Uncertainty in representation of ecosystem processes in Europe by the Community Land Model v5
- Agrosphere, IBG-3, Forschungszentrum Jülich, Jülich, Germany
Understanding hydrological and biogeochemical ecosystem process variability in response to a changing climate is important to improve land surface models and assess current and future states of ecosystem functioning. However, the representation of spatial heterogeneity of ecosystem processes in state-of-the-art land-surface models has not been evaluated thoroughly until today. Here we compare gross primary production (GPP) and evapotranspiration (ET) simulated by the Community Land Model version 5 (CLM5) for the period 1995-2018 over the Euro-CORDEX domain with in-situ data from eddy-covariance sites as well as remote sensing and reanalysis data. Additionally, we conducted a parameter sensitivity analysis to identify the impact of uncertainty coming from ecosystem parameters (in particular default parameters for given plant functional types) for selected sites in Europe.
Our results show that GPP and ET variation across hydroclimates show in general a good agreement between CLM5 and remote sensing and reanalysis products. However, both GPP and ET simulated by CLM5 show large differences with measured in-situ data, depending on the ecosystem type. Further, we identify sensitive parameters that will be adjusted to improve ecosystem representation in CLM5 in a future study. This work is important to improve land surface models and parameterization of plant functional types to understand and improve predictions of ecosystem functioning.
How to cite: Poppe Terán, C., S. Naz, B., Hendricks-Franssen, H.-J., and Vereecken, H.: Uncertainty in representation of ecosystem processes in Europe by the Community Land Model v5, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-13465, https://doi.org/10.5194/egusphere-egu23-13465, 2023.