UC Davis

Midday summertime flight data collected in the atmospheric boundary layer (ABL) of California's San Joaquin Valley (SJV) are used to investigate the scalar budgets of NOx , O3, and CH4, in order to quantify the individual processes that control near-surface concentrations, yet are difficult to constrain from surface measurements alone: These include, most importantly, horizontal advection and entrainment mixing from above. The setting is a large mountain- valley system with a small aspect ratio, where topography and persistent temperature inversions impose strong restraints on ABL ventilation. In conjunction with the observed time rates of change this airborne budgeting technique enables us to deduce net photochemical ozone production rates and emission fluxes of NOx and CH4. Estimated NOx emissions from our principal flight domain averaged 216 (33) t d-1 over six flights in July and August, which is nearly double the California government's NOx inventory for the surrounding three-county region. We consider several possibilities for this discrepancy, including the influence of wildfires, the temporal bias of the airborne sampling, instrumental interferences, and the recent hypothesis presented by Almaraz et al. (2018) of localized high soil NO emissions from intensive agricultural application of nitrogen fertilizers in the region and find the latter to be the most likely explanation. The methane emission average was 438 Gg yr-1 (143), which also exceeds the emissions inventory for the region by almost a factor of 2. Measured ABL ozone during the six afternoon flights averaged 74 ppb (D 9:8 ppb). The average midafternoon ozone rise of 2.8 ppb h-1 was found to be comprised of-0:8 ppb h-1 due to horizontal advection of lower O3 levels upwind,-2:5 ppb h-1 from dry deposition loss,-0:5 ppb h-1 from dilution by entrainment mixing, and 6.9 ppb h-1 net in situ photochemical production. The O3 production rates exhibited a dependence on NO2 concentrations (r2 D 0:35) and no discernible dependence on methane concentrations (r2 0:02), which are correlated with many of the dominant volatile organic compounds in the region, suggesting that the ozone chemistry was predominantly NOx-limited on the flight days. Additionally, in order to determine the heterogeneity of the different scalars, autocorrelation lengths were calculated for potential temperature (18 km), water vapor (18 km), ozone (30 km), methane (27 km), and NOx (28 km). The spatially diffuse patterns of CH4 and NOx seem to imply a preponderance of broad areal sources rather than localized emissions from cities and/or highway traffic within the SJV.