Evaluating litter decomposition in earth system models with long-term litterbag experiments: an example using the...

Area of Science:

• Earth System Science
• Ecology
• Biogeochemistry

Background:

• Decomposition significantly impacts the global carbon budget, yet its simulation in Earth system models remains largely unvalidated.
• Accurate modeling of litter decomposition is crucial for understanding carbon cycle-climate feedbacks.

Purpose of the Study:

• To test the litter decomposition parameterization of the Community Land Model version 4 (CLM4) using data from the Long-Term Intersite Decomposition Experiment Team (LIDET).
• To compare CLM4's performance with DAYCENT, an established ecosystem model, against field observations.

Main Methods:

• Conducted 10-year litter decomposition simulations using CLM4 across diverse North and Central American biomes and litter types.
• Constrained simulations with a climatic decomposition index from LIDET data, accounting for temperature and moisture effects.
• Performed parallel simulations with DAYCENT for comparative analysis.

Main Results:

• CLM4 exhibited a significant discrepancy with LIDET field data, simulating faster carbon loss and higher nitrogen immobilization than observed.
• Reduced decomposition rates, assuming severe soil mineral nitrogen limitation, were required for CLM4 to better match observations.
• DAYCENT demonstrated a closer agreement with observed carbon mass remaining and nitrogen dynamics, irrespective of nitrogen limitation.

Conclusions:

• CLM4's rapid litter decomposition may contribute to biases in simulated soil carbon levels within global Earth system models.
• Terrestrial biogeochemistry components of Earth system models require rigorous validation against real-world observational data.
• Long-term litter decomposition experiments like LIDET serve as essential benchmarks for evaluating and improving ecosystem models.