Quantifying Carbon Uptake by Vegetation for Europe on a 1km² Resolution

Kurzfassung

With this study we like to introduce our upgraded version of the Biosphere Energy Transfer Hydrology (BETHY/DLR) Model, which is designed for regional to continental applications and was validated with FLUXNET data. The BETHY/DLR model is an adaption of the well known JSBACH model which is frequently used to assess for instance carbon fluxes on a global scale. BETHY/DLR is designed to be driven by remote sensing data. Currently data derived from the SPOT-VEGETATION sensor are used as the land cover classification GLC2000 and leaf area index (LAI) time series. The model also needs meteorological input data, which are derived from the European Center for Medium Range Weather Forecast (ECMWF). Furthermore information about the altitude (SRTM) and a soil map (FAO/IIASA) are used as input. BETHY/DLR is a soil-vegetation-atmosphere-transfer (SVAT) model. SVAT models track the plant-mediated transformation of atmospheric carbon dioxide into energy-storing hydrocarbons such as sugars, a process known as carbon fixation. BETHY/DLR models photosynthesis, and takes into account environmental conditions that affect it. It separately treats the light and dark reactions of photosynthesis at the leaf level. A benefit of this design is that the photosynthetic rate can be mechanistically limited either by light availability or by the abundance of the carboxylation enzyme Rubisco, the key player in the Calvin cycle that fixes carbon. In addition, so-called C3 and C4 plants are distinguished in BETHY/DLR because significant differences exist between their carbon fixation physiologies; in particular, C4 plants such as sugar beet and corn can fix more atmospheric CO2 at higher temperatures than can C3 plants such as barley and wheat. Besides photosynthesis, other energy transfers also need to be tracked. BETHY/DLR’s energy balance model takes into account heat fluxes between the vegetation and the atmosphere above it, as well as the cooling effect of evapotranspiration from the soil and vegetation. Soil heat flux is also estimated, and the storage of heat in the canopy and in the air layer above the canopy is considered. Autotrophic respiration is modelled in BETHY/DLR as the sum of maintenance respiration and growth respiration. Maintenance respiration is determined by plant-specific dark respiration rates, while growth respiration is proportional to the difference between Gross Primary Productivity (GPP) and maintenance respiration. For estimating Net Ecosystem Productivity (NEP), heterotrophic soil microbe respiration is calculated as a function of temperature, scaled by the annual Net Primary Productivity (NPP) and the effect of soil moisture. The output of BETHY/DLR are time series of GPP and NPP in daily time steps, at the resolution and projection of the land cover classification. Here 1 km² resolution is used, in a latitude-longitude projection using the WGS84 (World Geodetic System 1984) datum. In order to assess the quality of the model performance a model run for the European continent for the years 2000 to 2010 was done and compared with time series of estimated GPP from about 100 FLUXNET tower sites. For each FLUXNET site we used the level 4 product of calculated GPP given as annual and monthly sums for all measured years. On an annual scale we found underestimation by BETHY/DLR of 20%-30% for forest sites and slightly better agreement for agricultural sites. However we found an agreement for most of the sites when comparing the course of monthly sums but with an offset. For most of the sites we found a RMSE (root mean square error) of not higher than 30% which is in good agreement with other vegetation models, such as BIOME-BGC, ORCHIDEE and LPJ.