Nitrification, denitrification, and competition for soil N: Evaluation of two Earth System Models against observations.

Earth System Models (ESMs) have implemented nitrogen (N) cycles to account for N limitation on terrestrial carbon uptake. However, representing inputs, losses, and recycling of N in ESMs is challenging. Here, we use global rates and ratios of key soil N fluxes, including nitrification, denitrification, mineralization, leaching, immobilization, and plant uptake (both NH and NO ), from the literature to evaluate the N cycles in the land model components of two ESMs. The two land models evaluated here, E3SM Land Model version 1 (ELMv1)-ECA and CLM5.0, originated from a common model but have diverged in their representation of plant-microbe competition for soil N. The models predict similar global rates of gross primary productivity (GPP) but have approximately two-fold to three-fold differences in their underlying global mineralization, immobilization, plant N uptake, nitrification, and denitrification fluxes. Both models dramatically underestimate the immobilization of NO by soil bacteria compared with literature values and predict dominance of plant uptake by a single form of mineral nitrogen (NO for ELM, with regional exceptions, and NH for CLM5.0). CLM5.0 strongly underestimates the global ratio of gross nitrification:gross mineralization and both models are likely to substantially underestimate the ratio of nitrification:denitrification. Few experimental data exist to evaluate this last ratio, in part because nitrification and denitrification are quantified using different techniques and because denitrification fluxes are difficult to measure at all. More observational constraints on soil nitrogen fluxes such as nitrification and denitrification, as well as greater scrutiny of the functional impact of introducing separate NH and NO pools into ESMs, could help to improve confidence in present and future simulations of N limitation on the carbon cycle.