Evaluation of ozone deposition models over a subalpine forest in Niwot Ridge, Colorado

In this study, we evaluated three conceptually similar ozone gas deposition models. These dry deposition models are frequently used with chemical transport models for calculations over large spatial domains. However, large scale applications of surface-atmosphere exchange of reactive gases require modeling results as accurate as possible to avoid nonlinear accumulation of errors in the spatially representative results. In this paper, model evaluation and comparison against measured data over a coniferous forest at Niwot Ridge AmeriFlux site (Colorado, USA) is carried out. At this site, no previous model calibration took place for any of the models, therefore, we can test and compare their performances under similar conditions as they would perform in a spatial application. Our results show systematic model errors in all the three cases, model performance varies with time of the day, and the errors show a pronounced seasonal pattern as well. The introduction of soil moisture content stress in the model improved model performance regarding the magnitude of fluxes, but the correlation between measured and modeled ozone deposition values remains low. Our results suggest that ozone dry deposition model results should be interpreted carefully in large scale applications, where the accuracy can vary with land cover sometimes are biased.