Emergence of the physiological effects of elevated CO₂ on land–atmosphere exchange of carbon and water

Abstract

Elevated atmospheric CO 2 (eCO 2 ) influences the carbon assimilation rate and stomatal conductance of plants, thereby affecting the global cycles of carbon and water.Yet, the detection of these physiological effects of eCO 2 in observational data remains challenging, because natural variations and confounding factors (e.g., warming) can overshadow the eCO 2 effects in observational data of real-world ecosystems. In this study, we aim at developing a method to detect the emergence of the physiological CO 2 effects on various variables related to carbon and water fluxes. We mimic the observational setting in ecosystems using a comprehensive process-based land surface model QUINCY to simulate the leaf-level effects of increasing atmospheric CO 2 concentrations and their century-long propagation through the terrestrial carbon and water cycles across different climate regimes and biomes. We then develop a statistical method based on the signal-to-noise ratio to detect the emergence of the eCO 2 effects. The eCO 2 effect on gross primary productivity (GPP) emerges at relatively low CO 2 increase (∆[CO 2 ] ~ 20 ppm) where the leaf area index is relatively high. Compared to GPP, the eCO 2 effect causing reduced transpiration water flux (normalized to leaf area) emerges only at relatively high CO 2 increase (∆[CO 2 ] >> 40 ppm), due to the high sensitivity to climate variability and thus lower signal-to-noise ratio. In general, the response to eCO 2 is detectable earlier for variables related to the carbon cycle than the water cycle, when plant productivity is not limited by climatic constraints, and stronger in forest-dominated rather than in grass-dominated ecosystems. Our results provide a step toward when and where we expect to detect physiological CO 2 effects in in-situ flux measurements, how to detect them and encourage future efforts to improve the understanding and quantification of these effects in observations of terrestrial carbon and water dynamics.

Elevated aerosol enhances plant water‐use efficiency by increasing carbon uptake while reducing water loss

“…Shapley values, originating from game theory, can partition the relative influence of each factor on the responsive variable. This approach offers a machine learning analog of partial regression for diagnosing the impact of individual factor (such as TotalR, DiffuseR, and VPD here) on RGR/SFD based on Shapley values (Maynard et al ., 2022; H. Wang et al ., 2022; Zhan et al ., 2022). Here, we used the R package ‘ fastshap ’ to calculate the Shapley values for each predictor in the RF model, using observations from cloudless days as input.…”

Estimating the CO2 Fertilization Effect on Extratropical Forest Productivity From Flux‐Tower Observations

“…The model setup and model simulations are identical with the “freeze‐CO 2 experiment” in Zhan et al. (2022). For better comparison with EC records, we take the last 20 years (1999–2018) in the simulations as the time period of the validation.…”

Estimating the CO2 fertilization effect on extratropical forest productivity from Flux-tower observations

“…The global network of eddy covariance (EC) flux towers observing the exchange of CO2 at the ecosystem scale provides a valuable resource, as it has accumulated sufficiently long time series data to potentially provide direct evidence of the CO2 fertilization effect (Knauer et al, 2017;Baldocchi, 2020;Zhan et al, 2022). Previous studies have attempted to attribute the CO2 fertilization effect on GPP by utilizing EC records.…”

“…We perform two simulations with identical climate but varying CO2 concentrations (transient CO2 as observed, and constant CO2 at levels of the year 1988) at 166 forested sites distributed across the globe. The model setup and model simulations are identical with the "freeze-CO2 experiment" inZhan et al (2022). For better comparison with eddy covariance records, we take the last twenty years(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018) in the simulations as the time period of the validation.…”