Sensitivity of ozone dry deposition to ecosystem-atmosphere interactions: A critical appraisal of observations and simulations

The response of ozone (O3) dry deposition to ecosystem‐atmosphere interactions is poorly understood but is central to determining the potential for extreme pollution events under current and future climate conditions. Using observations and an interactive dry deposition scheme within two dynamic vegetation land models (GFDL LM3.0/LM4.0) driven by observation‐based meteorological forcings over 1948‐2014, we investigate the factors controlling seasonal and interannual variability (IAV) in O3 deposition velocities (Vd,O3). Stomatal activity in this scheme is determined mechanistically, depending on phenology, soil moisture, vapor pressure deficit, and CO2 concentration. Soil moisture plays a key role in modulating the observed and simulated Vd,O3 seasonal changes over evergreen forests in Mediterranean Europe, South Asia, and the Amazon. Analysis of multi‐year observations at forest sites in Europe and North America reveals drought stress to reduce Vd,O3 by ~50%. Both LM3.0 and LM4.0 capture the observed Vd,O3 decreases due to drought; however, IAV is weaker by a factor of two in LM3.0 coupled to atmospheric models, particularly in regions with large precipitation biases. IAV in summertime Vd,O3 to forests, driven primarily by the stomatal pathway, is largest (15‐35%) in semi‐arid regions of western Europe, eastern North America, and northeastern China. Monthly mean Vd,O3 for the highest year is two to four times that of the lowest, with significant implications for surface O3 variability and extreme events. Using Vd,O3 from LM4.0 in an atmospheric chemistry model improves the simulation of surface O3 abundance and spatial variability (reduces mean biases by ~10 ppb) relative to the widely‐used Wesely scheme.