The buildup of ozone precursors and toxic air pollutants in nocturnal boundary layers has been recognized for decades but there is still uncertainty over how much this phenomenon affects midday ozone levels and human exposure to air toxics. Over flat, open terrain the effect on mid-afternoon ozone or other chemistry may be relatively minor, but in basins and valleys the effect can be larger. Numerical models can be used to simulate the effect of the accumulation of pollutants in the nocturnal boundary layer, but there are few measurements of vertical pollutant distribution or boundary-layer structure with which to evaluate model results.

To study these issues, we carried out a field campaign in Phoenix, AZ for a period of three weeks in June/July, 2001. The measurements focused on the nighttime accumulation of ozone precursors and related pollutants, and their distribution the next morning as the nocturnal boundary layer broke up. The campaign involved chemical and meteorological measurements taken at numerous ground sites, measurements at two different elevations from a downtown Phoenix office tower, and measurements taken aboard a research aircraft. This presentation will focus on the office tower measurements. At this site, continuous monitoring was carried out at 50 m AGL elevation for ozone, nitric oxide, total oxides of nitrogen, carbon monoxide, peroxyacetyl nitrate, formaldehyde, nitric acid, nitrous acid, formic acid, acetic acid, hydrogen cyanide, temperature, pressure, dew point temperature, relative humidity, and aerosol light scattering coefficient. Continuous monitoring was also performed at the 139m AGL elevation for ozone, nitric oxide, total oxides of nitrogen, carbon monoxide, peroxyacetyl nitrate, temperature, pressure, dew point temperature, relative humidity, and aerosol light scattering coefficient. These measurements are being used to elucidate the effects of nocturnal accumulation and production of trace species on the levels of ozone and other pollutants on the following day, and to evaluate the capability of numerical models to simulate these processes accurately.