ASHOE/MAESA End of Mission Summary of Preliminary Results Flights of the NASA ER-2 high altitude aircraft, carrying as many as 16 instruments, provide new observations to diagnose both the chemistry and the motion of air in the lower stratosphere. Measurements span February to November, 1994 and from the edge of Antarctica at 70#161#S to upper Canada at 60#161#N, in conjunction with observations from the ground, balloons, and satellites. Most measurements were in the southern hemisphere and in the tropics, both regions with few or no high-quality measurements of the abundances of trace gases and of the meteorology. This statement summarizes the observations and preliminary interpretations of the science team for Airborne Southern Hemisphere Ozone Experiment / Measurements for Assessing the Effects of Stratospheric Aircraft (ASHOE/MAESA), which occurred from February to November, 1994. A more detailed analysis will follow during the coming months and years. An important activity is the comparison of these results with those from previous missions, including the Stratosphere Troposphere Exchange Program (STEP) in 1987, the Airborne Antarctic Ozone Expedition (AAOE) in 1987, the Airborne Arctic Stratospheric Expedition (AASE) in 1989, the Airborne Arctic Stratospheric Expedition II (AASE-II) in 1991-1992, and the Stratospheric Photochemistry Aerosol and Dynamics Experiment (SPADE) in May, 1993. Operation The observations are based primarily on 45 flights of the NASA ER-2 high altitude aircraft. For most flights, the ER-2 carried 16 instruments to measure the abundances of trace gases, temperature, pressure, winds, ultraviolet light, and temperature profiles. For flights specifically to study dynamics and radiation, some of the trace gas instruments were removed and a instrument that measures infrared radiation was added. Among the 45 flights, 26 were from Christchurch, New Zealand, 5 were from Barber's Point, Hawaii, and 6 were transits between New Zealand, Fiji, Hawaii, and California. Also included are two full duration (8 hour) test flights north from Moffett Field in February, and a northbound flight from Moffett Field in November, 1994. Thirty-six flights had the full payload and 9 had the dynamics and radiation configuration. Observations from other instruments on the ground, balloons, and satellites provide valuable contributions to ASHOE/MAESA. Analyses of weather observations by UKMO, ECMWF, NMC, and GSFC aided flight planning and dynamical analyses. Satellite data from MLS and HALOE on UARS and TOMS on Meteor give a global context for the local measurements of the ER-2. The GSFC Lidar and NIWAR column measurements from Lauder, New Zealand, and the MacQuarie Island ozonsondes provided valuable intercomparisons for aerosol and ozone measurements on the aircraft, extending the observations well above and below the aircraft. All these observations will be combined with those from the ER-2 for study. Objectives The over-arching mission objective, as laid out in the mission booklet, was to investigate the causes of ozone loss in the wintertime midlatitudes in the southern hemisphere, which has been observed by satellites for the last 15 years. To meet this objective, ER-2 flights were designed to observe several aspects of ozone loss: the rapid ozone decline over Antarctica, the spread of this ozone-poor air to midlatitudes, and the activation of ozone-destroying chlorine directly in midlatitudes by reactions on the sulfate aerosols. They also enable us to determine the rate of chemical ozone destruction by reactive trace chemicals in the nitrogen, hydrogen, chlorine, and bromine chemical families and to test the amount of these chemicals that are in the reactive forms. Some flights focused on observations to study the mixing of air between the midlatitudes and tropics because this process helps establish the ozone distributions and rates of ozone change. The evolution of the Antarctic polar vortex, with its circumpolar winds and low temperatures, was studied during the course of ASHOE/MAESA. The rapid loss of ozone, reaching its maximum extent early in October each year that is known as the ozone hole appears mainly within this vortex. Measurements of trace gases and meteorological parameters were made early in the life cycle of the vortex (deployment I in March-April); after the vortex had formed and temperatures dropped below about 195 K, the formation temperature of polar stratospheric clouds (deployment II in May-June); after the temperatures had dropped below ice saturation, when large scale loss of the condensable vapors HNO3 and H2O by sedimentation in PSC particles occurs (deployment III in July-August); and when ozone was the most depleted (deployment IV, October). The midlatitude regions equatorward of the maximum in the polar night jet stream were also the subject of close observation and study. The other objective of ASHOE/MAESA was to obtain measurements for the assessment of the atmospheric effects of future stratospheric aircraft that are now being considered. ASHOE/MAESA addresses the three most uncertain aspects of this assess ment. First is the transport by air motions of the High Speed Civil Transports (HSCTs) exhaust from midlatitudes, where most HSCTs will fly, to the tropics, where the exhaust could rise up and spread out to midlatitudes at higher altitudes. Model calculations suggest that HSCT exhaust could efficiently destroy ozone above 25 km altitude, while at 20 km the exhaust would have little effect. Understanding the transport of air between the lower stratosphere where the HSCTs will fly and the middle stratosphere where the exhaust can do the most damage to ozone is essential. Second, the HSCT exhaust is calculated to increase stratospheric water vapor and gaseous nitric acid, which are the main components of polar stratospheric clouds (PSCs). HSCT exhaust near the cold regions at the winter poles and the equator could therefore possibly increase the frequency and range of PSCs, and also the reactivity of gases on cold, sulfate aerosols, which increases rapidly as the water content of the particles rises. Because ozone-destroying chlorine is rapidly produced on PSCs and cold sulfate aerosols by reactions of the reservoir gases on the particles, an increase in these clouds could result in increased ozone loss. Third, the uncertainties in the laboratory measurements of chemical reaction rates and products that are used in assessment models directly affect the uncertainty in the calculations of ozone loss expected for projected injections of HSCT exhaust. Measurements of trace gas concentrations test this chemistry and limit both the uncertainty and the possibility of some missing chemistry, when they are taken over a wide range of latitudes, altitudes, seasons, time of day, and abundances of chlorine, bromine, and nitrogen chemicals. An uncovenanted bonus was an opportunity to measure directly the chemical composition of the exhaust plume from a Concorde supersonic airliner as it approached Christchurch, New Zealand. Of the currently operating aircraft, the Concorde is the one most similar to the projected HSCTs. Measurements of the chemicals in its exhaust over a range of times tests the validity of the ground-based measurements of the emission indices of pollutants for the Concorde. It also tests atmospheric chemistry for the polluted conditions of the exhaust plume. Summary of Observations and Results These unique observations of variations of long-lived trace gases in the lower stratosphere will help define the transport and mixing of air among the polar, midlatitude, and tropical regions of the stratosphere and the troposphere. Observations and models find that the transport of polar air, primed for ozone loss, occurs as filaments that are shed into the midlatitudes. Studies underway seek to quantify the midlatitude ozone loss that results from this mixing process. Tropical air is also observed to be shed as filaments into midlatitudes.The aircraft observations suggest that midlatitude air mixes into the outer tropics from 23.5#161# to about 10#161# latitude, with the inner equatorial region showing signs of air which had entered the stratosphere more recently from below. The extent to which the outer tropical air can ascend to higher altitudes is crucial in the context of the potential effects of HSCT exhaust. Calculations show that such exhaust will destroy ozone if it gets to higher altitudes. These new observations are catalyzing a new way to look at transport of air in the stratosphere and between the stratosphere and troposphere. The ASHOE/MAESA observations will improve the understanding of the processes that lead to the Antarctic ozone hole and to ozone loss at midlatitudes. These processes include the formation of PSCs, the conversion of chlorine from its reservoir forms to its ozone-destroying reactive form on cold, sulfate aerosols, and of the return of chlorine to the inactive forms after all ozone is destroyed inside the ozone hole. The transport mechanisms which redistribute the effects of these chemical processes will also be better understood. There will be a better understanding of the detailed mechanisms of the processes, which will help to improve numerical models used to simulate the stratosphere. These observations, when compared to theoretical models, demonstrate a generally good understanding of the photochemistry in the lower stratosphere for the altitudes, latitudes, and seasons sampled during ASHOE/MAESA. The main disagreement between observation and model calculation is the amount of chlorine in the form of HCl in the Northern Hemisphere. Otherwise, agreement is generally far better than expected from the stated errors in the measurements and the model parameters. The simultaneous measurements of the reactive hydrogen, reactive chlorine and reactive bromine show conclusively that these species destroy ozone faster in the lower stratosphere than does reactive nitrogen, in the chlorine-laden stratosphere of the 1990's. Reactive nitrogen is now thought to be dominant only higher in the stratosphere, above 25 km. Thus, the calculated effect of HSCT exhaust on ozone is small throughout the lower stratosphere. The importance of reactions of reactive nitrogen on sulfate aerosols, which convert reactive forms to a reservoir form, is confirmed for a stratosphere with the low aerosol amounts characteristic of the period three or more years after a major volcanic eruption. Observations and results from ASHOE/MAESA The structure of this section is based in a consideration of three latitude regimes: polar, midlatitude, and tropical. A separate section is devoted to the Concorde exhaust encounter. Statements are presented as separate paragraphs under each heading. This document is not intended to be referenceable. Polar region Ozone loss in the Antarctic polar vortex in October was substantial, just as in the previous two years. Preliminary minimum ozone values reported by Meteor TOMS were 90 Dobson Units and the average areal extent of intense ozone loss was about 23 million km2, slightly larger than in 1992 and 1993 and almost twice the size of Antarctica. Ozone loss in the Antarctic polar vortex has now been severe in 7 of the last 8 years. The aircraft observations of ozone and tracers provide a high quality set which span a wide range of latitudes and altitudes between 16 and 21 km in the lower stratosphere over the whole period from March to October. They provide the database to examine ozone loss and its relationship to diabatic cooling across the polar night jet stream and into the vortex. Mean temperatures and other gross average indicators of dynamical activity associated with the Antarctic polar vortex were near the average values for the last 15 years, implying that observations from ASHOE/MAESA are representative. These indicators include the stratospheric zonal mean winds and heat flux at 60oS and zonal mean temperature gradient. Minimum temperatures, important for PSC activity, were slightly lower than average by a few degrees for June through October. Temperatures less than 195 K persisted at 70 mb within the vortex into the third week of October , and at 100mb into early November. Observations and model calculations demonstrate that air is transported out of the vortex as filaments that are sheared from the vortex edge and become thinner with time. The shedding of filaments was observed during all deployments, but was most intense in October. Studies of observed variations in the winds near the vortex edge will define the importance of smaller scale fluctuations for the exchange of air between the polar region and midlatitudes. Observations of ClO and HO2/OH were made in the air in and around the polar vortex during all four deployments. Analysis of these observations will determine the role of heterogeneous reactions on cold sulfate aerosols. Two processes can be tested: ClONO2 + H2O -> HOCl + HNO3; and HCl + ClONO2 -> Cl2 + HNO3. Laboratory studies suggest that ClONO2 + H2O on sulfate aerosols is important when the temperature is below about 205 K. The temperature at which this process will become more important is likely to be increased by the addition of water vapor and NOy from HSCT exhaust. Measurements made in type I PSC's at the vortex edge during late July challenge the commonly accepted notion that such clouds consist of particles of nitric acid trihydrate. The clouds appeared to be made up of externally mixed particles, and could contain both liquid and frozen particles of various hydrates and ternary mixtures of water, sulfuric acid and nitric acid. Surface areas of aerosols appeared to be much lower in cloud free vortex air which had earlier been colder than frost point. Tracer mixing ratios inside the vortex appeared to be higher than those encountered at similar altitudes and latitudes during the 1987 AAOE mission from Punta Arenas in 1987. Intense denitrification and dehydration were not encountered in horizontal flight above 16 km, except on one flight in early October when water vapor dropped sharply 7#161# of latitude poleward of the wind maximum; in 1987 the loss of condensable vapors was mostly evident closer to the wind maximum. There was no large asymmetry (#197#30%) of water vapor between the hemispheres outside the vortex, a feature which was observed in 1987. Preliminary measurements of molecular hydrogen during August and October constrain its value at flight altitudes to between 0.45 and 0.70 ppmv. There is a low amplitude inverse correlation between H2 and CH4 in the lower stratosphere. The total hydrogen budget appears to decrease poleward. Enhanced nitrification (evaporation of sedimented particles at a lower altitude) was observed in early June near the vortex edge at about 16 km. Denitrification (removal of NOy by particle growth and sedimentation) was observed in mid-winter (July) and subsequently, mainly at and below 16 km. The amount of denitrification may be important for the observed complete removal of O3 in the Antarctic polar vortex. Observed dehydration was minimal in early August; in early October, intense dehydration occurred about 7o latitude south of the vortex edge at cruise altitudes (18-20 km). At and below about 16 km, some obviously dehydrated air was observed further equatorward. Denitrification was observed on some flights at about 16 km in early June and early August, with no accompanying dehydration. The two processes are thus shown to be independent over Antarctica. Dehydrated air in August and October was also denitrified. The expected increase in ClO and decrease in HCl after the occurrence of PSCs were observed. Beginning in early June, ClO increased to 700-900 pptv in air which had been at temperatures below 195K, and HCl showed a proportional decrease. Activated chlorine in the form of enhanced ClO was seen in the polar vortex on four flights. Loss of HCl was seen on more flights, since the instrument concerned was on the dynamics & radiation flights as well. These observations of ClO and HCl confirm the role of PSCs in creating enhanced levels of ozone-destroying reactive chlorine. The observations confirm that ClOx and BrOx catalytic chemistry dominates ozone loss in the polar vortex. Although it had been suggested that OH and HO2 might have a significant role for polar ozone loss, the observations of small OH and HO2 abundances demonstrate an insignificant role for OH and HO2 for ozone loss inside and around the vortex for fall through to spring. Polar air nearly devoid of ozone was sampled in situ for the first time in side the vortex near 400 K (15 km altitude) on 10 and 13 October. Mid-October is late in the period of severe ozone depletion over Antarctica. This air contained little O3 (<0.4 ppmv), no ClO, high NO (1 ppbv), low NOy (2 ppbv), and HCl equivalent to total inorganic chlorine (2.6 ppbv). The almost total loss in ozone creates a condition where chlorine is shifted rapidly from ClO to HCl via the reaction of Cl with CH4 and NOx remains as NO. Thus, HCl was observed to be the dominant remaining chlorine species, whereas in the Arctic, with smaller ozone loss, ClONO2 is observed to be. The implications of these interhemispheric differences in the final phases of the polar vortices on ozone loss at midlatitudes are being examined. The dynamics & radiation payload provided evidence for a layer of warm air in the 14-17 km region at the base of the vortex which was not well reproduced in meteorological forecasts and analyses. Preliminary estimates of diabatic cooling rates in the polar night jet stream and in the vortex between the tropopause and 19 km showed somewhat larger cooling rates than those produced by numerical model calculations. Observations from the UARS satellite made by the MLS and HALOE instruments were provided to the aircraft site in Christchurch. Preliminary intercomparisons with the ER2 are encouraging, and lend both powerful support and a global context, including altitudes above the aircraft, to the in situ measurements. Middle latitudes The fast photochemistry of the lower stratosphere appears to be well understood for the altitudes, latitudes, and seasons that were sampled by the ER-2 during ASHOE/MAESA and SPADE. This ability to test the photochemical chain reactions is a direct result of careful, simultaneous measurements of a number of trace gases and metereorological parameters on the same platform, a capability which has not existed previously. The observed NO, OH, HO2, CO, and ClO are well simulated by a model that uses measured long-lived species as input data. Agreement is generally far better than expected from the stated errors in the measurements and the model parameters. This understanding greatly increases confidence in the chemistry incorporated into assessments models for studying the effects of HSCTs. However there is an important deviation from the generally good agreement between observation and model calculation: the HCl/Cly ratio in the Northern Hemisphere in May 1993 and in November 1994. It was observed to be 0.4-0.5 in the Northern Hemisphere in May 1993 and was about 2 times lower than model calculations. On the other hand, in the Southern Hemisphere midlatitudes during ASHOE/MAESA, the ratio was observed to be 0.75 in good agreement with model results. As a result, the Cly budget, which is constructed from measured organic chlorine, HCl, ClO, and ClONO2 inferred from ClO and NO, is balanced in the Southern Hemisphere this year but not in the Northern Hemisphere for 1993 and during one flight in early November 1994 at the end of ASHOE/MAESA. This observed difference is unexplained, and suggests that some aspects of stratospheric photochemistry of chlorine are poorly known. Filaments of both polar and tropical air are observed at middle latitudes for potential temperature surfaces between 340 K and 520 K (roughly 15 to 20 km altitude) as they were during AASE II and SPADE. These filaments, which are contorted into thin, sheets by wind shear, are often predicted by contour advection and trajectory dynamical models. They may be a dominant way that air is transported among the polar region, midlatitudes, and tropics. The combination between in situ observations and these new, advanced modeling tools will provide new insight into transport processes. The amount of aerosol in the stratosphere has decreased steadily since the eruption of Mt. Pinatubo. During ASHOE/MAESA, the aerosol surface area dropped from 2 x 10- 8 cm2 cm- 3 to less than 1 x 10- 8 cm2 cm- 3, with even lower values seen on 8 and 10 August. These values are comparable to background values observed before the eruption of Mt. Pinatubo. Thus, the influence of aerosols on the trace chemical species and on ozone loss is less now than it was for the previous missions: AASE-II in 1991-1992 and SPADE in 1993. The expected chemical response of the partitioning with in the nitrogen, halogen, and hydrogen chemical families for this range of aerosol surface areas has been confirmed for all latitudes. Because HSCT exhaust is calculated to increase the amount of ozone in the lower stratosphere in the presence of these aerosols, this confirmation is important. Observations of HO2, ClO, BrO, and NO during ASHOE/MAESA show that the HOx and ClOx catalytic cycles dominate ozone destruction globally in the lower stratosphere. These results are similar to those from SPADE, but extend them over a wide range of latitudes and seasons. On this basis small increases in NOy, such as those that might occur from HSCTs, would decrease ozone loss in the lower stratosphere. OH and HO2 have been measured in the upper troposphere for the first time, along with NO, NOy, O3, CO, and CH4. Analysis of these measurements, made during ascent, descent, and dives of the ER-2 into the troposphere, provides fairly complete first tests of the photochemistry for this region. Knowledge of the photochemistry of the upper troposphere is important for global climate change and for the effects of the current subsonic fleet. Tropics Prior to ASHOE/MAESA, many gases that indicate atmospheric motion were either poorly measured or unmeasured in the tropics. Observed during ASHOE/MAESA in both the tropics and midlatitudes are a number of gases, including N2O, CO2, CO, H2O, CFC-11, CH4, SF6, and O3. These gases have different trends and CO2, H2O, CO, and O3 have seasonal variations as well. The variation of the relationships among these gases with season, latitude, and altitude, especially when compared to tropospheric values, gives new information about transport into the stratosphere and between the stratosphere and troposphere. The relationships among these tracers is distinctly different in the tropics and middle latitudes. The tropical regime defined by trace gases is about 15#161# - 20#161# wide in latitude. The low value for NOy/O3 previously observed during STEP and AAOE in 1987 in the tropical regime was confirmed. Long-lived tracers approach tropospheric values in this region. The tropical observations in late October showed a layered structure in the vertical near the equator between 16 and 20 km, which was present in H2O, NOy and CO2; the minimum water was collocated with the tropopause. These measurements will shed new light on the entry and dehydration of air at the equatorial tropopause. Poleward of the changes in the tracers, the values of NOy/O3 and long-lived tracers are observed to be intermediate between the observed values deep within the tropics and at middle latitudes. These observations suggest that information about mixing between the tropics and midlatitudes can be obtained from analysis of these observations. This information is crucial for understanding the potential transport of HSCT exhaust in the upwelling of the tropics to the middle stratosphere, where it will chemically destroy ozone. The annual variation in CO2 shows a maximum occurring simultaneously in both hemispheres around March/April, and a minimum in October. Because CO2 is a tracer gas with little loss in the stratosphere and both a trend and a seasonal variation in the troposphere, the propagation of this seasonal signal into the stratosphere, observed during SPADE in 1993, gives an indication of both transport and mixing. Its variation is in phase with the seasonal variation in the Northern Hemisphere troposphere in March/April 1994 from 61#161#N to 68#161# S. This behavior of CO2 suggests that it enters the stratosphere in the tropics. The NO and NOy abundances are observed to increase, relative to expected values, in thin layers in the lower stratosphere and upper troposphere. The cause of these layers with enhanced NOy is not yet known. However, understanding these features has implications for the NOy budget in the lower straotsphere and upper troposphere, and thus for the assessment of the effects of current subsonic aircraft. The effects of enhanced albedo from high, tropical clouds have been observed for the first time on the abundances of the reactive gases NO, OH, HO2. Because the rates of chemical decomposition by sunlight are calculated and not measured for the stratosphere, these observations test these photolysis rates. The agreement is generally good between observed and calculated abundances of NO, OH, and HO2, implying that the calculated rates are reasonably accurate. Stratospheric sampling of Concorde and ER-2 On October 8th, 1994, off the northeast coast of New Zealand, the NASA ER-2 aircraft sampled the exhaust plume of a Concorde aircraft being operated supersonically by Air France. The ER-2 sampled a 320 km path that was traversed by the Concorde over a 10 minute period in three separate sampling legs, each of which required 30 minutes of sampling and was separated in altitude by about 1000 ft. The plume was encountered in the legs at and below Concorde flight altitude. Encounters with the Concorde exhaust plume were established by in situ observations of trace constituents made with instruments on board the ER-2 as follows: carbon dioxide (CO2), water vapor (H2O), nitrous oxide (N2O), carbon monoxide (CO), reactive nitrogen (NO, NO2, NOy), reactive hydrogen (OH, HO2) and condensation nuclei (CN). The plume was sampled on approximately 10 occasions with a duration that varied from 4 to 14 seconds. Instrument sampling rates varied from 0.3 Hz to 8 Hz corresponding to a spatial resolution between 600 and 25 m, respectively. The age of the plume as determined by the position of the encounter along the Concorde flight path varied from 10 to 70 minutes. CO2 reached peak values up to 2 ppmv above ambient values of 356 ppmv. The integrated change in CO2 above background values is needed to relate other constituent observations to the quantity of fuel burned. It is determined by using the measurement of N2O, which is not produced in jet engines, and the strong correlation between N2O and CO2 observed in the background atmosphere. The amount of water vapor observed in the plumes was less than the 1:1 mole ratio of H2O:CO2 expected from complete combustion. The slow response of the water vapor instrument is considered to be the cause of the low peak values. Carbon monoxide is a product of incomplete combustion. In all plume encounters, no CO increase was observed to coincide with that of CO2 and other species, whereas CO increases were found in the ER-2 exhaust plume measurements. The reactive nitrogen NOy, and the component species, NO and NO2, were observed in each encounter. The abundance of NOy ratioed to the abundance of CO2 yields the emission index (EI) of nitrogen oxides for the engine operating conditions. For the encounters with the largest integrated changes in CO2, the average EI value is 24 #177# 5 g (kg fuel)- 1 where NOy is expressed as equivalent mass of NO2. The EI for NOx reported during the CIAP program for the Olympus 593 Concorde engine is 18 g (kg fuel)- 1 as measured in ground based tests. Values found for the ER-2 exhaust were near 4 g (kg fuel)- 1. Measured ratios near unity for (NO + NO2)/ NOy indicate that NO and NO2 are the primary reactive nitrogen species in the exhaust plume up to an hour after emission. Condensation nuclei are a measure of small particles emitted by the engine or produced from the condensation of gases in the exhaust. All plume encounters contained CN concentrations elevated significantly above background values of 6 - 18 cm- 3. Peak values ranged up to 15,000 particles cm- 3. In most encounters, a large fraction of the particles were completely volatilized at 192#161#C, consistent with sulfuric acid composition. The absence of a plume signal in other particle sizing instruments on the ER-2 indicates that particle diameters are smaller than 0.09 micrometers. Values found in the ER-2 exhaust were generally below 200 particles cm- 3. Simultaneous measurements of OH and HO2 were made during each encounter with a spatial resolution of 25 m. In the youngest plume samples, OH was found to increase and HO2 to decrease. As the plume aged, the change in OH decreased. The perturbation of the ratio of OH to HO2 is consistent with the direct reaction of HO2 with NO in the exhaust. The Concorde observations are similar to those made in the exhaust plume of the ER-2 aircraft in that they both implicate HONO as the source of OH in the exhaust plume. This is consistent with the time evolution of the HOx signal in the Concorde samples and with observations of the ER-2 exhaust plume at night. Preliminary analysis of the HOx measurements establishes a minimum EI for HOx for the Concorde engine to be 8 #177# 2 percent of the NOx emission index. This is a lower limit since all the nascent OH is assumed to be converted to HONO in the immediate wake of the aircraft. The ratio of HOx/NOx EI values is comparable to values found for the ER-2 engine.