EGU2020-8004
https://doi.org/10.5194/egusphere-egu2020-8004
EGU General Assembly 2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Projected soil organic carbon stocks in German croplands under different climate change scenarios

Catharina Riggers1, Christopher Poeplau1, Axel Don1, Cathleen Frühauf2, and René Dechow1
Catharina Riggers et al.
  • 1Thünen Institute of Climate-Smart Agriculture, Braunschweig, Germany (catharina.riggers@thuenen.de)
  • 2Deutscher Wetterdienst, ZAMF, Braunschweig, Germany

Mineralization of soil organic carbon (SOC) is driven by temperature and soil moisture. Thus, climate change might affect future SOC stocks with implications for greenhouse gas fluxes from soils and soil fertility of arable land. We used a model ensemble of different SOC models and climate projections to project SOC stocks in German croplands up to 2099 under different climate change scenarios of the Intergovernmental Panel of Climate Change. Current SOC stocks and management data were derived from the German Agricultural Soil Inventory. We estimated the increase in carbon (C) input required to preserve or increase recent SOC stocks. The model ensemble projected declining SOC stocks in German croplands under current management and yield levels. This was true for a scenario with no future climate change (-0.065 Mg ha-1 a-1) as well as for the climate change scenarios (-0.070 Mg ha-1 a-1 to -0.120 Mg ha-1 a-1). Thereby, preserving current SOC stocks would require an increase in current C input to the soil of between 51 % (+1.3 Mg ha-1) and 93 % (+2.3 Mg ha-1). We further estimated that a C input increase of between 221 % and 283 % would be required to increase SOC stocks by 34.4 % in 2099 (4 ‰ a-1). The results of this study indicate that increasing SOC stocks under climate change by a noticeable amount will be challenging since SOC losses need to be overcompensated.

How to cite: Riggers, C., Poeplau, C., Don, A., Frühauf, C., and Dechow, R.: Projected soil organic carbon stocks in German croplands under different climate change scenarios, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8004, https://doi.org/10.5194/egusphere-egu2020-8004, 2020

Comments on the presentation

AC: Author Comment | CC: Community Comment | Report abuse

Presentation version 1 – uploaded on 04 May 2020
  • CC1: Pool quality initialization, Thomas Wutzler, 05 May 2020

    The chat discussion revolved around initial pool changes, that might be artefacts of model initialization. Knowing the stocks from an inventory, but not the distribution across pools, i.e. quality, is a common problem that might play a role here.

    It is well discussed in the review of

    Zhun Mao et al. Modeling soil organic carbon dynamics in temperate forests using Yasso07 Biogeosciences

    https://www.biogeosciences.net/16/1955/2019/bg-16-1955-2019-discussion.html

    • AC1: Reply to CC1, Rene Dechow, 07 May 2020

      In this study we assume indeed steady state conditions when initializing SOC models which might cause artefacts. However, the applied models, carbon input estimations and initialisation methods have been tested using a longterm observational network (139 sites, measured SOC time series of around 20 years each) under common agricultural management. In average, carbon stock changes could be well described by modelled trends although site specific errors were high (Riggers et al. 2019). Since both data sets, the longterm observational network and the agricultural soil inventory (used in this study) represent common managed arable soils (before and during the measurement period) we assumed that the model approach tested on the longterm observational network is also applicable for the sites of the agricultural soil inventory. Modelled C-stock changes (agricultural soil inventory) were in the same range as observed C stock trends (observational network).

      Artefacts by inappropriate initialisation might occur in the first decades of the modelling period but should diminish when longterm SOC changes are addressed since the influence of model initialisation method on modelled SOC stocks decreases with increasing length of the simulation period. Here the modelling period is about 83 years.  In a prior test, we compared recent measured SOC stocks of the agricultural soil inventory with modelled SOC stocks of longterm spinup runs bringing the C stock of ASI sites into equilibrium (based on recent management and weather conditions). Resulting trends and direction of SOC change after establishing an equilibrium compared well with the no climate change scenario. Mean SOC trends in the first two decades of the no climate change scenario were comparable with observed mean annual changes of the longterm observational network. Based on these results I assume that artefacts caused by inappropriate model initialisation are less relevant in this study although the initialisation method based on the partial equilibrium assumption in Mao et al(2019) would be an interesting alternative.

      Riggers et al. (2019) Multi-model ensemble improved the prediction of trends in soil organic carbon stocks in German croplands;