How to upscale reaction-diffusion models

Denitrification is an important microbial process and potential source for nitrous oxide (N₂O), an important greenhouse gas and destructor of stratospheric ozone. Quantitative predictions of denitrification in soils are difficult as denitrification, an anaerobic respiration process,  appears to occur even in well-areated soils. It is assumed that denitrification is active in dense aggregates over short time periods - so called hot spots and hot moments. While soil microbial metabolism occurs at the pore scale, the interest in denitrification is mostly at the field and landscape scale. Simulating both scales simultaneously is not feasible. Therefore, denitrification has to be upscaled from the pore to the aggregate scale without losing essential properties of the aggregates. An important key to effectively upscale is the anaerobic aggregate volume fraction.

In order to compare different upscaling techniques we conducted pure culture experiments with varying spatial structures. To avoid confounding effects associated with the transition of bacteria switching from oxic respiration to denitrification, we used bacteria only capable of the former. The investigated upscaling techniques include simplifying the microbial reaction as well as creating an effective one-dimensional diffusion model. We compare computation intensity and approximation quality of experimental results.