One of the main sources of uncertainty during the design and operation of wind energy projects is associated with our current limited ability to predict wind and turbulence at spatial and temporal scales relevant to wind turbine operation, particularly over complex terrain. Many mountainous regions with high wind energy potential (including the Jura region in Switzerland and the Carpathian Mountains in Romania) are characterized by multi-scale variability of land surface properties (topography and vegetation cover), which strongly affects the spatial distribution of wind and turbulence and, in turn, wind-turbine performance. Despite the recent efforts to develop high-resolution eddy-resolving flow simulation techniques such as Large-Eddy Simulation (LES) for assessing wind energy projects, their application to mountainous regions is still in its infancy. In order to be effective, LES needs to be properly coupled with high-resolution information of the relevant land surface properties, namely topography, aerodynamic surface roughness, and vegetation structure of tall canopies. This information could potentially be obtained using the latest advances in wavelet-based multi-resolution digital terrain modeling and vegetation modeling. The proposed research aims at developing and assessing a framework that integrates terrain and vegetation modeling concepts and tools in support of accurate wind modeling for wind energy applications over complex terrain. To achieve this, we propose a multidisciplinary approach that consists of coupling the following main modeling elements: (1) a new-generation tuning-free Large-Eddy Simulation technique for high-resolution predictions of wind and turbulence over complex terrain, with and without wind farms; (2) very high resolution Digital Elevation Models linked with novel, wavelet-based generalization and filtering techniques to provide description of the surface properties at the relevant scales; and (3) landscape and vegetation models to predict the potential feedbacks between atmospheric boundary layer processes (fluxes), as affected by the wind farms, and vegetation patterns. The resulting modeling framework will be applied to two case study areas for which high resolution terrain data will be available: one in the Swiss Jura region, and the other in the Romanian Carpathians. The proposed new modeling framework is expected to be a powerful tool for optimizing the design and operation of wind farms. In particular, it will be useful to maximize wind energy production and minimize fatigue loads (and associated maintenance costs) in wind farms. It will also allow us to study the effects of wind farms on land-atmosphere exchanges and fluxes of momentum, heat and water vapor, which are expected to affect the near-surface air temperature and moisture and, in general, the local meteorology. (...)
LPJmL-FIT is process- and trait-based vegetation model. It is a subversion of the model LPJmL simulating patches of competing individual trees with flexible functional traits and empirically derived relations between these traits. Trait composition, productivity and stability of a forest are a result of environmental and competitive filtering. A detailed description of LPJmL-FIT (basic features and differences to LPJmL) is given by Sakschewski et al. (Sakschewski et al., 2015, https://doi.org/10.1111/gcb.12870). LPJmL-FIT was originally developed for tropical forests and has been adapted to European forests by Thonicke et al. (2020).
The data covers southern and central Europe (29.5°N – 62°N, 11°W – 36°O) on a spatial resolution of 0.5° covering the years 1901-2013. Tree height, leaf and stem traits (specific leaf area, wood density, leaf longevity; aggregated), individual traits of simulated trees, vegetation distribution (foliage projected cover, FPC), vegetation carbon and fire carbon emissions are given on an annual basis. Gross primary productivity is provided monthly. Tree height and leaf and stem traits are biomass weighted.
For evaluation of the dataset R- and MATLAB-scripts are provided.
An overview and description of all variables are found in the file description.