Tropospheric Distribution Of Reactive Nitrogen

Knowledge concerning the tropospheric distribution of NO, is critical given its importance to ozone photochemistry. Over much of the remote atmosphere, NO concentrations hover near the critical level necessary for net photochemical produc tion of ozone. This critical NO level varies from as low as about 5 pptv to near 20 pptv depending on ambient conditions (Crawford et al, 1997). As a general rule, observations have shown NO to typically fall below critical levels over remote marine boundary layer environments away from NOx sources where the lowest ozone values in the atmosphere also occur. By contrast, ozone production in the upper troposphere appears to be ubiquitous given observations of NO consistently above the critical level.

Despite the pivotal role NOx plays in tropospheric photochemistry, current knowledge of its tropospheric distribution is based on very limited data, especially for remote regions. Reliable methods for measuring NO over the full range of its tropospheric variability (a few pptv to tens of ppbv) have been available since the late 1970s. Even so, field measurements ofNOx from ground, ship, and aircraft platforms provide only limited spatial and temporal coverage. Nevertheless, these observations do reveal some basic features in the global distribution of NOx (Emmons et al, 1997; Bradshaw et al, 2000). For instance, gradients in NOx observations are greatest near the surface. NOx in urban areas is typically in the parts-per-billion range and a few hundred parts-per-trillion are common even in rural areas. For remote oceanic regions, however, NOx levels are generally less than 50 ppt and often only a few parts per trillion. These trends in NOx at the surface are consistent with most NOx sources being land based and the short atmospheric lifetime for NOx of a day or less. In the upper troposphere, gradients in NOx are weaker owing to both the longer lifetime for NOx and generally faster transport. NOx values in the range of 50 to 200 pptv are often observed; however, observations ranging from only a few pptv to more than a ppbv of NOx are not uncommon. These extremes are most likely due to the convection of NOx-poor air in marine environments contrasted by the convection of NOx-rich polluted air with additional inputs from lightning in continental regions. As a consequence of the difference in NOx gradients at the surface and high altitude, NOx is generally observed to increase with altitude over remote locations.

Some of these trends can be seen in data collected from NASA's DC-8 aircraft (see Fig. 3). NOx in this figure has been estimated from daytime measurements of NO (solar zenith angle < 70°) by assuming photochemical equilibrium conditions for N02. These data have been taken from the following field campaigns: Pacific Exploratory Mission (PEM)-West A (1991) (Hoell et al, 1996), Transport and Atmospheric Chemistry Near the Equator-Atlantic (TRACE-A, 1992) (Fishman et al, 1996), PEM-Tropics A (1997) (Hoell et al, 1999), and Subsonics Assessment Ozone and Nitrogen Oxides Experiment (SONEX, 1998) (Singh et al, 1999). These campaigns have focused on taking measurements to characterize the remote oceanic troposphere with an emphasis on the upper troposphere. While flown in different years, each campaign was conducted during the fall season (September-November).

Figure 3 shows data for the boundary layer (0 to 1 km), the lower free troposphere (1 to 6 km), and the upper troposphere (6 to 12 km). In general, NOx over the Atlantic is greater than over the Pacific at all altitudes. This is due to a closer proximity to NOx sources, which are predominantly land based. The seasonal nature of NOx sources is also important to the elevated levels of NOx over the South Atlantic since measurements were taken during the biomass burning season

A TROPOSPHERIC DISTRIBUTION OF REACTIVE NITROGEN

Upper Free Troposphere (6-12 km)

Longitude Lower Free Troposphere (1-6 km)

135 E

30 N

135 E

0 E 45 E Longitude

Boundary Layer (0-1 km)

135 E

135 W 90 W

Figure 3 (see color insert) Distribution of NO, based on measurements taken from NASA's DC-8 aircraft during fall (see text for details). Data are averaged on a 5 x 5 latitude-longitude izrid for three altitude ranges. See ftp site for color image.

Upper Free Troposphere (6-12 km)

Longitude Lower Free Troposphere (1-6 km)

90 N

135 E

135 E

135 E

ISO E

180 E

ISO E

90 S

Boundary Layer (0-1 km)

90 N

0 E 45 E Longitude

90 S 180 W

135 W 90 W

45 W 0E

Longitude

30 N

NOx (pptv)

m

>200

m

1D0-ZD0

m

50-100

m

25-50

M

10-25

<10

Figure 3 (see color insert) Distribution of NO, based on measurements taken from NASA's DC-8 aircraft during fall (see text for details). Data are averaged on a 5 x 5 latitude-longitude izrid for three altitude ranges. See ftp site for color image.

74 NITROGEN OXIDES AND OTHER REACTIVE NITROGEN SPECIES

for South America and Africa. While boundary layer data over and near continental areas are sparse, the strong gradient in surface NOt is evident in the low values (typically < lOpptv) over the South Pacific. Gradients are weaker at higher altitudes. The increase in NO t with altitude over remote oceanic regions is also evident for both the South Atlantic and South Pacific data. The decrease in NO t with altitude over continental areas is less evident since data is sparse, but NO t values over South America, Southern Africa, and the South China coast do exhibit a decrease with altitude.

Information concerning the distribution of NO^ is even more limited than that for NOt. The only NO^ species other than NOt that have been measured with any regularity are HN03 and PAN. Total NO^ measurements have been more common, but there are still questions as to what these measurements represent since they often exceed expected values based on the sum of all NO^ constituents (Crosley, 1996). Measurements of HN03 in the remote troposphere have consistently fallen well below levels expected based on theory (Liu et al., 1992; Ridley et al., 1998; Schultz et al., 2000). This problem has been particularly troubling since theory predicts HN03 to be the dominant NO^ species in the remote troposphere. Heterogeneous mechanisms recycling HN03 to NOt have been hypothesized as a potential solution (Fan et al., 1994; Chatfieldi, 1994; Hauglustaine et al., 1996; Lary et al., 1997). Possible underestimations in wet removal and partitioning of HN03 between gas and aerosol phases have been cited as well (Liu et al., 1992; Wang et al., 1998).

Measurements of PAN support the contention that it plays a strong role in sustaining NOx in air masses as they are transported away from regions of strong industrial or biomass burning emissions. Decomposition of PAN was found to be adequate to explain NOx observations at low altitude over eastern Canada (Fan et al., 1994), the South Atlantic (Jacob et al., 1996), and the western, North Pacific (Crawford et al., 1997). For even more remote regions, the decomposition of PAN in descending air masses can also be responsible for sustaining NOx. This condition was observed by Schultz et al. (1999) over the remote South Pacific.

While global models can be used to estimate the distributions of NOt and NO^, the accuracy of these estimates is still very uncertain and are complicated by several factors. First is the level of uncertainty that remains for various NOx sources, especially natural source strengths as well as spatial distributions. Second, there are still major questions concerning the recycling of NOx from NO^ reservoir species and the potential role of aerosol in both the removal and recycling of NOx. Finally, the wide range of photochemical lifetimes for NOx requires atmospheric models to accurately represent small-scale transport processes (e.g., convective vertical transport of NOx and wet deposition of HN03). The scales for these processes remain significantly smaller than the resolution of current photochemical transport models.

REFERENCES 75

0 0

Post a comment