Project leaders: Prof. Dr. Doerthe Tetzlaff (IGB Leibniz Institute of Freshwater Ecology and Inland Fisheries), Prof. Dr. Gerald Jurasinski (University of Greifswald), Prof. Dr. Julia Pongratz (Ludwig Maximilians University München)

Although the concept that peatlands are closely linked to and embedded in their landscapes is old, the quantification of energy, material, and information flows, based on a thorough understanding of the associated processes involved is relatively new. For example, there is still limited understanding of how physical and biogeochemical processes and the microclimate in fens influence important ecosystem functioning of entire peatland landscapes. It is also fairly unclear how global change will affect the peatland capacities in terms of adaptation and mitigation. While projects C1, C5 and A7 focus on specific scales and on the processes the models used in these projects were specifically developed for, synthesis project S4 will integrate these processes and feedbacks across and beyond individual models’ answering ORQ4: How do rewetted peatlands interact with and feed back to the landscape and beyond? In S4, we will integrate novel data obtained from the monitoring in WETSCAPES2.0 into process-based modeling approaches, with improved model parametrisation via data assimilation products and machine learning approaches to quantify physical and biogeochemical process interactions and feedbacks between rewetted peatlands and the landscapes they are imbedded in. These feedback mechanisms include ecohydrological fluxes (such as transpiration, evaporation, lateral subsurface and surface fluxes), surface energy fluxes (such as from altered roughness length and bowen ratio), and GHG fluxes with the novel, improved, integrated modeling of S4 allowing us to quantify these feedbacks across spatial scales. Thus, S4 aims to quantify the heterogeneous spatio-temporal process patterns and dynamics in restored wetscapes as well as their potentials of climate change mitigation and adaptation at various scales. The specific objectives are: (1) to quantify physical and biogeochemical process-interactions and feedbacks between peatlands, their landscapes and beyond, (2) to identify the challenges and uncertainties in quantifying these feedbacks at larger areas and across scales via multiple, integrated approaches; (3) to use process-based models as learning frameworks by informing and enhancing data collection and experimental setup schemes as well as model structures to reduce uncertainties in the future. Different process-based modeling frameworks (details in C1, C5 and A7) will be integrated to assess the consequences of peatland rewetting from local to regional scale. We are neither proposing interactive model coupling nor ensemble modeling, but rather take advantage of the individual modeling strengths and foci, while improving model parametrization through collaboration and exchange both with the different projects actually measuring on the sites as well as the modeling projects, and making use of the experimental WETSCAPES2.0 sites. The site-level simulations of peat formation and decomposition (A7, S2) will complement the assessment of GHG fluxes and climate impacts providing process insights to improve the landscape-scale ecohydrological model (C1) and the regional-scale dynamic vegetation model (C5). The tracer-based landscape-scale model (C1) provides simulated quantification of water fluxes (E, T, Q, GW recharge) and stores (interception, storages). C5 provides (simulated results on) biomass productivity, vegetation CO2 fluxes, albedo, roughness length, heat and land surface energy fluxes. Linked to the landscape-scale ecohydrological model (C1), the calibrated DAMM-GHG model from A7 will be forced by landscape-scale differences in soil moisture to compare GHG emissions from rewetted peatlands and the surrounding landscapes. C1 will further use the local climate data from C5. It is important to note and to distinguish between the projects A7, C1 and C5: the actual processes will be quantified in these respective projects, but to be able to quantify the feedbacks across scales and between fens and landscapes, an integrated approach like the one proposed in S4 is crucial. Close collaboration during parametrization will ensure maximum consistency between the models. To improve model parametrization, S4 will further incorporate spatially and temporally highly resolved information on biophysical conditions from Earth Observation (B6, C4, S3: e.g. soil moisture, vegetation type/fractions, biomass, state variables). Through these inter-linked modeling experiments, we can improve model parametrization moving beyond the use of literature values (which is the common approach) and holistically assess GHG as well as water and energy fluxes. Parametrization will further be improved by Machine Learning Models, used to extrapolate sparse parameter sets in time and/or space (closely collaborating with S3). These modelling experiments will provide the basis for a later assessment of the importance of temperate fen rewetting for the global climate system and sets up the process-based modeling framework for the explicit consideration of different land use management scenarios in phase 2 of WETSCAPES2.0. Further, S4 has a strong focus on upscaling and testing which parametrization is required at coarser grid scales of the regional climate modeling systems. Therefore, a major added value from S4 derives from linking peatland landscape scales with regional scales as well as linking water and energy fluxes and water storage with vegetation and peat dynamics. S4 will fuse process-based modelling and machine learning models with findings from field and remote sensing data (from the relevant projects), and thus, provide an integrated peatland-landscape perspective on ecosystem functioning and feedbacks to local and regional climates when temperate fens are rewetted or land use is altered to investigate the potential (extreme) impacts of land use decision making. The resulting novel understanding gained by S4 will be cross-fertilized with insights from synthesis projects S1-S3 to reconcile our process understanding of peatland rewetting, from the micro to the macro scale.