Similarities and differences in the sensitivity of Soil Organic Matter (SOM) dynamics to biogeochemical parameters for different vegetation inputs and climates

by   Giulia Ceriotti, et al.

The biogeochemical complexity of environmental models is increasing continuously and model reliability must be reanalysed when new implementations are brought about. This work aim to identify influential biogeochemical parameters that control the Soil Organic Matter (SOM) dynamics and greenhouse gas emissions in different ecosystems and climates predicted by a physically-based mechanistic model. This explicitly accounts for four pools of organic polymers, seven pools of organic monomers, five microbial functional groups, and inorganic N and C species. We first benchmarked our model against vertical SOM profiles measured in a temperate forest in North-Eastern Bavaria, Germany (Staudt and Foken, 2007). Next, we conducted a sensitivity analysis to biogeochemical parameters using modified Morris indices for target SOM pools and gas emissions from a tropical, a temperate, and a semi-arid grassland in Australia. We found that greenhouse gas emissions, the SOM stock, and the fungi-to-bacteria ratio in the top soil were more sensitive to the mortality of aerobic bacteria than other biogeochemical parameters. The larger CO2 emission rates in forests than in grasslands were explained by a greater dissolved SOM content. Finally, we found that the soil N availability was largely controlled by vegetation inputs in forests and by atmospheric fixation in grasslands


Generalized Autoregressive Score Trees and Forests

We propose methods to improve the forecasts from generalized autoregress...

Inverting Regional Sensitivity Analysis to reveal sensitive model behaviors

We address the question of sensitivity analysis for model outputs of any...

Uncovering dark matter density profiles in dwarf galaxies with graph neural networks

Dwarf galaxies are small, dark matter-dominated galaxies, some of which ...

Direct numerical simulation of the combustion of a suspended droplet in normal gravity

DropletSMOKE++ is a multiphase CFD framework based on OpenFOAM, original...

Biomimetic use of genetic algorithms

Genetic algorithms are considered as an original way to solve problems, ...

On the Validation of Gibbs Algorithms: Training Datasets, Test Datasets and their Aggregation

The dependence on training data of the Gibbs algorithm (GA) is analytica...

Multi-label Classification with Optimal Thresholding for Multi-composition Spectroscopic Analysis

In this paper, we implement multi-label neural networks with optimal thr...

Please sign up or login with your details

Forgot password? Click here to reset