AAAR 33rd Annual Conference
October 20 - October 24, 2014
Rosen Shingle Creek
Orlando, Florida, USA
Abstract View
Understanding the Sensitivity of SOA Formation to Various Uncertain Modeling Parameters Using a Variance-Based Statistical Approach
MANISHKUMAR SHRIVASTAVA, Chun Zhao, Yun Qian, Richard Easter, Alla Zelenyuk, Jerome Fast, Pacific Northwest National Laboratory
Abstract Number: 444 Working Group: Aerosol Chemistry
Abstract Several physical and chemical processes affect the formation of secondary organic aerosol (SOA), one of the most important but uncertain fraction of fine particles in the atmosphere. Models typically select a set of parameters to represent various processes such as dry deposition, the emissions of volatile organic compounds (VOCs), NOx, SOA precursor emissions and SOA yields. But all these parameters have a big range of uncertainty, which are not accounted in the models. In addition, there are complex non-linear interactions among these parameters e.g. changing the emissions of volatile organic compounds (VOCs) and NOx would change oxidant fields and NOx regimes, which affect SOA yields. Also, the sensitivities of SOA formation to VOC and NOx regimes may depend on the properties of SOA such as its low volatility, phase and viscosity. In this study, we investigate the sensitivity of SOA to seven parameters related to the emissions of anthropogenic and biogenic VOC and NOx, as well as anthropogenic semi-volatile and intermediate volatility species (SIVOC), phase and volatility changes of SOA particles, and dry deposition parameters of SOA precursors. We perform an ensemble of 128 simulations using the Weather Research and Forecasting Model coupled to Chemistry (WRF-Chem), and simulate SOA using our modified volatility basis-set (VBS) approach over the domain where the CARES field study in Sacramento, CA was conducted in June 2010. We adopt a quasi-Monte Carlo (QMC) sampling approach to effectively sample the high-dimensional parameter space and conduct a variance-based sensitivity analysis to quantify the contribution of each parameter to the overall SOA loading uncertainty. We also characterize the spatial and temporal statistical significance of these parameters. Finally we investigate how the interactions between these parameters affect the variance of SOA formation in the atmosphere. Results are expected to provide insights that would help better constrain the SOA modeling to improve model-measurement agreement.