Jim opened the session by outlining current thoughts on NOX production from thunderstorms. Knowledge remains limited on many fronts: how much NOX is produced per flash, how many cloud-to-ground and intracloud flashes occur per storm, and where in each storm lightning occurs. "All of these are research problems; we really don't have answers to them," Jim said. For instance, new satellite readings have dropped the standard estimate for global lightning frequency by half, from 100 flashes per second to 40-50 flashes per second.
New clues to the lightning-NOX problem are coming from field work done in northeast Colorado last summer as part of the Stratosphere-Troposphere Experiment: Radiation, Aerosols, and Ozone (STERAO). Jim and colleagues are now working on analysis of STERAO results. In the data set, large NOX enhancements are evident in the uppermost part of thunderstorm anvils. According to Jim, this implies that a large fraction of a storm's lightning is being generated in the updraft region, from which the resulting NOX is swept into the anvil. Jim also discussed a storm from last summer that produced very few cloud-to-ground flashes once it entered its supercell phase: "That's a finding that's somewhat characteristic of supercells, though not universal."
Brian presented results from STERAO and an exploratory project in New Mexico showing the large enhancements of NOX in storm anvils. However, he noted that no obvious NOX enhancement was found below cloud base in either region--a surprising finding, since cloud-to-ground flashes have been considered the dominant source. Brian also pointed out the importance of where NOX is deposited: its lifetime in the upper troposphere is much longer than in the boundary layer, so its impact can be much larger.
Peter presented early simulations from a coupled chemical-transport/mesoscale model that examined Colorado storms on one of the STERAO days. Such a model can depict NOX production by lightning as well as residual effects from NOX transport by strong thunderstorms. Peter discussed assumptions used in the simulations and emphasized the need for improved observations, such as those provided by STERAO.
Morris and Jenkins discussed the ozone maximum over the South Atlantic and how lightning might help explain it. "There's been a lot of discussion on whether that max is due to biomass burning in South America," said Morris. However, Jenkins noted that, in Africa, more than 400,000 fires a year are observed, and central Africa is home to the world's highest annual density of lightning flashes. Morris and Jenkins believe the Atlantic ozone maximum could result from an interaction among African biomass burning and circulations induced by large, long-lasting storm complexes. In September and October, the largest areas of African fires are collocated with the peak regions for storms; the highest Atlantic ozone levels occur around this time of year as well.
"The big problem in dealing with the Atlantic is data--there's just not enough," said Jenkins. Several new satellite platforms will help, including the Measurement of Pollutants in the Troposphere (MOPITT) instrument, scheduled for launch in 1998 with data retrieval algorithms developed at NCAR. (See the lead story in the latest UCAR Quarterly.) Meanwhile, Jenkins and Morris are planning laboratory experiments to help quantify the amount of NOX produced per lightning flash.
The STERAO work has helped settle the matter of lightning-based production of ozone. According to Jim and Brian (principal investigators in STERAO), the experiment confirmed earlier findings in New Mexico that lightning doesn't produce significant amounts of ozone at high altitudes. "There could be minor production of ozone," said Brian, "but it's hard to detect [when ambient levels are much higher]."