Collaborators: H. Akimoto (FRSGC), M.J. Prather (UCI), J. Sundet (U. Oslo) and I.S.A. Isaksen (U. Oslo)
Meteorological processes control the production of ozone from precursors over polluted urban/industrial regions, as well as the export of ozone from these regions, through the effects of cloud cover, humidity, temperature and rainfall. The regional and global impacts of urban/industrial emission sources are coupled through regional meteorology, and the time scales for ozone formation and destruction differ greatly between clean and polluted environments and between the surface and the upper troposphere. Here we explore how regional meteorological processes affect the quantity and location of ozone production, with particular interest in the relative importance of regional and free-tropospheric ozone production from surface industrial sources. We focus on the East Asian region during the springtime, when meteorological conditions are particularly variable, and study March 2001, when the NASA Transport and Chemical Evolution over the Pacific (TRACE-P) measurement campaign was held.
We use the Frontier Research System for Global Change (FRSGC) version of the University of California, Irvine (UCI) global chemical transport model running off-line at T63 (1.875°x1.875°), 37-layer resolution with meteorology generated with the ECMWF Integrated Forecast System by Jostein Sundet (University of Oslo). Figure 1 shows an example of the ozone distribution and the net chemical ozone production rate at two altitudes over the western Pacific on March 4, during the passage of a cyclone from the mainland of Asia out to the Pacific. This system brought with it both polluted continental outflow and dry, descending stratospheric air. Extensive ozone production occurs in the outflow region behind the front, where it is capped in the lower troposphere by subsidence; the clean marine conditions ahead of the front show substantial ozone destruction.
Figure 1. Examples of the ozone distribution (ppbv, left panels) and net ozone formation rate (ppbv/day, right panels) at 4:00 GMT (local noon at 120°E) on March 4, 2001, during the TRACE-P measurement campaign, showing the influence of a large cyclone at 44°N, 133°E and the associated frontal systems. The upper left panel shows the distribution of ozone from the stratosphere at 500 hPa, clearly illustrating the dry intrusion behind the cold front; net ozone production is found at the same level ahead of the front, where precursors from surface sources have been lifted by convection and slower frontal lifting (upper right panel). At 800 hPa, ozone formed from tropospheric sources is greatest in the outflow region behind the cold front (lower left panel), and ozone production in this region reveals the impact of formation from mainland Asian sources (lower right panel). The ozone mixing ratio is lowest in the clean, marine conditions in the warm sector ahead of the cold front, which shows considerable ozone destruction. The tracks of the DC-8 and P-3B aircraft, flying from Guam to Hong Kong, are shown in white; contours show the mean sea level pressure.
The daily ozone formation over mainland China in the spring period is shown in Figure 2. There is a notable increase in ozone production over the 3-month period due to longer hours of daylight and increasing temperature, dominated by the northern parts of the region. While over southern regions such as Hong Kong there is little variation in production over this period, the northern regions around Beijing show a dramatic increase in production. Total net boundary layer ozone formation over the region is 11.9 Tg over the period, with a production rate of 134±20 Gg/day, and a trend of 0.7 Gg/day².Figure 2. Ozone formation over mainland China between February 1 and April 30, 2001, demonstrating the daily variability induced by meteorological systems and the net increase in production during springtime.
To investigate the meteorological impacts on regional production in more detail, we apply one-day pulses of precursor emissions over key emission regions, and follow the subsequent regional and global ozone responses. Figure 3 shows examples of ozone changes due to additional emissions from Shanghai for two days in March 2001 showing differing responses. An anticyclone dominated the region on March 12, and clear skies and a capped boundary layer led to rapid ozone formation and substantial build-up in the boundary layer over the region, though little effect at higher altitudes for the first 2-3 days. In contrast, on March 16 a nearby region of cyclogenesis led to extensive cloud cover, and very little ozone build-up in the regional boundary layer. However, substantial quantities of precursors were lifted into the free troposphere, and ozone production continued at a slower pace for the next few days. The global responses, while quite different for the first few days due to the different time scales for ozone formation, are very similar in magnitude after the first week, suggesting that while regional impacts are strongly dependent on local meteorology, the global impacts may nevertheless be quite similar.Figure 3. Ozone formation from one-day pulses of precursor emissions over Shanghai on March 12 and March 16, expressed as a perturbation to the ozone budget over the region (red), the globe (black) and over different parts of the troposphere.
The peak additional ozone over the Shanghai region and over the globe (the maximum of the red and black curves in Figure 3, respectively) are shown for each day of March in Figure 4. The month-mean in the peak regional perturbation is 0.97±0.24 Gg and of the global perturbation is 1.62±0.16 Gg, suggesting that regional ozone production dominates the global impact on ozone at this location in this season. Note that the variability in the perturbation (defined here by one standard deviation) is about 25% over the region but only 10% over the globe, highlighting the greater impact of meteorology on the time scale for production than on the total production itself.Figure 4. Peak ozone perturbations over Shanghai (red) and over the globe (blue) due to one-day pulses of precursor emissions over Shanghai for each day of March 2001.
The days of heaviest cloud cover (March 7, 16 and mid-20s) typically show greater than average ozone formation beyond the emission region due to greater export of precursors. This effect is dominated by the impacts of lifting processes; convection associated with the cloud cover lifts precursors into the free troposphere, where ozone production efficiencies are higher than in the boundary layer, see Figure 5. This demonstrates the important role that the lifting of precursors plays in the subsequent global formation of ozone, and contributes to the much smaller variability seen in global production than in regional production.Figure 5. Excess ozone formation above the boundary layer, defined as the difference between the peak global and boundary layer ozone perturbations, versus the additional NOx lifted out of the boundary layer, demonstrating the role of vertical transport of NOx (principally by convection) in ozone production in the free troposphere.
This study has started to address the variability in regional and global ozone production from a single emission region on a daily basis, and to explore how the links between them are related to the regional meteorology. Warm, sunny, anticyclonic conditions typically lead to a large build up of regional ozone, with consequences for boundary layer air quality. However, we find that the global ozone production from the same emissions is typically no more than on cooler, overcast days, suggesting that the impacts of regional emissions on the global distribution of ozone, and hence on climate forcing, are much less sensitive to these meteorological influences. Further studies over different regions and seasons are needed to address how general these results are.