In this dissertation we investigate different factors controlling summertime surface ozone (O3) in the western U.S., including the impacts from increased wildre emissions, the modulation by North American summer monsoon as well as long-range transport of O3 and its precursors from outside of North America.
We first analyze the surface ozone observations from the Clean Air Status and Trend Network (CASTNet) using a global chemical transport model (GEOS-Chem) to investigate the impact of biomass burning on surface O3 in the western U.S. (WUS) mountain ranges during the June-October fire season of 2007, one of the stronger fire years in the WUS in the past decade. GEOS-Chem O3 captures the observed seasonal, synoptic and daily variations. Model daily afternoon average surface O3 concentrations at the CASTNet sites are within 2 ppb of the observations, with correlation coefficients of 0.51-0.83 and Taylor scores of 0.64-0.92. Observed maximum daily 8-hour (MAD8) surface O3 concentrations are 37-58 ppb at the sites, while the corresponding model results are higher by 6 ppb on average. Model results show July-September maximum surface O3 enhancement of ~9 ppb on average because of biomass burning. Peaks in fire-contributed surface O3 correspond broadly with high levels of potassium (K), reaffirming a strong fire influence. We find a policy relevant background (PRB) O3 of 45.6 ppb on average during July-September. Fire-contributed O3 accounts for up to 30% of the PRB O3, highest in the intense fire region (Montana, Idaho, and Wyoming) with maxima in August and September.
We also examine an unexpected summertime surface O3 minimum (30–45 ppb) in July–August 2007 observed throughout the Southwestern U.S. (SWUS) by interpreting observations of O3 and rainfall from the Clean Air Status and Trends Network (CASTNet) with a global chemical transport model. The O3 minimum reflects competing chemical and dynamic factors as well as anthropogenic and natural influences. Its reoccurrence annually in 2000–11 corresponds to the seasonal rainfall maximum during the North American summer monsoon (NASM) (negative O3 and positive rainfall anomalies at the CASTNet sites, r = -0.5 to -0.7, p < 0.05). Relative to June 15–July 15, 2007 (pre-onset of the NASM), increased cloudiness in July 15–August 15 (post-onset) weakens photochemistry, reduces O3 production from anthropogenic emissions, thereby depresses O3 throughout the lower troposphere and at the surface (-5 ppb at Chiricahua, AZ and -3 ppb on average across the SWUS). The resulting changes are largest at rainfall maxima, particularly in the core of the Great Plains low-level jet. Enhanced lightning NOx emissions post-onset augments O3 production in the middle troposphere followed by downward mixing in convective downdrafts, thereby increases O3 significantly throughout the tropospheric column and non-negligibly at the surface (+2 ppb at Chiricahua and +1 ppb averaged over the SWUS). The resulting O3 changes is largest (+8 ppb) in the middle troposphere in the anti-cyclonic circulation associated with the reoccurring summertime high over the Southern U.S. Weaker photochemistry post-onset dominates the overall O3 change near the surface, while enhanced lightning dominates in much of the free troposphere. Additionally, we find that transport leads to a net export of O3 throughout the tropospheric column and the influence from stratospheric intrusion is vanishingly small. These competing effects suppress O3 in the lower troposphere (O3 change up to -5 ppb) while enhance O3 at higher altitudes (O3 change up to +7 ppb) across the SWUS during the monsoon. Better understanding of these effects is critical to estimate present and predict future background O3 in the U.S. Southwest as the NASM changes under a changing climate.
Lastly we use the GEOS-Chem 3-D global tropospheric chemical transport model and its adjoint to quantify the source contributions to O3 pollution observed at Mt. bachelor Observatory (MBO) during the summer of 2008. The adjoint computes the sensitivity of O3 concentration at the receptor site to O3 production rates at 2?x2.5? resolution over the history of air parcels reaching the site. We found that MBO experienced distinct O3 pollution episodes from Siberia wildfire emissions. During the O3 pollution episode from June 30th to July 4th in year 2008, 7.5 ppb of MBO O3 is produced over Siberia, comparable to the amount of O3 (8ppb) produced over North America. A significant amount of O3 (18ppb) production took place over the North pacific, with maxima just off the west coast of the U.S. where subsidence of air masses causes decomposition of PAN (peroxyacetylnitrate, a thermo-unstable NOx reservoir species) and drives further ozone production. We also used the adjoint of GEOS-Chem to show the model O3 at MBO is largely sensitive to NOx emissions from biomass burning sources in Siberia and northern California, lightning sources over southwestern U.S. and Mexico, and anthropogenic sources in western U.S. and eastern Asia. For the CO emissions, the largest O3 sensitivity is to the biomass burning sources in northern California and Siberia. The peak sensitivity to biomass burning CO emissions is comparable to the peak O3 sensitivity to anthropogenic NOx emissions.
