Assessing dynamic vegetation model parameter uncertainty across Alaskan arctic tundra plant communities

Area of Science:

• Ecology
• Climate Science
• Computational Modeling

Background:

• Arctic ecosystems are rapidly changing due to global warming.
• Terrestrial Biosphere Models (TBMs) are essential for predicting these changes but face uncertainties in their parameters.
• Improving TBMs requires understanding parameter sensitivity and uncertainty.

Purpose of the Study:

• To incorporate an arctic-specific Terrestrial Ecosystem Model (TEM) into the Predictive Ecosystem Analyzer (PEcAn) framework.
• To quantify model sensitivity and uncertainty related to 21 parameters influencing gross primary production.
• To assess how parameter variations affect carbon fluxes and pools across diverse Alaskan tundra types.

Main Methods:

• Utilized the Predictive Ecosystem Analyzer (PEcAn) framework with an arctic-focused Terrestrial Ecosystem Model (TEM).
• Treated model parameters as probability distributions, estimating them from synthesized field data.
• Analyzed sensitivity and uncertainty of carbon fluxes (net primary productivity, heterotrophic respiration) and pools (vegetation C, soil C) across a latitudinal gradient in Alaskan tundra.

Main Results:

• Model sensitivity was highest for parameters regulating photosynthesis temperature.
• Model uncertainty was primarily driven by parameters related to leaf area, photosynthesis temperature regulation, and stomatal light responses.
• Parameter sensitivity and uncertainty varied spatially, with some sites showing broader parameter influence.

Conclusions:

• The study highlights the complexity of parameter uncertainty in heterogeneous Arctic tundra.
• Temperature regulation of photosynthesis and leaf area are key factors influencing TBM predictions in the Arctic.
• The developed framework allows for iterative testing of new field data to improve Arctic change forecasting.