| 17 Aug 2026
Impacts of ozone-vegetation feedbacks on surface ozone in a coupled climate-chemistry-ecosystem model
Abstract. Surface ozone (O3) is strongly influenced by interactions between atmospheric chemistry and terrestrial ecosystems, yet these feedbacks remain insufficiently represented in Earth system models. Here, we develop a fully coupled O3-vegetation interaction framework by integrating the terrestrial ecosystem model iMAPLE into the ECHAM-HAMMOZ Earth system model. The coupled model explicitly represents three major land-atmosphere interaction pathways, including dynamic biogenic volatile organic compound (BVOC) emissions, vegetation-mediated O3 dry deposition, and O3-induced vegetation damage. These feedback pathways exert heterogeneous impacts on surface O3 concentrations due to spatial variations in vegetation characteristics and nonlinear chemical sensitivities. Changes in monoterpene emissions cause the largest global effect with widespread reductions in surface O3, whereas isoprene-driven responses vary among regions with different chemical sensitivities. In regions with high anthropogenic emissions, such as East Asia, vegetation-mediated changes in BVOCs and dry deposition substantially amplify O3 responses. The O3-induced vegetation damage further modifies atmospheric chemistry by reducing BVOC emissions through leaf area loss while increasing surface O3 by inhibiting stomatal conductance. Incorporating these processes decreases simulated global surface O3 concentrations and alleviates the positive bias in the original model, particularly over China (relative mean bias decreases from 31.5 % to 29.7 %) and the U.S. (from 19.0 % to 7.8 %). These results demonstrate that interactive O3-vegetation feedbacks are essential for improving the representation of atmospheric composition and ecosystem–atmosphere interactions in Earth system models.
The study by Zhao et al. investigates ozone-vegetation feedbacks by coupling iMAPLE with ECHAM-HAMMOZ. The model includes dynamic isoprene and monoterpene emissions, vegetation-driven ozone dry deposition, and ozone-induced vegetation damage. A sequence of sensitivity experiments is then used to separate the effects of these processes on surface O3 and ecosystem productivity. The results show that the response of O3 strongly depends on the chemical regime, with BVOC changes producing different, and in some regions opposite, responses under NOx-limited and VOC-limited conditions, while ozone damage also feeds back through changes in GPP, LAI, BVOC emissions, and dry deposition.
The manuscript is well written, easy to follow, and addresses an important topic. The stepwise model experiments are useful, and the authors also acknowledge several limitations. However, the novelty should be stated more clearly, since fully coupled ozone-vegetation feedbacks have already been explored in previous modelling studies. The specific advance of the present framework should therefore be better defined. I find the paper useful for the literature and would recommend publication after considering the following points.
General comments
Specific comments
L23-24: This should mention the chemical-regime dependence directly in the abstract, since the sign of the response is not universal.
L29-31: The improvement over China from RMB 31.5% to 29.7% appears modest. Please provide some measure of variability or uncertainty, or otherwise discuss whether this difference is meaningful.
L71-76: The statement that current models generally prevent a “fully two-way feedback” should be softened and better positioned relative to previous work. The novelty here seems to lie more in the specific coupled implementation and process attribution than in two-way coupling itself.
L123-124: “isoprene and monoterpenes, the two dominant BVOC species” is too simplified. They are major BVOC classes, but the wording ignores other highly reactive compounds, especially sesquiterpenes and oxygenated VOC. I suggest rephrasing this and acknowledging the restricted BVOC representation.
L134-140: Given that the MTP pathway becomes the dominant global O3 response, more discussion is needed on the uncertainty associated with the simplified temperature-dependent monoterpene emission scheme. Recent work has shown that monoterpene temperature sensitivity varies substantially among vegetation types and that alternative parameterizations can significantly affect simulated atmospheric chemistry, including ozone (https://www.nature.com/articles/s43247-023-01175-9). The authors could discuss how uncertainty in this temperature response may influence their results.
L274. Probably a typo (2-6ppb instead of 26ppb)
L296: A global mean O3 increase of 0.11 ppbv is extremely small. I would avoid presenting this as a result without some context on variability or uncertainty.
L394-408: This section is useful, but I would add explicitly the restricted BVOC composition, especially the omission of SQTs and other reactive/stress-related VOCs, and the possible implications for O3 chemistry.
L405-408: Since the authors themselves note the importance of temperature and extreme events, I would encourage them to discuss whether high-O3/high-temperature episodes could reveal stronger feedbacks than the seasonal averages used here.