Our regional air quality forecast modeling system, AIRPACT-5, consistently underestimate surface level fine particulate matter (PM2.5) concentrations in the summer at both urban and rural locations in the Pacific Northwest, primarily result of errors in organic particulate matter. A number of chamber and field measurements have shown that atmospheric organic aerosols produced from wildfires are significantly underestimated in the emission inventories used for air quality models for various applications such as regulatory strategy development, impact assessments of air pollutants, and air quality forecasting for public health. In this work, we implement a) improved chemical speciation and emission factors, based on FLAME-IV (fourth Fire Lab at Missoula Experiment) and other lab/field measurements, and b) updated parameterizations of secondary organic aerosols (SOA) formation and aging mechanisms into the AIRPACT-5 modeling system. We test our model for two major fire events occurred in the Pacific Northwest in summer 2013: the Big Windy Complex Fire and the Colockum Traps Fire. We compare model results from CMAQ version 5.2 (CMAQv5.2) which has a better treatment for anthropogenic SOA formation (as a base case) and modified parameters used for the SOA module in the model to parameterize physic-chemical properties of SOA from fire sources (fire-soa case) based on findings from Barsanti et al. (2013). Using the observed aerosol chemical composition and mass loadings for organics, nitrate, sulfate, ammonium, and chloride from the aircraft during the Studies of Emissions and Atmospheric Composition, Clouds, and Climate Coupling by Regional Surveys (SEAC4RS) and the Biomass Burning Observation Project (BBOP), we assess how new knowledge gained from wildfire measurements improve model predictions for SOA and its contribution to the total mass of PM concentrations.