Project

Utilizing deep learning for streamflow prediction in the context of ecosystem service modelling

Hydrological ecosystem services are among the most important services that ecosystems provide to society. It is therefore crucial that these services are included in natural capital accounts. However, to date, a clear conceptualization of hydrological ecosystem services for natural capital accounting, is still lacking. Moreover, existing hydrological models are very specific in terms of the data needed, and therefore not scalable to other watersheds. This makes scaling up these models wall-to-wall at a national scale (as is required for accounting) very difficult. This PhD research project will develop and test hydrological ecosystem services models compatible with natural capital accounting approaches in watersheds in Europe and Brazil.