Modelling water balance components in a temperate forest in Germany: A comparative analysis of pine, oak, and beech

Understanding the water balance components in forests is crucial for sustainable land and water management. The Frankfurt Rhine-Main metropolitan region in Germany is heavily dependent on groundwater, with the Hessian Ried forest being one of the main sources. However, climate change, population growth, continuous land sealing, and the expansion of farmland and irrigation in the region, have increased the pressure on water resources, exacerbating conflicts over water use between the affected sectors. Therefore, the region requires comprehensive solutions for a sustainable and flexible water management.

This study focuses on modelling the water balance components in three monitoring plots located in the Hessian Ried. Each plot is dominated by a different tree species — pine, oak, and beech. The aim of the study is to assess the impact of tree species and soil physical properties on water dynamics and availability. We employed the LWF-Brook90R package for the implementation of the LWF-Brook90 model considering climatic boundary conditions, vegetation parameters and soil physical parameters at different depths. The study covers the period of 2005 to 2023 allowing the assessment of seasonal variations over several years. Moreover, we performed the assessment of different parameter sets and a Bayesian calibration in order to analyse the variations in the resulting water balance components for each plot. We compared our simulations to throughfall and soil water content observations.

Our findings revealed complex interplays between tree species and water balance components, highlighting the importance of species-specific considerations when modelling forests. We obtained a good agreement between our results and observed throughfall, indicated by an R² ≥ 0.7. The different parameter sets and the calibration delivered highly similar statistical indicators of observed versus simulated throughfall. However, the calibration did not improve the throughfall simulations in all cases. Regarding actual transpiration and interception rates, the pine plot exhibited larger variations depending on the parameter set used, compared to oak and beech. Both deciduous stands presented a larger transpiration deficit as water stress indicator compared to the pine plot. The transpiration deficit increased considerably in the three plots after calibrating interception and soil physical parameters, compared to default datasets. Additionally, the simulations of the pine plot resulted in the lowest drainage rates among the plots, due to a combination of factors including the evergreen canopy and predominant sandy soil texture along the entire rooting depth. We achieved a more comprehensive and improved estimation of the soil water content — and consequently soil water storage in the root zone — after calibrating the soil physical parameters in contrast to pre-established soil datasets. This allowed for uncertainties in the estimation of soil water content in the unsaturated zone, which is a key consideration when modelling water balance components.

The insights gained from this research have implications for climate change adaptation and mitigation. As climate patterns shift, understanding how different tree species influence water availability and utilization becomes paramount. The presented models serve as a valuable tool for predicting and managing water resources in diverse forested landscapes, supporting the development of adaptive strategies for sustainable forest management.