Evaluating SoilGen2 as a tool for projecting soil evolution induced by global change

Area of Science:

• Soil Science
• Environmental Modeling
• Predictive Soil Science

Background:

• Predicting soil evolution over decadal to centennial timescales is crucial for managing soil threats from human activities and global change.
• Mechanistic soil evolution models offer a valuable tool for these long-term projections.

Purpose of the Study:

• To assess the SoilGen model's capability in simulating soil characteristic changes relevant to various soil threats.
• To evaluate the model's sensitivity to initial and boundary conditions for predicting soil evolution.

Main Methods:

• Functional sensitivity analysis was employed, varying initial conditions (parent material properties) and boundary conditions (climate, fertilization, tillage, agriculture duration).
• Simulations included anthroposequences in Luvisols across two sites, with multiple variants for initial and boundary conditions to reflect uncertainties.

Main Results:

• The SoilGen model demonstrated sensitivity to climate and agricultural practices for all simulated soil properties.
• Long-term simulation results were less affected by boundary condition uncertainties but were influenced by initial condition uncertainties.
• Organic carbon projections yielded the best results, while bulk density and <2μm fraction patterns require model improvements.

Conclusions:

• The SoilGen model, after calibration for organic carbon under agricultural use, is suitable for projecting soil responses to global change.
• Enhancements, such as incorporating dynamic vegetation growth and refining the treatment of clay formation and bulk density, would improve model performance.