equivalent to small municipalities) within the county. Finally, GAME estimates the impervious surface footprint of this new growth for each MCD. GISHydro is a GIS-based tool that melds the data needed by hydrologic models of water quality and quantity with the algorithms needed to interpret and prepare this data for input into the hydrologic models. In this modeling approach SLEUTH output in the form of anticipated future land use across the DelMarVa peninsula serves as the driving input for two distinct types of hydrologic models. The first type of these hydrologic models is the EPA Chesapeake Bay Program nutrient loading model already embedded within GISHydro. This hydrologic model produces estimates of nitrogen, phosphorus, and sediment loadings from user-defined locations within the DelMarVa peninsula. The second type of hydrologic models supported by GISHydro are flood models. These models predict floods (e.g. the 100-year flood) as a function of watershed size, slope, soils, location, and land use. The SLEUTH land use informs the flood models as to the relative quantities and intensities of urban, agricultural, and forest lands present in the watershed area being studied.
The emphasis of this presentation is on the structure of linkages, feedbacks, and interpretation of information flowing between the SLEUTH, GAMe, and GISHydro models. Outputs from one model were not always precisely suited as input to another model. Creativity, critical thinking, and an appreciation of model sensitivity and uncertainty were crucial for our project team to attain the goal of linking these disparate tools for the purpose of estimating hydrologic change on the DelMarVa peninsula. Examples of how information was shaped and reformed as it flowed from one model to the next will be provided in this presentation.
|
