MiWaRes: Digital Twin for Climate-Resilient Mixed Forests

Retaining water in the forest

Forests are increasingly affected by heat, drought, and heavy rainfall. At the same time, they play a vital role in flood protection, groundwater recharge, and the preservation of biodiversity. The MiWaRes project is developing a digital monitoring, decision-making, and forecasting system that supports water-balance-based management of mixed forests.

The Bellheimer Forest near Offenbach an der Queich in the Southern Palatinate serves as a real-world hydrological laboratory. The historically developed network of ditches and the connection to the Queich River provide suitable conditions for scientifically investigating targeted water retention in the forest. The goal is to reduce drought stress in forest stands, retain water within the landscape, and thereby also help prevent both high and low water levels.

Project goal

MiWaRes is developing a digital twin for water-balance-based management of climate-resilient mixed forests. This digital twin integrates sensor data, weather and water level data, hydrological models, AI methods, and ecological and silvicultural indicators.

Among other things, the focus is on:

• the monitoring of soil moisture, groundwater and surface water levels, and tree vitality,
• the study of water distribution, infiltration, and water retention in the network of ditches,
• the modeling of soil moisture, groundwater levels, and water balance components,
• the development of short-term forecasts for periods ranging from days to weeks,
• the assessment of impacts on tree species, soil, groundwater quality, and biodiversity,
• the development of scientifically sound recommendations for forestry operations and government agencies.

Scenario assessment using AI-based models

In addition to the digital twin of the demonstration area, we at Fraunhofer IOSB are developing AI-based surrogate models for operational water balance management. These models replicate key results from the detailed water balance model ParFlow-CLM and can therefore be used for real-time assessment of scenarios—for example, with regard to soil moisture, infiltration, water levels, or potential waterlogging. Finally, the MiWaRes Assist prototype supports the assessment of potential water discharges while taking hydrological, ecological, and technical constraints into account.