The model is tuned with plastic pollution data from peer-reviewed studies on rivers in the US. You input basic data about your watershed and your sampling parameters, and we give you an expected concentration. We even have a two-way particle size range transformation that allows you to specify a particle size range of interest and have the model output concentrations in that range. As always, models are never perfect, and this one has a ways to go, but we think this is a good framework to start with and think that using numbers from here should provide reasonable ballpark estimates. On our current dataset, the model is performing 70% better than a mean-only model.