The chemical evolution of the Sacramento urban plume (California)

Observations of hydrocarbons, nitrogen oxides, and ozone in the Sacramento, California region are analyzed using a Lagrangian plume model. The analysis indicates that the Sacramento urban plume is typically diluted to 29% of the urban core concentration and oxidized by an average of 1.4 × 10 7 molecules cm−3 of OH radicals over a five hour period. The model is capable of accurately reproducing hydrocarbon, NO _x, and NO_y observations at the University of California - Blodgett Forest Research Station (UC-BFRS); however, it overestimates ozone mixing ratios within the urban plume by 25 ppb or 35%. This discrepancy demonstrates that our Lagrangian model constrains the average atmospheric ozone production rate to within a factor of two. This degree of precision is sufficient to distinguish between different methods of calculating ozone production rates. The Lagrangian model represents a tool to assess if current photochemical mechanisms accurately capture all relevant tropospheric ozone photochemistry. For the model to accurately describe ozone, we require a combination of minimal reactive hydrocarbon and formaldehyde sources downwind of the Sacramento urban core, which reduce the modeled ozone production rate by 55%, and more vigorous mixing in the first half of the plume than in the second half.; Our Lagrangian analysis of NO_x and NO_y in the Sacramento urban plume suggests that much could be learned from a more complete analysis of NO_y partitioning. Here we also evaluate a new technique for observing speciated NO_y, Thermal Dissociation - Laser Induced Fluorescence (TD-LIF), by comparison to a metalcatalysis/ozone chemiluminescence NO _y detector (CL). We compare 960 hours of simultaneous 1 minute observations at the UC-BFRS taken by the two instruments. The instruments agree to within 10%. This agreement varies with season: the CL observations were greater than the TD-LIF observations by 6% both in the fall and winter months and the TD-LIF observations were greater than the CL observations by 20% and 10% in the spring and summer months, respectively. The conversion of NH₃ to NO ₂ in the TD-LIF HNO₃ channel is consistent with the magnitude and seasonal cycle of the instrumental differences based on internal TD-LIF diagnostics, prior observations of HNO₃ and NH₃ in the Sierra Nevada, a theoretical NH₃ decomposition mechanism, and preliminary laboratory experiments. During the summer mornings, the CL observations were greater than the TD-LIF observations by 500 ppt (30%) when corrected for the NH₃ interference. This additional signal in the CL instrument may be HONO given the unique conversion method of the TD-LIF instrument, the time in which the signal is observed, and prior observations of HONO in the Sierra Nevada mountains.