Changing Stomatal Conductance in Response to Anthropogenic Climate Change: a Model-Data Comparison

Abstract:

Understanding the response of stomatal conductance to increasing anthropogenic CO₂ is of critical importance for the evaluation of the global hydrological cycle, and has implications for future climate change, particularly flood and risk. Via physiological and morphological changes in leaf structure, stomatal conductance has been shown to decrease under elevated CO₂ conditions, stimulating higher vegetation water use efficiency and water retention within the land system. However, an assessment of how the simulation of changing stomatal conductance in an Earth System Model compares with observational data under varying CO₂ concentrations (Free Air CO₂ Enrichment - FACE - studies) has not yet been conducted, and is crucial for considering the significance of the issue of flooding for global policy making. Here we utilise the Community Land Model, Version 4.5 (CLM4.5), performing climate simulations over a range of atmospheric CO₂ concentrations from 350ppm to 700ppm (50ppm increments), and compare the model data with results from a conclusive set of FACE studies (350ppm – 700ppm range). We show general agreement between the climate model and FACE data with regards to decreasing stomatal conductance in response to increasing CO₂ concentration. However the magnitudes of the conductance changes differ, spatially within the model, and across species in the FACE studies. We show how the model can be used to assess global stomatal conductance changes on spatial scales where FACE cannot (ie. most FACE studies are located across the 30°- 60° latitude belt). This is useful for understanding the role of groundwater retention and potential flood risk at regional scales across the globe. Additionally the model demonstrates seasonal-spatial changes in stomatal conductance in response to CO₂ forcing. Such seasonal changes are generally absent throughout FACE studies, as measurement is predominantly tasked during summer months. Our results reiterate the importance of climate models in the assessment of future vegetation induced flood risk, especially since FACE studies are thus far limited to maximum CO₂ concentrations of ~700ppm.