Models often provide an excellent approach to better understand how pollutants may behave under a range of conditions that cannot be directly measured. Some models result in numeric predictions, such as expected concentrations under different flows in a stream. Other models are not good at predicting specific concentrations but may help understand how pollutants vary or respond to a change in management. It is critical that the strengths and limitations of any model are well understood before they are used.
Models can also be helpful in prioritizing or designing BMP implementations. For example, a landowner wishes to move a corral off of a creek. The landowner’s preferred location is 25 feet off of the creek. A model may suggest this location will result in a 75% reduction of nutrients to the creek. A similar project that would result in a projected 95% reduction may receive a higher funding priority, benefiting the landowner. The use of the model to inform the landowner that moving the corral 40 feet from the creek may result in a projected 95% reduction in nutrients might be sufficient information to allow the landowner to change their plans for the final corral location.
Because modeling may require unique skills, you may need to subcontract this work. It is critical, however, that those involved in other aspects of managing and implementing the BMP understand modeling sufficiently to make informed decisions about the modeling process, including an understanding of the strengths and limitations of a particular modeling approach. Click here(link to model references at the end of this page) for some specific models to consider.
Types of Models: Deterministic: The model outcome is always the same given the same inputs. Stochastic: Incorporates uncertainty into the model so the results are not always the same.
Analytical: Have a closed-form mathematical formulation. Simulation: Lack a single, general mathematical solution; represents complex, non-linear relationships.
Process-Based: Components represent specific hydrologic and ecological processes. Empirical: Based on simple correlations; derived from data.
Temporal and Spatial Scale Models: Point scale: Point-in-space approach with no spatial distribution (field, reach, or plot scale). Spatial: Systems that vary in space (e.g. watershed models). Static: Point-in-time approach with no temporal distribution (e.g. event based model). Dynamic: Systems that change in time (long-term simulation model).
If you are planning to use a model, consider the following:
• What question(s) are you trying to answer using the model?
• What type of model should be used to model the system? The monitoring objectives and the pollutant of concern will determine this selection.
• What is the application scale of the model? Does it match your project scale e.g. plot, field, watershed, or basin? A complex, spatially explicit watershed scale model that uses a 30 m resolution may not be appropriate to assess the impacts of a 10 m wide buffer strip installed next to a single agricultural field.
• Is the model process-based (attempts to mimic the natural processes in a system)? If so, it is important to understand the dominant processes of the pollutant in the system to well enough so that your model will provide useful results?
• Is the model based on statistical relationships of previously collected data? If so, it is important not to extend predictions beyond the limits of these data.
• Is the model an event model (single storm response) or a long-term simulation model (month-decades)?