4. Conclusion
The TEOTIL2 model has been extended to simulate fluxes of eight metals (arsenic, lead, cadmium, copper, chromium, mercury, nickel and zinc) at the scale of Norway’s “regine” catchment network. The spatial distribution of “diffuse source” metal concentrations in Norwegian surface waters has been derived from the 2019 “1000 Lakes” dataset, while temporal changes at regional scale are inferred from long-term monitoring in major rivers collected as part of Elveovervåkingsprogrammet. Direct point discharges (e.g. from industry) are also included, based on data extracted from Miljødirektoratet’s discharge licensing database.
For most metals and most locations, the model adequately reproduces patterns in observed metal fluxes. Assessment of the model’s performance demonstrates it is clearly superior to the current “default” approach of considering only point discharges, as done e.g. for OSPAR reporting. There is some evidence that the point discharge datasets used to drive the model may contain significant errors for some locations and years, leading to sudden implausible spikes in the model output. With further refinement of the input datasets, it may be possible to improve these aspects of model performance.
Despite these limitations, TEOTIL Metals seems capable of producing plausible simulations of metal fluxes for the period from 1990 to 2019. The model is simple and fast (run times are typically less than 10 seconds) and it will be straightforward to use the model to simulate fluxes for future years by assimilating new water chemistry data as it becomes available.
The entire TEOTIL2 modelling framework (including TEOTIL2 Metals) is Open Source and available to download from GitHub (here).