Comparing optimal and empirical stomatal conductance models for application in Earth system models

Area of Science:

• Earth system science
• Ecology
• Plant physiology

Background:

• Land surface models (LSMs) use canopy conductance to simulate energy, water, and CO2 fluxes.
• Traditionally, canopy conductance models are semi-empirical, while newer models use optimization theory.
• The predictive advantage of optimization-based models over empirical ones remains untested.

Purpose of the Study:

• To compare the performance of a widely used empirical canopy conductance model (BB) and a newer optimization-type model (MED).
• To evaluate the necessity of species-specific parameterization for these models, particularly concerning CO2 sensitivity.
• To develop a new method for global parameterization of canopy conductance models using leaf physiological data.

Main Methods:

• Field data from diverse forest systems were used to compare BB and MED models.
• Coupled photosynthesis-conductance models and the CLM5 LSM were employed for simulations.
• Leaf δ13C data were utilized to derive intercellular to ambient CO2 concentration ratios (ci/ca) for global parameterization.

Main Results:

• Both BB and MED models showed similar performance in simulating canopy conductance, leaf gas exchange, and ecosystem fluxes.
• Key model parameters (g1B and g1M) demonstrated significant species specificity and CO2 sensitivity.
• Failure to account for CO2 sensitivity led to overestimations of evapotranspiration in CLM5 simulations under high-CO2 conditions.

Conclusions:

• The choice between BB and MED models may not significantly impact current LSM simulations of water and carbon fluxes.
• Species-specific parameterization, especially CO2 sensitivity, is critical for improving the accuracy of LSMs.
• A novel, globally applicable parameterization method for BB and MED models was developed, linked to leaf physiology and precipitation patterns.