Sensitivity Analysis of Nested Photochemical Simulations

Conclusions

It is apparent that the evolution of the concentration fields, and thus, their predictability using a numerical model, is strongly influenced by the uncertainty in boundary concentration fields and the emission inventories. In this work, we applied an automatic differentiation tool (ADIFOR) to investigate the sensitivity of model outputs with respect to perturbations in the boundary conditions of 25 species and emission rates of 10 species. Then, we computed the relative importance of the most important perturbations in the system and in O₃ fields. The inserted perturbations had a magnitude similar to observational uncertainties.

The tangent linear model solutions, which describe the evolution of perturbations along trajectories of a time-dependent non-linear base state, represent well the corresponding non-linear perturbation fields accurately for the modelled period, which included the 98%tile O₃ concentration. This type of sensitivity analysis identifies, for the same change in each variable, which variable may include a larger forecast error at some specific time. Although this sensitivity information does not include non-linear effects, it does provide the fundamental characteristics of the changes in solution behaviour and forecast error. All simulations demonstrate that the highest sensitivity in an atmospheric system arises from perturbations at the boundary conditions of O₃.