Exploring the Influence of Forest Dynamics on Flow Regimes in Alpine Watersheds

Many alpine watersheds are complex and highly engineered systems designed to mitigate the risks associated with flooding and debris flows, thereby protecting downstream areas and safeguarding land. However, these engineering interventions often come with significant financial costs and environmental challenges. Thus, less engineered solutions which exploit the connections between land cover and hydrology-sediment functions to reduce risk need to be explored as alternatives. We develop and apply a novel approach that integrates forest dynamics into hydrological simulations within a landscape evolution model for Alpine watersheds.

We employ two simulation models: LandClim, a dynamic forest landscape model, and HAIL-CAESAR, a landscape evolution model. Integration between these models is achieved through two critical parameters: the m value and Potential Evapotranspiration (PET).

The m value in the HAIL-CAESAR model defines a soil scaling parameter which affects soil transmissivity and the soil water deficit in time, and thereby surface runoff and baseflow. This parameter was spatially implemented and calibrated based on hydrological response units combining soil and land use data. To investigate parameter robustness, we performed the calibration on 46 widely different catchments across Switzerland. We hypothesize that soil water storage capacity is greater in forested areas compared to pastures, agricultural land, and unproductive land, which leads to lower flood peaks and longer recession times.

The second key linkage vegetation dynamics and runoff is Potential Evapotranspiration (PET), which is calculated spatially and adjusted based on the forest leaf area simulated by LandClim. PET captures the maximum amount of water that can be lost to the atmosphere by evaporation and transpiration, thus directly impacting soil moisture levels and, consequently, the timing of surface runoff. By integrating PET into the landscape evolution model, we improved its accuracy in simulating hydrological processes, specifically enhancing our understanding how forest cover affects flow regimes.

Our findings highlight the role of forests in moderating flood peaks and timing in Alpine watersheds. We found a strong gradient of m values, with forests exhibiting slower transmissivity, which delays the onset of surface runoff. This helps to reduce and delay flood peaks, offering a natural buffer against extreme events. However, as saturation levels increase, this mitigation effect diminishes. We conclude that integrating forest dynamics into watershed management tools is a promising way to assess cost-effective and environmentally sustainable alternatives to conventional engineering approaches.