Optimal hydrological restoration of Swedish wetlands with deep learning and hydrogeodesy

Agricultural expansion, forestry, hydropower and road infrastructure have heavily altered the hydrological regimes of Swedish wetlands. As a result, there is a growing interest in their hydrological restoration.

Yet, hydrological restoration should also target climate change mitigation by controlling resulting greenhouse gas emissions.

Knowledge of hydrological regimes is compulsory to guide optimal wetland restoration, but a lack of hydrological monitoring limits this understanding.

This project combines artificial intelligence, earth observations, and state-of-the-art techniques tracking water levels from space to define the hydrological regimes in 53 wetlands of international importance.

Our main objectives are to quantify comprehensive water surface extent and changes, the spatial and temporal variability of water levels, and define guidelines to prioritise wetland restoration, maximising ecosystem health and climate change mitigation. Our communication plan targets wetland researchers, environmental authorities, and the general public.

It includes a phone-based game application where the public can “play” the principles of optimal wetland restoration.

We expect our project to refine the Swedish Wetland Inventory and double the number of water resources with real-time monitoring of water levels.

Finally, we expect to provide the hydrological knowledge that Swedish environmental authorities need to prioritise and restore degraded wetlands.