The scale at which environmental transport and fate modeling occurs is often influenced by regulatory needs, scope of potential impacts, complexity, and/or data availability. Local modeling at the edge-of-field or effluent discharge point is often used in a regulatory context as the inputs associated with the model results are likely less complex and variable. Watershed or catchment scale modeling requires additional information about the landscape, hydrological conditions over time, variability in chemical emissions, as well as more complex and time consuming models. Regional or national models generally use input data with less spatial resolution but encompass a wider extent, thus reducing the need to extrapolate a small set of model results from limited areas in order to understand the national context.
Local models operate at the edge-of-field, individual point source discharge, or single location. These models are often used for screening level assessment due to their ability to parameterize in a consistent and conservative way, while avoiding many other aspects of environmental variability and uncertainty.
Watershed models incorporating the hydrological connectivity of surface water within a basin require additional data beyond that needed for local scale models. Watershed models incorporate aspects such as variability of landscape factors, heterogeneous chemical use/emissions across the watershed, and hydrologic aspects such as time of travel. Waterborne uses Geographic information systems (GIS) in combination with models such as SWAT, PRZM, RIVWQ, and RICEWQ to compute estimated environmental concentrations in watersheds. Waterborne have modeled watersheds in the United Kingdom, United States, Japan, Australia, and Europe.
Consistent modeling input data across large regions are required for exposure modeling at the national/regional scale. This can be especially difficult when the study area crosses national boundaries, as in this examination of groundwater vulnerability in Europe using the metaPEARL, geoPEARL, MEDRICE and RICEWQ model for rice paddies. Even when contained within a single country with a common dataset, variability in environmental factors and use patterns can result in a large number of model runs, as demonstrated in this case study using PRZM to model soil runoff across the continental US. Even “watershed” models can be super-sized to a national extent, where the iSTREEM model uses a river network of over 200,000 miles to model in-river concentrations based on effluent from more than 10,000 wastewater treatment plants.