Is there a relationship between the total area of the watershed and the threshold to be set for discretizing the watershed into sub-basins? I am looking for literature that addresses this.
Depending on the modeling goal as soon as you
get good topographic data,
learned your watershed (size, slope, land use) and
understand what happens in the watershed, then:
- explore different methods (flow accumulation, stream order, slope)
- test sensitivity - try various thresholds and see how they affect your model.
- choose a threshold that gives accurate results for your objectives.
- confirm your choice with observed data (validation).
- Ensure the sub-basin sizes make sense for your project.
- Keep records of your choices for transparency and be ready to adjust the threshold if needed through iteration and expert input.
My rule of thumb for the WEPP watershed tools is fairly simple and likely applies to other models. Figure out how many subwatersheds you computer can handle. Divide the total watershed area by the maximum number of subwatersheds you computer can handle, and you have the number. You may determine the maximum number of subwatersheds/hillslopes/management units or whatever you model calls them by how long a computer run will last. In the case of the WEPP watershed model, it can handle 1000 hilllslopes. Keep in mind that finer resolution DEMS can generate a greater number of subwatersheds. Once you have completed a large watershed run, you can always go back and look at subwatersheds wihtin the larger watersheds for finer modeling. Another concern is rainfall distribution. If you can’t do distributed weather modeling, then you may want to limit your watershed area to a smaller area where the weather is similar. In a mountainous study we did recently, we divided the study up into three climate zones as there was a significant differerence in climate with elevation. Good luck with your watershed modeling.
- Badges
- Science topic