Evaluation of emission control strategies for regional scale air quality: Performance of direct and surrogate techniques
S.S. Isukapalli, T. Kindap, A. Vemuri, S.W. Wang, P.G. Georgopoulos
Environmental and Occupational Health Sciences Institute, Piscataway, NJ
Evaluation of emissions control strategies often involves several variables such as region-specific and source category-specific controls, and requires several photochemical air quality model simulations for combinations of various potential controls. Different sensitivity analysis techniques that can be applied include direct techniques such as the Direct Decoupled Method (DDM) and the Automatic Differentiation (AD), as well as surrogate techniques such as the High Dimensional Model Representation (HDMR). These provide alternative means for assessing model response without requiring simulations for each combination of different variables. The DDM requires addition of equations for the sensitivity calculations to the original model source code, and relies on local gradients. The AD technique is based on automatic processing of the model source code and also relies on local gradients. In contrast, the HDMR technique can be applied without alteration to the model and relies on a global response spanning the ranges of different dimensions. In this study, these sensitivity methods are comparatively evaluated for assessing the model response in the context of emission reductions of VOCs and NOx on ozone and particulate matter levels using the CMAQ model for ten states in Mid-Atlantic and North East US. The different sensitivity techniques are compared for the quality of the sensitivity measures such as gradients. The estimates of these methods are also compared with the full model simulations at multiple combinations for addressing the quality of the response estimation and source apportionment under different control strategy options.
This work is funded in part by the USEPA (Cooperative Agreement CR-83162501) and the New Jersey Department of Environmental Protection (NJDEP, through base funding to the Ozone Research Center). This work has not been reviewed by and does not represent the opinions of the funding agencies.