Open Access
Chen, Shuang
Graduate Program:
Master of Science
Document Type:
Master Thesis
Date of Defense:
July 18, 2008
Committee Members:
  • William Henry Brune, Thesis Advisor
  • OH
  • HO2
  • radical
  • chemical mechanism
  • comparison
  • modeling
  • atmosphere
Several chemical mechanisms have been developed to describe the chemistry of troposphere. Although model calculations using different mechanisms have been compared before, they have not also been compared to measurements during a field campaign. In this study, a comparison of several widely used mechanisms (RACM, CB05, LaRC, SAPRC-99, SAPRC-07, and MCMv3.1) has been conducted based on TexAQS-2006 field data. The concentrations of hydroxyl (OH) and hydroperoxy (HO2) radicals (called HOx collectively) were calculated by a zero-dimensional box model with each mechanism and then compared with OH and HO2 measurements. The OH and HO2 calculated by the different mechanisms show similarities and differences with each other and with the measurements. For daytime, CB05 predicts the highest values and SAPRC-99 predicts the lowest. The median modeled-to-observed ratio is 0.82 and 0.77 (CB05), 0.81 and 0.66 (RACM), 0.78 and 0.61 (LaRC), 0.70 and 0.66 (MCMv3.1), 0.65 and 0.77 (SAPRC-07), and 0.59 and 0.54 (SAPRC-99) for OH and HO2, respectively. For nighttime, the highest nighttime OH and HO2 mixing ratios are predicted by RACM and CB05 with modeled-to-observed ratios of only 0.16 and 0.40, respectively. In general, OH and HO2 were underestimated by all mechanisms. The modeled and measured ratio of HO2/OH suggests good simulation when NO is about 1 ppbv, although it is too high when NO was less and too low when NO was more, in agreement with previous studies. OH concentrations are determined by its production and loss, which are in balance on time scales of a few seconds. The former can be calculated based on measurements, and was controlled by the reaction of HO2 with NO for this campaign. The latter is the product of the OH concentration and the OH reactivity (the inverse of the OH lifetime), which was also measured during this campaign. The calculated OH reactivity based on each mechanism generally agrees with the observation to within the observational uncertainty, although the largest discrepancy can reach up to 20% of the observed value for the morning peak time. This agreement between measured and calculated OH reactivity indicates that (1) all important OH reactants were measured; (2) different lumping methods of organic species in these mechanisms lead to little difference in OH loss. In addition, OH production and loss agree to within measurement uncertainties for most days and for most times of the day. However, OH production was overestimated by a factor of up to 3 at morning rush hour when NO has average peak concentration of 14 ppbv, as has been seen in other studies. This overestimated OH production is difficult to understand. Mechanism-mechanism discrepancy and model-measurement difference were sensitive to the environmental conditions. The deviation of modeling results by each mechanism reduces under polluted condition but the modeled-to-measured ratio increases slightly under clean condition. The overall under-prediction of HOx still occurs under different conditions for all mechanisms, which further underestimate O3 production in the model because lower-than-measured HO2 was calculated.