VOTE/TOTE Mission Plan

M. R. Schoeberl and O. B. Toon, Co-Project Scientists

Synopsis

The main objective of these experiments is to examine the production and dispersal of filaments from the tropical and polar vortex regions. Polar vortex air and tropical air are chemically distinct from midlatitude air. It currently is believed that some background exchange of air between the midlatitudes and these two regions takes place continuously through the intrusion of small-scale streamers or filaments of air into midlatitudes. Filament exchange is in contrast to the large exchange events observed in the satellite data. The purpose of this mission is to provide a quantitative assessment of the exchange process and our ability to predict filament exchange.

There are two added scientific objectives for this mission: examination of ozone/methane mixing ratios in the ozone loss region at the edge of the Arctic vortex, and examination of denitrification processes associated with high cold cirrus near the tropopause. The in situ measurements will also be helpful in assessing the level of stratospheric/tropospheric exchange taking place along the flight track.

To achieve our experimental objectives, we plan to use the DC-8 ozone/aerosol lidar and the new methane/water/temperature lidar to observe and follow filaments. In situ trace gas measurements (methane (CH4), nitrous oxide (N2O), ozone, and water (H2O)) are essential for providing DC-8 flight level information and for calibrating the lidar measurements. In a similar fashion, cross calibration of the lidar temperature measurements will be performed by the microwave temperature profiler (MTP). The addition of in situ nitric oxide (NO), reactive nitrogen (NOy), and aerosol instruments will allow us to achieve the added scientific objective of investigating NOy in aerosols and the possible denitrification of the lower stratosphere or upper troposphere by high cold cirrus.

The mission strategy calls for four test flights from Ames in June to July 1995; a tropical deployment in December 1995, and an Arctic deployment in early February 1996. We anticipate four to six tropical science flights and seven Arctic science flights. The tropical deployment will be from Hawaii, the polar deployment from Anchorage, Alaska. Each science flight will be from 8-10 hours in duration, requiring an augmented DC-8 flight crew. The tropical and polar missions will slightly exceed two weeks in duration so that the 30 flight hours per week restriction is met. We will need science overflight permissions for Greenland, and Canada.

1. Introduction

This document discusses the scientific objectives for a pair of aircraft missions to examine lower stratospheric transport processes. As a conceptual model, the tropical, midlatitude, and polar stratospheres can be considered separated by mixing barriers. Observations and modeling studies discussed below show that sharp gradients in long-lived trace gases will form at the edges of these barriers. Mixing appears to take place by the formation of streamers or filaments of air which break out of these regions. The formation, motion, and dispersal of these filaments is central to our understanding of the transport of trace gases between these regions.

The small scale of the filaments makes diagnosis of their morphology difficult. Satellite instruments can detect only the largest structures, thus we propose an aircraft investigation of tropical and polar filaments.

a. Tropical processes

Transport within the Tropics and between the Tropics and midlatitudes is much less understood than elsewhere in the stratosphere. In situ tracer observations are few and, because of the vertical range of the ER-2, mostly confined to a few kilometers above the high tropical tropopause. Further, because of the breakdown of balance approximations in the Tropics and the small number of tropical radiosonde sites, tropical winds are poorly known. These are serious limitations to our understanding of stratospheric transport. For one thing, tropospheric source gases (chlorofluorocarbons, CH4, N2O) enter the stratosphere in the tropical lower stratosphere; it is important to know how they are dispersed from there to the rest of the stratosphere. For another, knowledge of tropical transport is crucial to our ability to predict the fate of stratospheric aircraft emissions from proposed high-speed civil transports operating in midlatitudes: the transport rate of these emissions into the Tropics will determine not only interhemispheric transport but, by controlling the rate at which the emissions spread into the tropical upwelling, will also determine the transport to higher stratospheric levels, where residence times are long and chemistry rapid.

Increasingly, observations suggest that tropical air is, at least to some degree, isolated from midlatitudes. Aerosols from tropical volcanic eruptions tend to form `reservoirs' in the Tropics which decay slowly by erosion into midlatitudes [Trepte and Hitchman, 1992]; these reservoirs may terminate in a sharp edge in the subtropics [Grant et al., 1994, 1995]. Long-lived tracers also show such edges [Randel et al., 1993] and some pairs of tracers display interrelationships in the Tropics that are different from those of midlatitudes [Murphy et al., 1993]. Observations [e.g., Leovy et al., 1985; Randel et al., 1993] and contour advection calculations [Waugh, 1993; Waugh et al., 1994a, Polvani et al., 1995] also show filaments of tropical air being entrained into the midlatitude surf zone, sometimes all the way to the polar jet (see Figure 1). The subtropical edge appears to be most sharp, and entrainment most intense, in the winter hemisphere when midlatitude winds are westerly (see Figure 2). This has been confirmed by recent studies of the aerosol distribution. Grant et al. [1995] show the clear narrowing of the tropical barrier during winter (see Figure 3) in both hemispheres.

Figure 1. Location after 15 days of a material contour initially located on the 20-degree latitude circle, and advected with winds (from United Kingdom Meteorological Office analyses) on the 500K isentropic surface. [Waugh 1994]

Figure 2. Seasonal variation of poleward transport across 30deg N from 20deg N within 15 days. Note the brief summer minimum. [Waugh 1994]

Figure 3. Width and position of the boundary of the tropical stratospheric reservoir boundary for the year following the eruption of Mount Pinatubo, based on Advanced Very High Resolution Radiometer (AVHRR) weekly global aerosol optical thickness (AOT) maps. [Grant et al., 1995] Wide bars indicate that the aerosol boundary is diffuse, while the narrow one indicate sharp boundaries. Note that the Northern Hemisphere boundary is sharp beginning in November.

It is less clear to what extent midlatitude air is entrained into the Tropics [Plumb, 1995]; trajectory calculations using analyzed winds are unreliable within the Tropics and there few relevant observational data available. This is an important issue for the high-speed aircraft assessments because, if polluted air is entrained into the Tropics, it can be lofted to regions more sensitive to NOx ozone loss. Recently, Avallone and Prather [1995] have argued that the falloff of N2O with altitude is inconsistent with total tropical isolation and suggest that substantial mixing from midlatitudes would be needed. On the other hand, numerical model results [Norton, 1994; Polvani et al., 1994] indicate that isentropic transport into the Tropics may be weak but that the rate depends strongly on circumstances. Observational documentation will be crucial in addressing this issue.

b. Polar vortex - midlatitude interaction

The lower stratospheric winter polar vortex appears to be mostly isolated from midlatitudes [Schoeberl and Hartmann, 1989; Schoeberl et al., 1992]; however, Waugh et al. [1994b] have shown that a small amount material erodes from the polar vortex almost continuously in the form of small-scale streamers or filaments. Also, under rarer circumstances, material can be injected into the polar vortex [Plumb et al., 1994]. These filaments are predicted by contour advection (CA) calculations and may be the dominant form of vortex erosion below 20 km during most of the winter [Waugh et al., 1994b]. The scientific question posed by this erosion process is: to what extent could polar stratospheric cloud (PSC) processed air be responsible for influencing midlatitude chemistry and, possibly, midlatitude ozone loss reported by Stolarski et al. [1991]?

Seasonal CA calculations suggest that too little material is exported from the vortex to account for the midlatitude ozone loss, yet these calculations depend entirely on meteorological analysis and predictions that have had little validation. Figure 4 shows a contour advection calculation from Plumb et al. [1994] with the corresponding DC-8 flight track of January 24, 1992. During this period the DC-8 crossed the vortex edge and passed through a filament over western Canada. Figure 5 shows aerosol lidar measurements taken by the ozone/aerosol lidar aboard the DC-8. The lower part of the figure shows the CA calculated vortex elements. The filament extruded from the vortex is depleted in aerosol (far left side of the figure) and corresponds to the predicted position of the filament using the CA calculation. Although the CA calculation captures the position of this filament, the vertical structure is slightly off; that is, the filament tilts outward with height which is not predicted by their CA calculation.

Figure 4. Contour advection computation of the evolution of the vortex edge for January 24, 1992 from Plumb et al. [1994]. Dark line shows DC-8 flight track.

Figure 5. Aerosol lidar measurements from the DC-8 showing vortex structure. Note detached section with low aerosol amounts to the right of the figure. This piece corresponds to the filament shown in Figure 1.

A new method of computing streamers, reverse domain filling (RDF), has been developed. The technique is similar to the domain-filling method of Fisher and O'Neill [1993], except the calculation is run backwards from a uniformly distributed, dense distribution of parcels to a few days earlier [Sutton et al., 1994]. The final potential vorticity (PV) field is then plotted on the initial gridded domain. Figure 6 shows the RDF version of Figure 4. RDF provides more detail and a more complete picture of all the filamentation. The RDF cross-section of the Alaskan filament at 120deg W (see Figure 7), roughly corresponding to Figure 5, shows the filament extended off the vortex and tilted outward as seen in the data. The RDF method significantly enhances our filament predictability since all filaments are computed at once.

Figure 6. Reverse domain filling calculation of vortex filamentation corresponding to Figure 4.

Figure 7. Reverse domain filling cross-section of the vortex at 120deg W corresponding to Figure 5.

2. Objectives of the Tropical and Vortex Ozone Transport Experiments

The remarkable agreement of CA and RDF predictions and the aerosol observations in the polar regions [Plumb et al., 1994] for two DC-8 flights suggests an important breakthrough in understanding of transport processes erosion. Nonetheless, it would be difficult to predict the gross behavior of tropical-midlatitude and vortex-midlatitude exchange based upon these limited observations. The filament reported in Plumb et al. [1994] was quite robust; can smaller filaments be predicted? How accurate are our overall predictions? Can we reliably use CA calculations or RDF computations to quantify the amount of material being stripped from the Tropics or from the vortex through filamentation? How does the structure of the filaments vary with altitude? How do they dissipate? Midlatitude air is observed to be breaking into the polar air mass; does midlatitude air break into the Tropics? If so, this could have a major impact on our understanding of how trace gases are dispersed throughout the middle atmosphere as discussed above.

The main purpose of the Tropical Ozone Transport Experiment (TOTE) is to look at filament generation near the tropical barrier and to look at possible intrusion of midlatitude air into the tropical reservoir. The main purpose of the Vortex Ozone Transport Experiment (VOTE) is to make a quantitative assessment of filament generation and dissipation and to determine our ability to predict filament formation and structure.

To accomplish the above objectives, we propose to use a pair of DC-8 lidars to track ozone, aerosols, and the long-lived tracers CH4 and H2O to a height of about 20 km (8 km above the aircraft). Aerosol measurements from the lidar are shown in Figure 5. Aerosols, ozone, CH4, and H2O have significantly different vertical profiles within the vortex compared to the exterior region [Proffitt et al., 1993; Russell et al., 1993; Browell et al., 1993]. As filaments are pulled off the tropical barrier or the vortex, these trace gases retain their original amounts. Thus, tropical and vortex air can be differentiated from midlatitude air. The lidars will allow us to track the vertical structure of the filaments. An important element of this experiment is the need for in situ measurements of the same tracers as measured by the lidars to provide aircraft level validation for the lidars and data. Thus, in situ ozone, CH4, aerosol, and H2O measurements are important components of this experiment.

Ozone loss in the Northern Hemisphere vortex is an increasing concern as levels of stratospheric chlorine levels continue to rise. In situ ER-2 and DC-8 lidar aircraft measurements first demonstrated that ozone loss was occurring in the Arctic vortex [Schoeberl et al., 1990; Browell et al., 1990]. Satellite measurements appear to have confirmed the widespread ozone decrease using potential vorticity analysis [Manney et al., 1994]. Because ozone loss in the Arctic vortex is small compared to the Antarctic vortex, quantifying the amount of loss from year to year is difficult. One of the best indicators of ozone loss is the change in the ozone/long-lived tracer ratio [Schoeberl et al., 1990, Proffitt et al., 1993]. The planned payload for the DC-8 allows us to look at the ozone/methane ratio near the vortex edge where the ozone loss will be largest [Salawitch et al., 1993]. This objective of the mission can be met by performing vortex transects. However, because the VOTE mission will have only a 2-3 week duration in February 1996, we cannot determine the change in ozone/methane ratio during the winter season. One option is to perform a transect during the TOTE phase of the mission in late fall 1995; alternatively, we could compare measurements made during VOTE with previous ER-2 ozone/CH4 ratios from the first and second Airborne Arctic Stratospheric Expeditions (AASE I and II). Another option would be to use satellite data, but the Upper Atmosphere Research Satellite (UARS) Halogen Occultation Experiment (HALOE) sampling pattern does not provide deep vortex penetration during early winter.

The planned in situ payload for VOTE will also allow us to examine some of the lower stratosphere/upper troposphere processes related to the impact of nitrogen oxides from aircraft exhaust. One of the outstanding questions is whether or not high cold cirrus can act as a denitrifying agent in the upper troposphere, or can such cirrus act as sites for heterogeneous chemistry as do PSCs in the vortex? Thus, with the addition of instruments which measure NOy and NO, we can estimate the denitrification by high cold cirrus. Using particle sizing instruments, we should be able to tell if the nitrogen is present in aerosols or cloud droplets. We may also be able to tell if the particles are frozen or not (from the aerosol lidar depolarization), and whether the amount of NOy present in particles is consistent with theory (from measurements of NOy, particle volume, H20, and temperature). We will also be able to further study the nitrogen chemistry in any aircraft plumes which may be intersected, including that of the DC-8.

Finally, the extensive transects made during these missions will allow us to examine upper troposphere trace gases in detail and possibly assess the level of stratospheric/tropospheric exchange in the flight regions.

3. Instrument Payload

Instrument descriptions are included in another section of this document. The strategy for the experiment requires lidar measurements of ozone and aerosols. This can be accomplished by the NASA Langley Research Center Differential Absorption Lidar (DIAL) (E. Browell, PI). Long-lived tracer measurements can be made by the NASA Goddard Space Flight Center CH4 lidar, which also measures H2O and temperature (W. Heaps, PI). The temperature measurement is critical for placing the observations on a potential temperature surface.

In situ measurements of long-lived trace gases (CH4, N2O, carbon monoxide (CO), and H2O) can be made by the NASA Langley tracer payload (G. Sachse, PI). Ozone, NO, and NOy can be measured by the National Center for Atmospheric Research (NCAR) trace gas payload (B. Ridley, PI). In situ aerosol measurements can be made with the NCAR Multiple-angle Aerosol Spectrometer Probe (MASP) (D. Baumgardner, PI) and the NASA Ames Forward Scattering Spectrometer Probe (FSSP) Model 300 (R. Pueschel, PI). The Jet Propulsion Laboratory Microwave Temperature Profiler (MTP) (B. Gary, PI) will provide near-aircraft temperature measurements.

Test flights of the new Pennsylvania State University OH instrument (W. Brune, PI) will occur during TOTE/VOTE as the opportunity arises.

4. Mission Strategy

a. TOTE

Tropical PV calculations show that there are two Pacific sites where the tropical barrier extends northward, near the international dateline in the mid-Pacific and in the eastern Pacific. The mid-Pacific would thus be a preferred region for filaments to extend from the tropical region. Our overall plan is to fly near the tropical barrier on a series of survey flights searching for filament formation based upon the meteorological analysis. The ideal location for this mission is Hawaii, from which we can range either southeast or southwest toward the barrier.

Transit flights between perhaps Guam and Hawaii would also cover a region where the tropopause temperature approaches nitric acid trihydrate condensation levels.

The tropical Pacific and Atlantic regions would also be zones of `clean' air which can be contrasted with polluted flight corridor regions. We will be able to contrast the data taken in this region with data taken as part of the western Pacific Exploratory Mission (PEM-West). The TOTE flight series will begin with a northern survey flight which, along with VOTE data, will be used to estimate ozone loss relative to the methane distribution.

b. VOTE

Using CA methods, filament calculations have been run for the last three winters. Filaments tend to be generated all winter (December through March), and become most separated from the vortex near western Canada in the Aleutian anticyclone region. Thus we plan to use Anchorage or Fairbanks, Alaska as a base of operations to observe filaments well removed from the vortex. One of the advantages of using the DC-8 is that its long range will allow us to find and track filaments. Real-time flight planning of this type was used during AASE II to probe the vortex edge. Occasionally filaments do not form for a period (e.g., when the vortex is quiet). These periods are not generally longer than two weeks as indicated by our last three year integrations; thus, we believe that a two-week mission is needed to maximize the number of filament observations.

To meet the objective of measuring vortex ozone loss, the VOTE mission should be flown in February 1996 when high chlorine monoxide levels will have had time to deplete ozone. This time frame also meets with DC-8 availability and allows for adequate instrument integration time. The proposed flight plan would be as follows: 29 January 1996, DC-8 to Anchorage; 30 January to 14 February, deployment out of Fairbanks to study filaments and to look at ozone loss; 15 February, transit to Barbers Point, Hawaii; 16-18 February, Hawaii deployment in conjunction with the Stratospheric Tracers of Atmospheric Transport (STRAT) aircraft campaign; 19 February, return to NASA Ames. These dates are approximate and for illustrative purposes only. The actual dates depend on DC-8 availability and on meteorological conditions.

During the VOTE mission, if high cold cirrus formations are predicted, the aircraft will be diverted to try and sample those regions. The missions will be mostly at night for lidar operation. In addition, moonless or near-moonless nights are preferred to improve signal to noise in the methane lidar. Vertical profiles between flight altitude and about 20 kft will be needed to calibrate the CH4 lidar and to perform in situ profiling. At least one pass through the DC-8 exhaust trail will be performed, and possibly others will be done through a DC-8 contrail during daylight and at night. The objective of these wake crossings is to continue analyses of aircraft emission factors and to look for evidence of nitrogen chemistry occurring on contrails.

5. Meteorological/Chemical Modeling

Forecasts of filament positions and structure and chemical forecasts of ozone loss will be necessary for mission success. Tropospheric ozone production and losses must also be characterized. These forecasts will be provided by a meteorological/ chemical modeling team. Meteorological forecasts for flight planning will be provided by NASA Goddard Space Flight Center, which has produced the appropriate products on previous ER-2 and DC-8 missions. These products will be faxed to field sites as required. Chemical (`point') and trajectory models will be applicable for both TOTE and VOTE, but the tropospheric ozone maps use a technique which is applicable only in the Tropics and, hence, will not be appropriate for VOTE.

Below are suggested meteorological/chemical modeling team members:

R. A. Plumb, Massachusetts Institute of Technology; and D. Waugh, Monash University (Australia) - Filament position forecasts
A. R. Douglass and R. Kawa, Goddard - Chemical model forecasts
P. A. Newman, L. R. Lait, Goddard; and L. Pfister, Ames - Flight planning meteorological information
A. M. Thompson, Goddard; and R. D. Hudson, University of Maryland - Tropospheric chemical and trajectory modeling and tropical tropospheric ozone maps

6. Schedule

June-July 1995 - Integration of instruments on aircraft and test flights with Synthetic Aperture Radar (SAR). This payload will be limited to modified and new instruments: NOy, CH4 lidar, aerosol, and MTP.
13-14 September 1995 - TOTE/VOTE science team meeting at Ames.
27 October to 4 December 1995 - TOTE integration of instruments on DC-8, two test flights.
7 December 1995 - Survey flight north to Alaska to provide a pre-winter assessment of trace gas distributions. TOTE experiment begins. Deployment will consist of transit flight to Hawaii and four science flights in the tropics (deployment from Hawaii).
Early January 1996 - Test flights for VOTE or alternatively, if TOTE flights are delayed by planned DC-8 maintenance, tropical deployment will take place during this period.
29 January 1996 - VOTE experiment begins. Deployment will consist of science flight to Anchorage or Fairbanks; science flights from Anchorage; and science flight to Thule, Greenland and back. Possible STRAT coordination flight in mid-February out of Hawaii.
19 February 1996 - Return to Ames.
11-12 June 1996 - Follow-up science team meeting on data analysis in Colorado.

7. References

Appendices:

Instrument and Model Summaries

Range of Proposed Flight

Radius of operations from Guam, Hawaii, Ames, Anchorage, and Thule

Tropical Ozone Transport Experiment (TOTE) and Vortex Ozone Transport Experiment (VOTE)