Improved estimation of global gross primary productivity during 1981-2020 using the optimized P model

Area of Science:

• Earth and Environmental Sciences
• Ecology
• Biogeochemistry

Background:

• Accurate terrestrial gross primary productivity (GPP) estimation is vital for understanding global carbon cycling.
• Light use efficiency (LUE) models are common for GPP estimation but face challenges with maximum LUE (LUEmax) determination and downregulation.
• The P model, a photosynthesis-based model, offers improved mechanistic understanding and simplified parameterization compared to traditional LUE models.

Purpose of the Study:

• To evaluate the effectiveness of five water stress factors within the P model for global GPP estimation.
• To develop and validate a new long-term global monthly GPP dataset (PGPP) using the optimized P model.
• To compare the performance of the new PGPP dataset against existing GPP products.

Main Methods:

• The P model was optimized by comparing the effectiveness of five integrated water stress factors.
• A new global monthly GPP dataset (PGPP) was generated at a 0.1° × 0.1° resolution for the period 1981-2020.
• The PGPP dataset was validated using data from 109 globally distributed FLUXNET sites.

Main Results:

• The optimized P model demonstrated superior performance in estimating GPP.
• Validation showed PGPP outperformed three widely-used GPP products, with R² = 0.75, RMSE = 1.77 g C m⁻² d⁻¹, and MAE = 1.28 g C m⁻² d⁻¹.
• Analysis revealed a significant increase in PGPP across 69.02% of global vegetated regions between 1981 and 2020.

Conclusions:

• The PGPP dataset offers a valuable new resource for global ecological research and carbon cycle studies.
• The comparison of water stress factors provides insights for future improvements in GPP modeling.
• The P model shows promise for accurate global GPP estimation, advancing our understanding of terrestrial ecosystem responses.