Peatlands store vast amounts of soil carbon and are significant sources of greenhouse gases, including carbon dioxide (CO2) and methane (CH4) emissions. The traditional approach in biogeochemical model simulations of peatland emissions is to simply divide the soil domain into an aerobic zone above and an anaerobic zone below the water table (WT) and then calculate CO2 and CH4 emissions based on the assumed properties of these two discrete zones. However, there are major potential drawbacks associated with the traditional WT-based approach, because aerobic or anaerobic environments are ultimately determined by oxygen (O2) concentration rather than water content directly. Variations in O2 content above and below the WT can be large and thus may play an important role in partitioning of carbon fluxes between CO2 and CH4. In this paper, we propose an oxygen-based approach, which simulates the vertical and radial components of O2 movement and consumption through the soil aerobic and anaerobic environments. We then use both our oxygen-based and the traditional WT-based approaches to simulate CO2 and CH4 emissions from an Alaskan fen peatland. The results of model calibration and validation suggest that our physically realistic approach (i.e., oxygen-based approach) cause less biases on the simulated flux of CO2 and CH4. The results of model simulations also suggest that the traditional WT-based approach might substantially under-estimate CH4 emissions and over-estimate CO2 emissions from the fen due to the presence of anaerobic zones in unsaturated soil. Our oxygen-based approach can be easily incorporated into existing ecosystem or earth system models but will require additional validation with more extensive field observations to be implemented within biogeochemical models to improve simulations of soil C fluxes at regional or global scale.
- Water table