15031188-Integrated-Study-of-Clouds-and-Radiation-of-the-Upper-Troposphere-and-Lower-Stratosphere by stephinrazin


									NCAR UTLS White Paper DRAFT v1a. October 17, 2003

Integrated Study of Dynamics, Chemistry, Clouds and Radiation of the
Upper Troposphere and Lower Stratosphere
Laura Pan, Brian Ridley, Bill Randel, Andrew Gettelman, Andy Heymsfield, Mary Barth,
Alan Fried, Chris Cantrell, Todd Lane, Don Lenschow, Steve Massie, Owen Cooper,
Alyn Lambert, Mike Coffey, Sue Schauffler, Paul Wennberg, Guy Brasseur

Outline                                                                     Page

I.     Summary                                                                 1
II.    Background and Key Issues                                               3
III.   Research Tools                                                          7
III.1. Airborne Observational Capability                                       7
III.2. Satellite Platforms                                                     8
III.3. Models                                                                  8
IV.    Scientific Objectives and Proposed Studies                              9
IV.1. Water Vapor, Radiation, and Microphysics of the TTL                      9
IV.2. Extratropical Stratosphere-Troposphere Exchange                         14
IV.3. Ozone and Radical Budget Chemistry in the UTLS                          21
IV.4. Aerosol and Cloud Particle Composition in the UTLS                      27

V.     Concluding Remarks                                                     31
References                                                                    32
Figures                                                                       38
 1   I.       Summary
 2           Identifying and understanding the dynamical, chemical and physical processes that
 3   control water vapor, ozone, radical constituents, aerosols, and clouds and their impact on
 4   the radiative balance of the Upper Troposphere and Lower Stratosphere (UTLS) are
 5   critical for advancing the reliability of predictions of climate change or of trends in global
 6   air quality. The UTLS is a highly coupled region: dynamics, chemistry, microphysics and
 7   radiation are fundamentally interconnected. For example, the abundance of two
 8   radiatively important trace gases, water vapor and ozone, in the UTLS is controlled by
 9   dynamics including deep convection and two-way stratosphere-troposphere exchange,
10   chemical processing including multiphase chemistry, and cloud microphysics, which in
11   turn are all influenced by temperature and aerosol distributions.
12           The UTLS has been an under-sampled region. The altitude range is often below the
13   detection range of the spaceborne instruments and there are only a few high altitude
14   airborne observing platforms. The strong gradients in thermal and chemical structure near
15   the tropopause pose challenges to most global and regional models. A suite of new
16   observational capability, which includes the high altitude NSF HIAPER aircraft for
17   detailed process studies, the NASA A-Train satellite data for larger scale investigations
18   and temporal changes, and the significant advances in in-situ measurement techniques for
19   a large suite of chemical and aerosol/cloud particle constituents, together with the recent
20   advances in in-house and community models present an unprecedented opportunity for an
21   integrated UTLS study and provide the incentive for this NCAR Strategic Initiative.
22           The objectives of this white paper are to identify the critical areas for
23   understanding the role of the UTLS in global change, to outline the approaches we plan
24   to take in addressing the outstanding issues, and to define achievable goals for the
25   initiative. The issues we identified for the project are defined into four interrelated
26   themes. Each theme will be addressed through integration of the modeling, satellite, and
27   in situ tools.
28   (1) Tropical UTLS water vapor, clouds, microphysics, and radiation. The focus is on
29        improving our understanding of the processes that maintain the observed water vapor
30        distribution and trends in especially the tropical UTLS, the microphysics of cirrus
31        formation and evaporation, and the role of deep convection and their effects on the
32        radiation budget. Water vapor is a major source of OH, a driver of photochemical
33        processing.
34   (2) Two-way stratosphere-troposphere exchange (STE) processes. The focus is to
35        establish an improved climatology of STE in the extratropics, to provide needed input
36        for refining model simulations, and to improve our understanding of how dynamical
37        processes over multiple scales control the location and seasonality of irreversible
38        exchange and mixing processes.
39   (3) Chemistry that controls the budgets of ozone and radical species in the UTLS.
40        The focus is on gaseous and multiphase processes that control the sources and sinks
41        of radical constituents (HOx, NOx, ROx, ClOx, BrOx…), and hence the processes that
42        control the budget of O3 and those that ultimately control chemical transformations
43        and removal of many chemical pollutants. Of particular interest is the role of deep
44        convection in which a large variety of near-surface biogenic and anthropogenic
45        emissions or oxidation products of near-surface photochemistry can be rapidly
46        transported to the UTLS, or produced and injected by lightning activity.
47   (4) Composition of aerosol and cloud particles in the UTLS. The focus is on
48        determining the processes controlling formation of aerosols and cloud particles in the
49        UTLS, the chemical composition of aerosols, and how the composition influences the
50        generation of cirrus particles. Furthering our understanding of the formation of polar
51        stratospheric clouds (PSCs) and their role in ozone destruction in the polar

     NCAR UTLS White Paper DRAFT v1a                                                             1
 1       stratosphere is another focus. Refining our understanding of multiphase processing of
 2       chemical constituents in/on liquid and ice particles is of particular importance to
 3       advance both detailed microphysical/chemical models and sub-grid scale
 4       parameterizations in global models.
 5          The goal of the UTLS initiative is to plan and to conduct integrated studies of
 6   these issues using the new HIAPER aircraft in conjunction with observations from the
 7   NASA A-Train satellites and with NCAR modeling tools.
 8          Process studies using the HIAPER platform are integral to this initiative and five
 9   specific targets are summarized below. The timing of these projects will depend critically
10   on instrument development/modification for the new aircraft. These projects will be
11   conducted in concert with observations from AURA (and other A-Train satellites). In
12   particular, we expect the HIRDLS, MLS, and TES instruments to provide unprecedented
13   data for the UTLS. The satellite data and/or predictions from global, regional, and
14   mesoscale models may well provide impetus for the final design of airborne projects. In
15   turn, the airborne data will provide comparisons for satellite instrument validation.
16   (1) Extratropical stratosphere-troposphere exchange (STE). This experiment will
17       investigate STE in the extratropical tropopause region near the subtropical jet using a
18       variety of tracers together with high resolution temperature and wind measurements.
19       Additional objectives are to perform instrument comparisons during the start-up
20       phase of HIAPER flights and to contribute to AURA validation. The initial phase of
21       this experiment will be proposed to NSF as a part of the HIAPER “Progressive
22       Science” program (possibly July-December 05, details TBD by NSF). A more
23       extensive experiment is planned for Spring 06.
24   (2) Tropical water vapor, convection and clouds experiment. A tropical composition,
25       cloud, chemistry and climate coupling experiment (TC4) has been formulated by a
26       group of scientists in conjunction with AURA satellite validation. We propose the
27       HIAPER’s participation in the ‘07 phase as its altitude capability makes it uniquely
28       suitable for probing the main convective outflow region (10-14 km) and altitudes
29       within the middle troposphere.
30   (3) Radical budget experiment. The objectives are to determine radical production
31       rates, identify critical radical precursors, and compare measurement and model
32       calculations of HOx in air masses affected and not affected by convection, by cirrus
33       particles, and by high NOx levels. Studies over the relatively clean lower troposphere
34       (the upper Great Plains) and over more polluted regions (the southeastern U.S.) are
35       proposed.
36   (4) Polar PSC experiment. Objectives are to identify the chemical composition and
37       particle size distributions of PSCs, to observe the PSCs under a variety of conditions
38       to investigate formation mechanisms, and to measure the effects of PSCs on reactive
39       species (e.g. O3, NO2, HNO3). Scandinavia would be the base of operation.
40   (5) Midlatitude cirrus cloud formation near the tropopause and aerosol
41       distributions in the UTLS. The objective is to investigate the microphysics of
42       formation of in-situ and anvil derived cirrus near the extratropical tropopause, the
43       interactions with gaseous constituents, and the chemical composition of aerosols and
44       ice particles.
45   The design of these experiments and the possibility of combining objectives will be
46   refined at the upcoming October workshop. The involvement from both the experimental
47   and modeling communities will be formalized at the workshop. Collaborations with the
48   international community are currently under discussion with the science steering group of
49   SPARC.

     NCAR UTLS White Paper DRAFT v1a                                                          2
 1   II.     Background and Key Issues

 2   The UTLS region is roughly defined as the part of atmosphere between 5 and 20 km.
 3   Figure 1 highlights the structure of this region and how various activities couple.
 4   Dynamically, the region is a transition zone between the convectively dominated
 5   troposphere and the stably stratified stratosphere, and separates radiatively and
 6   chemically distinct air masses. The tropopause is fundamentally identified by the change
 7   in static stability between the troposphere (low stability) and stratosphere (high stability).
 8   Several criteria such as the thermal lapse rate, the temperature minimum (cold point),
 9   static stability, and potential vorticity have been used to define the tropopause. The
10   region is characterized by strong gradients in many trace constituents with tropospheric
11   or stratospheric origin (such as water vapor and ozone), and such tracers have also been
12   used to define the tropopause. Because different physical processes determine the
13   tropopause behaviour in the tropics and the extratropics, it is useful to separate the
14   discussion of these regions.

15   In the tropics, the tropopause is relatively high, with the cold point near 17 km. The
16   tropospheric lapse rate (up to 12-14 km) is determined by radiative-convective
17   equilibrium, while the thermal structure above ~14 km is primarily in radiative balance
18   (characteristic of the stratosphere; Thuburn and Craig, 2002). Overall the region of the
19   tropical atmosphere between ~12 km and the altitude of the cold point (17 km) or slightly
20   above has characteristics intermediate between those of the troposphere and stratosphere,
21   and is referred to as the Tropical Tropopause Layer (TTL). Thin (sometimes subvisible)
22   cirrus clouds are observed over large areas of the TTL (Wang et al., 1996; Winker and
23   Trepte, 2000), although their formation mechanism(s) and effects on the large-scale
24   circulation and humidity are poorly known.

25   In the extratropics the tropopause level is highly variable depending on the location of the
26   polar jet stream which meanders back and forth across this region during all seasons.
27   Typical tropopause heights are 12-14 km on the sub-tropical side of the polar jet stream
28   and 8-10 km on the polar side. The tropospheric lapse rate in midlatitudes is determined
29   by baroclinic adjustment (the net effect of tropospheric baroclinic waves), while the
30   stratosphere is again in radiative equilibrium. Annually averaged, there is a slow mean
31   downward motion in the extratropical stratosphere, in balance with tropical upwelling.
32   The extratropical tropopause region is characterized by a high degree of synoptic and
33   smaller-scale variability, related to transient baroclinic wave activity (synoptic weather
34   systems, and associated fronts, tropopause folds, and high frequency wave variability).

35   Recent work has highlighted the importance of the UTLS region for a number of issues,
36   including the radiative effects of clouds, aerosols and water vapor in the present and
37   future climate (IPCC, 2001), cloud life cycles and mechanisms of dehydration in the
38   tropics, understanding trends of stratospheric water vapor and ozone, the details of UTLS
39   radical chemistry and its impact on ozone production/loss, and mechanisms of
40   extratropical stratosphere-troposphere exchange. In each case, current global models are
41   beginning to incorporate the three-dimensional resolution and physical mechanisms that

     NCAR UTLS White Paper DRAFT v1a                                                              3
 1   control the relevant dynamical, radiative, microphysical and chemical processes. The key
 2   issues of UTLS research are therefore fundamentally interconnected. We have identified
 3   the following topics as key issues for improving our understanding of the UTLS region:
 4   (1) Detailed radiative balance in the tropical tropopause region; (2) Processes controlling
 5   the dehydration of air entering the stratosphere; (3) Cloud microphysics; (4) Variability
 6   of aerosols; (5) Ozone and radical budgets, and oxidation processes; (6) Effect of deep
 7   convection; (7) Multiphase chemistry; (8) Stratosphere-troposphere exchange.

 8   II.1.   Detailed radiative balance in the tropical tropopause region

 9   In the tropics, the background clear sky radiative balance shifts from cooling in the
10   troposphere to heating in the stratosphere, with the transition occurring around 15 km.
11   The region of heating above ~15 km corresponds to mean upward motion into the lower
12   stratosphere. This marks the base of the stratospheric Brewer-Dobson circulation, so that
13   the physical and chemical characteristics of the air in this region behave as boundary
14   conditions for the global stratosphere.

15   Folkins (2002) and Hartmann and Larson (2002) have noted the coupling between
16   radiatively determined mass fluxes and the altitude where tropical cirrus clouds detrain.
17   There are, however, significant differences in the simulated radiative balances in current
18   large-scale models. Detailed observations of both the vertical structure of water vapor in
19   the upper troposphere, and the radiative fluxes in the far IR, are necessary to better
20   understand radiative forcings and feedbacks in the UTLS.

21   II.2.   Processes controlling the dehydration of air entering the stratosphere

22   One of the current outstanding issues in stratospheric research is the origin of the
23   increasing trend in stratospheric water vapor observed from a long record of balloon
24   measurements at Boulder, Colorado during 1980-2002 (Oltmans et al., 2000). The
25   observed trends of approximately 1%/year can have substantial radiative and chemical
26   effects if representative of the global stratosphere, and if they continue in the future.
27   However, the cause of the Boulder trends is unknown at present, making future
28   predictions uncertain. A simple explanation could involve a slight warming (~1
29   K/decade) of the tropical tropopause region, which modulates the humidity of air entering
30   the stratosphere. However, observations show the tropical tropopause has cooled slightly
31   over the 1980-2000 time period (Randel et al., 2000; Seidel et al., 2001). This calls into
32   question current understanding of the processes that control stratospheric humidity, in
33   particular dehydration near the tropical tropopause, and interactions with deep convection
34   and tropical cirrus (as discussed in the idealized models of Holton and Gettelman, 2001,
35   and Sherwood and Dessler, 2001).

36   II.3.   Cloud microphysics

37   Upper tropospheric clouds are crucial components of the earth’s energy and water

     NCAR UTLS White Paper DRAFT v1a                                                          4
 1   balance system. They exert strong influences on the radiative balance and affect the
 2   distribution and amount of precipitation in the troposphere. They are also key elements
 3   in determining the sources and distribution of water vapor in the stratosphere. Cirrus
 4   near the tropopause influences radiative transfer in a complex manner. If the cirrus is
 5   above clear sky regions, then a net warming occurs (McFarquhar et al, 2000), whereas if
 6   the cirrus is over deep convective anvils, then there is a net cooling (Hartmann et al.,
 7   2001). Current work with NCAR’s CAM2 indicates that cirrus may be important for
 8   maintaining (i.e. warming) tropopause temperatures in the tropics (Boville, 2003,
 9   personal communication). Since deep convection only occupies a small percentage of the
10   tropics, roughly 8% of the geographical region, cirrus near the tropopause will primarily
11   have a warming effect on the UTLS. Details of cirrus formation, maintenance, removal,
12   space-time variability of the particle size distributions, and interactions with chemical
13   constituents are poorly understood.

14   Seasonal distributions of cirrus observed from satellites reveal distributions primarily
15   over Indonesia (the maritime continent), Africa, and South America (Wang, 1996;
16   Massie, 2000). While the cirrus is highly correlated to deep convection, observations also
17   indicate that cirrus can form in-situ (Pfister et al., 2001; Massie et al., 2002), and may be
18   enhanced by the temperature perturbations of atmospheric waves (Jensen et al., 1996;
19   Boehm and Verlinde, 2000) The specification of the radiative effects of subvisible cirrus
20   in global circulation models therefore first requires a better understanding of the
21   processes that produce the cirrus, both from deep convection activity and in-situ
22   formation processes.

23   II.4.   Variability of aerosols

24   The distribution, composition, formation, and radiative properties of UTLS aerosol
25   particles are also critical for the Earth’s radiative balance. As stated in the Strategic Plan
26   for the Climate Change Science Program Review draft, November 2002, One of the
27   largest uncertainties about the impact of aerosols on climate is the diverse warming and
28   cooling influences of the very complex mixture of aerosol types and their spatial
29   distributions. In addition to their radiative impact, a subset of the aerosol population act
30   as cloud condensation nuclei (CCN). In response to lifting and humidity, CCN will grow
31   to haze or cloud droplets, causing a substantial increase in radiative effects. These
32   aerosols may also serve as sites for subsequent ice particle nucleation. Processes for
33   aerosol-cloud-radiation inter-connections are therefore complicated and challenging
34   issues. Since the radiative properties (warming or cooling by particles) are dependent
35   upon composition, the determination of aerosol composition in the UT/LS is an important
36   issue.

37   II.5.   Ozone and radical budgets, and oxidation processes

38   The tropopause region distinguishes the transition of global change concern from
39   depletion of ozone in the lower stratosphere to growth of ozone in the troposphere due to

     NCAR UTLS White Paper DRAFT v1a                                                             5
 1   trends in release of anthropogenic contaminants. Ozone production in general exceeds
 2   destruction in the UT as a result of reactions involving HOx (HO2, RO2, OH, …) radicals
 3   and NOx (NO, NO2, ..) fueled by sunlight, and to a large extent by CO and CH4. Gas
 4   phase photochemical processing is thought to be reasonably well understood in the
 5   background UT (and LS). More recent studies have shown that the production of HOx in
 6   the dry UT can be strongly perturbed by oxidation of longer lived VOCs (volatile organic
 7   carbon constituents) (Wennberg et al., 1998) and by shorter lived VOCs, peroxides, and
 8   halogen radical precursors that can be rapidly transported to the UT, together with a large
 9   variety of other lower tropopshere pollutants, in deep convection or by uplift in frontal
10   systems. Indeed it is likely that some reactive constituents that can perturb the radical
11   production rates, and hence the ozone budget, are yet to be identified. Since O3 is an
12   effective greenhouse gas in the UTLS region, changes in ozone can have an impact on
13   climate. Changes in climate, e.g., changes in the water vapor distribution, will impact
14   chemistry.

15   II.6.   Effects of deep convection

16   Lightning production of NOx in electrically active deep convection can also significantly
17   alter, even saturate, the local net ozone production rate and also affect HOx partitioning.
18   Such processes may also transport sources of halogen radicals from the oceans to the
19   UTLS region that ultimately increase O3 loss rates. Thus deep convective systems can
20   present an unusual suite of reactive constituents to near the tropopause, the upwelling
21   region into the stratosphere in the tropics, or to regions of smaller scale troposphere to
22   stratosphere transport in the extra-tropics.

23   Deep convection may also perturb the tropopause altitude or even penetrate the
24   tropopause and transport the pollutants and chemical products from biomass burning into
25   the stratosphere (Fromm & Servranckx, 2003). In current global models the treatment of
26   deep convection is a subgrid-scale parameterization and remains a weakness. Process
27   studies of such systems in combination with detailed cloud models are required to
28   improve the large-scale model treatment and combined chemical effects.

29   II.7.   Multiphase chemistry

30   Deep convective outflow can also be a source of cirrus layers to the UT and possible ice
31   crystal/chemistry interactions or other gas-aerosol processes that are only beginning to be
32   explored observationally and theoretically. For example, it is not known whether uptake
33   of chemicals affects the microphysics of cirrus formation or whether radical loss is
34   significantly increased in the presence of ice crystal or other aerosols. Incorporation of
35   chemicals into the aqueous phase and their reactions in the cloud drops as air parcels are
36   convectively transported from the boundary layer to the upper troposphere may enhance
37   or reduce the source of radicals in the UT. In polar regions, there remain outstanding
38   details concerning the formation, evolution, chemistry, and composition of polar
39   stratospheric clouds responsible for triggering major ozone loss.

     NCAR UTLS White Paper DRAFT v1a                                                          6
 1   II.8.   Stratosphere-troposphere exchange

 2   Two-way stratosphere-troposphere exchange (STE) couples the UT and LS and is a key
 3   controlling process for the UT ozone budget and LS water vapor budget. It is also a key
 4   factor for understanding the anthropogenic impact, such as biomass burning, to long term
 5   climate change. Mixing between LS and UT air alters not only the mixing ratios of
 6   various species, but also the chemistry and reaction rates. For example, a box-model
 7   study by Esler et al. (2001) suggests that the mixing of stratospheric ozone and
 8   tropospheric water vapor leads to enhanced hydroxyl radical concentrations compared to
 9   background tropospheric and stratospheric values, which increases the oxidation of CO,
10   methane and higher hydrocarbons. The extent, the seasonality and the mechanisms of
11   exchange need to be better quantified, in order to quantify the ozone, water vapor and the
12   distribution of aerosols in the UTLS.

14   These issues present a set of interconnected questions, some of which are: What controls
15   the level of mean convective outflow in the tropics? How does cloud microphysics
16   impact the dehydration process? How is the formation of cirrus clouds related to deep
17   convection and gravity waves? How do homogenous and heterogeneous freezing
18   contribute to cirrus formation? How is UTLS chemistry affected by convective transport
19   of constituents from the lower troposphere, from land regions versus ocean regions?
20   What are the multiphase processes that affect the production and loss of radical species
21   (HOx, NOx, ClOx, BrOx, IOx, …) in the UT and across the tropopause to the LS? What is
22   the impact of VOC oxidation on HOx radical budgets? What short-lived VOC
23   constituents are important sources of HOx in deep convective outflow? What is the
24   contribution of NOx and possibly other constituents from lightning? How is the lightning
25   production distributed over altitude? How can we better understand the mechanisms of
26   PSC formation in the polar region and the impact on ozone trends? How can we improve
27   our ability to accurately model ozone depletion in the polar regions? What is the impact
28   of multiphase chemistry processes in the UTLS region? What chemical reactions occur
29   on cirrus particles? What is the importance of new particle formation and outflow from
30   convection on regional and global scales? How can we better quantify the contribution of
31   chemistry and transport to the ozone budget using satellite remote sensing, in situ
32   measurements, and model calculations? How are processes of different scales coupled to
33   produce the transport and mixing between the stratosphere and troposphere?

34   To effectively address some of the questions raised above with the available and new
35   tools requires that we optimize the relation between aircraft and satellite measurements to
36   modeling over a variety of spatial scales.

37   III.    Research Tools

38   III.1. Airborne Observational Capability
39   The HIAPER aircraft is capable of carrying a payload of 6000 lbs initially to 12.5 km and

     NCAR UTLS White Paper DRAFT v1a                                                          7
 1   to a maximum of 15.5 km as fuel is depleted, at a cruise speed of 850 km/hr. The aircraft
 2   has a long endurance of ~9 hours with simple cruise flight tracks and thus can sample
 3   polar regions or remote ocean regions from convenient airports. Flights requiring
 4   repeated profiling will be necessarily shorter. At mid-to-high latitudes flights well into
 5   the LS can be made. In the tropics, the LS will be inaccessible for in situ instrumentation
 6   but sampling in the altitude region of major convective outflow will be possible.

 7   A concerted effort of new instrument development (preferably autonomous) or
 8   modification will be required in conjunction with flights for test and evaluation. Critical
 9   design of in-situ sampling inlets and instrumentation is necessary to minimize weight and
10   size to achieve reasonable sampling time (> 1hr) at altitudes approaching the maximum
11   flight altitude and to allow the large suite of instrumentation required to achieve progress
12   on the key issues described above.

13   The C-130 can carry much larger payloads (13000-23000 lbs) but only to 7.5 km altitude.
14   Its cruise speed is ~460 km hr-1. For the field studies proposed, the C130 would deploy
15   remote sensing instrumentation (e.g., O3, aerosol Lidars) to probe the UTLS and in deep
16   convective studies employ both remote and in situ sensors to chemically characterize the
17   low altitude source regions and possibly the mid-level outflow and inflow regions.

18   III.2. Satellite Platforms

19   A new generation of satellite instruments is poised to contribute measurements of
20   meteorological and chemical structure in the UTLS region with unprecedented horizontal
21   and vertical resolution. Several instruments that will contribute directly to UTLS
22   research at NCAR include (1) the Atmospheric Infrared Sounder (AIRS), (2) the
23   Moderate Resolution Imaging Spectroradiometer (MODIS) instrument on the NASA
24   AQUA satellite, and (3) the High Resolution Dynamics Limb Sounder (HIRDLS) and
25   Troposphere Emissions Sounder (TES) instruments on the NASA AURA satellite (to be
26   launched in January 2004). AIRS will contribute high vertical resolution (~1 km)
27   temperature and water vapor measurements with an approximate 50 km horizontal
28   sampling (globally), which will allow novel analyses of processes controlling UTLS
29   humidity (especially in the tropics). Figure 2 gives examples of AIRS data. On the other
30   hand, MODIS is providing unprecedented information on clouds. The CALIPSO LIDAR
31   and the CLOUDSAT microwave experiments will measure the vertical structure of sub-
32   visual cirrus and larger (several hundred microns in diameter) cirrus near and below the
33   tropopause, in addition to particle state (liquid vs solid form). HIRDLS will contribute
34   high resolution trace constituent measurements in the UTLS, including water vapor,
35   ozone and other species. These constituent measurements will be especially valuable for
36   studying exchange processes near the tropopause (such as tropopause folds in
37   midlatitudes), for providing global perspective to localized aircraft data sets, and for
38   direct comparison with model simulations.
40   III.3. Models

     NCAR UTLS White Paper DRAFT v1a                                                           8
 1   A suite of models is available to study in conjunction with aircraft and satellite
 2   observations the dynamics, chemistry, microphysics and radiation of the UTLS. The
 3   participating models span scales from the cloud scale to global, and exist in MMM, RAP,
 4   ACD and CGD. Model analyses will be an important part of the experimental design and
 5   post analysis and interpretation. A partial list of the models and their role in this initiative
 6   are outlined below.

 7   Several mesoscale models will be used in conjunction with airborne experiments. The
 8   Weather Research and Forecast (WRF) Model will be used for regional modeling efforts,
 9   both for forecasting to support mission planning and for post analyses to provide
10   meteorological variables. The WRF-chemistry model will also be used for cloud-scale
11   simulations of deep convection to determine the processes (e.g. multiphase chemistry,
12   precipitation formation, lightning production, etc.) that create the distribution of relevant
13   chemical species. The Clark-Hall cloud model (Clark et al. 1996) can be utilized to
14   represent small-scale processes such as gravity wave and turbulence generation by
15   convection and the jet stream. Several global models will be used to work with satellite
16   observations and to connect the aircraft observations to the satellite results. These
17   models include WACCM and CAM2, coupled chemistry versions of these models based
18   on the MOZART platform, and the Chemical Lagrangian Model of the Stratosphere
19   (CLaMs) (McKenna et al., 2002). MOZART 3, a unique global CTM that covers the
20   troposphere and stratosphere, is ideally suited to analysis of the UTLS region. Figure 3
21   gives an example of MOZART3 simulation for the tropopause region near the subtropical
22   jet, showing the intrusion of stratospheric ozone into the upper troposphere. Finally, a
23   suite of process models, particularly chemical solvers and microphysical models, are also
24   available for this effort, that help bridge the scales of the other models, and act as
25   parameterizations or checks on parameterizations for larger models. This suite of
26   simulation tools enables us to work across scales from modeling of individual cirrus
27   clouds particles, to individual clouds, to regional transport, to global climate implications.

28   IV.     Scientific Objectives and Proposed Studies

29   Section II describes a broad range of possible UTLS investigations. Here we identify
30   four more focused projects that are centered on the new tools available. In practice, the
31   field programs considered that are centered on HIAPER will likely be combinations of
32   parts of several projects. Instrument development required for some projects may well be
33   the rate-limiting process. On the other hand, many of the models are currently available
34   for investigation and analysis. In October 2003 a workshop is to be held that will refine
35   ideas presented here or define the development of other projects.
37   IV.1. Water vapor, Radiation, and Microphysics of the TTL
38   The tropical UTLS is the main radiating region of the heat engine of the climate system.
39   Chief among the uncertainties is the role of water vapor, and the ‘water vapor feedback’
40   to increasing greenhouse gas concentrations. There remain critical uncertainties in the
41   basic large scale physical relationships among clouds, water vapor, and radiation in the
42   tropical UTLS. These uncertainties inhibit understanding of the boundary conditions for

     NCAR UTLS White Paper DRAFT v1a                                                               9
 1   the stratosphere and limit the ability of global climate models to reproduce observed
 2   climates and variability, as well as our ability to predict future changes to the climate
 3   system. We propose a multi-pronged approach to reduce these uncertainties.

 4   IV.1.1.      Objectives

 5   The main objectives of this project are to address the following issues:

 6   IV.1.1.1.     What processes maintain the observed water vapor distribution in the
 7                tropical UTLS?

 8   Water vapor is controlled to a high degree by temperature, horizontal transport and ice
 9   microphysics (eg. Sherwood, 1996; Gettelman et al., 2000), but the local and global
10   effects of deep convection on UTLS water vapor are currently debated (eg. Sherwood and
11   Dessler, 2001). A wide spectrum of wave activity contributes to dynamical variability
12   and coupling with convection in the TTL, spanning diurnal, synoptic and intraseasonal
13   time scales (eg. Lau et al., 1991). Analyses of global data sets have identified convective
14   coupling in the tropics due to equatorially-trapped and inertia-gravity waves (eg. Wheeler
15   et al., 2000), together with Rossby-wave forcing from midlatitudes (Kiladis and
16   Weickman, 1992; Waugh and Polvani, 2000). A key focus of tropical UTLS studies will
17   be to quantify the inter-relationships among temperature, water vapor and ice clouds, and
18   to understand the processes that need to be simulated in large-scale models. A further
19   focus will be to understand the physical process(es) of dehydration near the cold tropical
20   tropopause (which regulates stratospheric humidity). Specifically, what are the relative
21   roles of convection, and large- and small-scale circulations in controlling the dehydration
22   process? Finally, recent studies have shown the importance of the NH summer monsoon
23   circulations and associated deep continental convection for influencing water vapor and
24   other constituents in the lowermost stratosphere (Randel et al., 2001; Park et al., 2003;
25   Fromm and Servranckx, 2003). Details of UTLS monsoon dynamics, and coupling with
26   water and convection, are ongoing research topics.

27   IV.1.1.2. What are the effects of cirrus and convective clouds on dehydration, and
28             their impact on the radiation budget?

29   Upper tropospheric clouds are crucial components of the earth’s energy and water
30   budget. They exert strong influences on the radiative balance and are also key elements
31   in determining the sources and distribution of water vapor in the stratosphere.

32   Measurements of the microphysical and radiative properties of UTLS cloud layers,
33   especially in the subtropical and tropical regions that are central sources and sinks of
34   stratospheric water vapor, are extremely limited. Many of the in-situ observations of
35   UTLS cloud layers are based on a few UT ice cloud measurements that were sampled in
36   the 1970’s (Heymsfield, 1986; McFarquhar et al., 2000). Recent observations of UT
37   cloud layers have been conducted during survey flights in the northern and southern

     NCAR UTLS White Paper DRAFT v1a                                                         10
 1   hemispheres during the European INCA (Inter-hemispheric differences in cirrus
 2   properties from anthropogenic emissions) project, but the measurements were primarily
 3   in cirrus that were well below the TTL cirrus layer heights. Recently, more
 4   comprehensive measurements have been made in convective outflow cirrus during the
 5   NASA CRYSTAL-FACE program over south Florida.

 6   In addition to cirrus clouds, deep convective clouds are important for transporting water
 7   substance and impacting the radiation budget. The radiative balance of the tropical UTLS
 8   controls the upward transport of air and constituents into the stratosphere, with strong
 9   feedbacks to the climate system. Deep convection in the troposphere and cirrus within the
10   TTL can significantly influence the local radiative balances and derived vertical
11   circulations. Recent work (Hartmann and Larson 2002) has raised several interesting
12   questions regarding the relationships between convection and the large-scale radiative
13   balance. Convective clouds may produce large positive or negative radiative feedbacks,
14   depending on whether they cover larger regions and cool the surface by radiating more
15   energy to space, or whether they supply more water vapor to clear sky regions. The
16   uncertainties are largest in understanding large cloud ensembles.

17   Our objective is to characterize and document the spatial distribution of UTLS water
18   substance, especially the portion that is in the condensed (ice) phase. A major goal is to
19   improve the understanding of processes that affect the temporal and spatial evolution of
20   UTLS ice cloud layers, whether produced directly or indirectly through deep convection
21   or through large-scale uplift. A secondary goal is to assess the role of anthropogenically-
22   produced aerosols and ice nuclei on UTLS cloud layer formation and dehydration. The
23   final goal will be to quantify the observed relationships between cirrus and convective
24   clouds and radiative fields (across a spectrum of scales), and accurately simulate the
25   observed behavior in models.

26   IV.1.2.      Proposed Studies

27   IV.1.2.1. Water Vapor and Temperature Distributions

28   We will attempt to understand the large-scale variability, and inter-relationships among
29   temperature, water vapor and clouds in the UTLS. This will be accomplished by a
30   combination of satellite data and in-situ measurements (existing and planned by the
31   community), combined with global models. Much of the current observational data for
32   the tropical UTLS region is derived from satellite measurements, and new satellite data
33   sets will soon offer increased spatial and spectral resolutions. Improved temperature data
34   sets include high vertical resolution measurements from GPS radio occultation (Randel et
35   al., 2003) and the Atmospheric Infrared Sounder (AIRS) on the AQUA satellite (since
36   middle 2002). New satellite water vapor data sets will include AIRS retrievals, plus
37   HIRDLS and MLS measurements from AURA (beginning early 2004). We will also use
38   the ongoing long-term record of stratospheric water vapor from HALOE (since 1991),
39   plus new measurements from SABER (since 2002). While past studies of UTLS water

     NCAR UTLS White Paper DRAFT v1a                                                         11
 1   vapor have relied on a few data sets with relatively low horizontal and vertical resolution,
 2   the near future will be data rich, and substantial progress should be realized. The global
 3   data sets will be used to quantify space-time variability and co-variability between
 4   temperature and water vapor fields, with focus on controlling processes. Statistical
 5   analyses and case studies will be used to isolate dynamical wave influences on
 6   temperature and water vapor. Nearly identical diagnostics can be applied to output from
 7   global chemistry and climate models for direct comparisons. Available aircraft
 8   observations from recent and planned tropical field campaigns (CRYSTAL-FACE, AVE-
 9   Tropics, CRYSTAL-TWP and TC4) will also be incorporated into the analyses. Two
10   regions of specific interest for more detailed process studies are over Indonesia during
11   NH winter (the coldest region with lowest water vapor), and over the NH summer
12   monsoons (cold, persistent high tropopause).

13   IV.1.2.2. Impact of clouds on radiation in the tropical UTLS

14   We also propose to investigate the radiation balance of the tropical UTLS region. This
15   work will be based on detailed column radiation codes and parameterized codes for
16   GCMs (such as CAM2) to explore the sensitivity of clear and cloudy sky heating rates to
17   variations in radiatively active gases, especially water vapor, ozone and carbon dioxide.
18   Recent analyses of the radiation budget in the tropical UTLS (Gettelman et al, in
19   preparation, 2003) indicates large uncertainties in the treatment of the infrared cooling
20   from water vapor in the region. Apart from this, there is generally good understanding of
21   the clear sky radiation budget. The radiation budget in cloudy skies is less certain.
22   Proposed investigations will incorporate deep convective clouds as well as examining the
23   radiative effects of thin cirrus clouds. We intend to tightly constrain these model studies
24   with available in-situ observations of clouds and the radiation budget from various field
25   programs (including CRYSTAL-FACE and observations from tropical ARM sites).

26   With better understanding of the effects of clouds on radiation in column models, we
27   further propose to use satellite observations (CERES instrument on AQUA, CALIPSO to
28   detail UTLS ice cloud layers with high spatial resolution) and models to investigate cloud
29   radiative feedbacks. We hope to test the ability of global models at NCAR (such as
30   CAM2) to properly represent the radiation balance of the tropical UTLS. This will enable
31   better understanding of the entry of air into the stratosphere, and to better constrain the
32   climate sensitivity of global models.

33   IV.1.2.3. Impact of waves on the water vapor budget

34   A broad spectrum of wave activity occurs in the tropics, and studies of coherent wave
35   variability offer an opportunity to quantify coupling mechanisms (eg. Wheeler et al.,
36   2002). We also propose a multi-scale investigation of the effects of waves on the water
37   vapor budget. We will study large-scale waves in the tropics (such as equatorially trapped
38   Kelvin waves) using satellite observations and global models. Randel et al (2003) have
39   recently used GPS data to examine the coherence between deep convection and

     NCAR UTLS White Paper DRAFT v1a                                                          12
 1   temperature variability in the tropics, and we will link this to newly available water vapor
 2   observations.

 3   Radiosonde observations and GPS radio occultation measurements both indicate
 4   enhanced small-scale wave activity in the tropical UTLS region, probably associated with
 5   inertia gravity waves (Alexander et al., 2002). These waves have large amplitudes (~2-3
 6   K) and periods up to several days, and can significantly influence water and ice cloud
 7   behavior. Figure 4 is an example of the wave structure observed in the GPS data.
 8   Gettelman et al (2002) have noted the potential for small scale waves to have an
 9   important impact on dehydration of air in the Tropical Tropopause Layer. We will study
10   small-scale gravity waves and their effects in the UTLS using available observations of
11   water vapor, temperature and winds from aircraft platforms and high resolution models.
12   We will also investigate the water budget within and below gravity waves initiated by
13   tropical or subtropical convection, using a combination of measurements and cloud
14   modeling.

15   The modeling effort will draw upon simulations of gravity waves by Lane et al. (2001) to
16   define the vertical velocity, phase, and amplitude of gravity waves forming under a
17   variety of conditions. Detailed cloud models will be adapted to study the microphysical
18   and radiative properties and water budgets of cirrus clouds formed by gravity waves and
19   by large scale ascent in the Tropical UT. The effort will focus on how particles affect
20   cloud formation and evolution, as well as the impact of chemical interactions and
21   anthropogenic emissions on particles and cloud evolution.

22   IV.1.2.4. Role of cirrus in dehydration

23   The processes by which particles in anvil and thin cirrus clouds nucleate and evolve may
24   be central in the dehydration process in the UTLS. The small particles are thought to
25   originate by homogenous freezing (without the need for ice nuclei) of cloud droplets at
26   temperatures below –34C, although it is not known how frequently this occurs.
27   Measurements in updrafts near the homogenous freezing level in continental convective
28   updrafts will be compared with similar measurements made over oceanic storms. In
29   particular, we will focus on the advection and evolution of small particles.

30   To better understand the regional and global effects of these clouds, satellite observations
31   of thin cirrus in the tropics (ongoing measurements from HALOE and SAGE II) and new
32   data (e.g. MODIS and AIRS beginning in 2002, HIRDLS in 2004, CALIPSO in 2005)
33   will be analyzed in relation to deep convection (as indicated by fields of outgoing long-
34   wave radiation) and large-scale meteorological fields. Temporal variability of the cirrus
35   fields will be analyzed in conjunction with temperature and water vapor fields, to
36   understand formation and maintenance mechanisms. Cirrus particle sizes will be derived
37   from an analysis of HIRDLS extinction data, and later, from analysis of CALIPSO (lidar
38   in space) data.

     NCAR UTLS White Paper DRAFT v1a                                                          13
 1   We also propose a detailed intercomparison of cirrus cloud models with observations,
 2   with the aim of improving and validating the parameterization schemes used in GCM’s,
 3   specifically CAM2 and WACCM2. We will use data from field campaigns (CRYSTAL-
 4   FACE and upcoming follow-on missions) with MODIS, AIRS and HIRDLS data to form
 5   an operational set that can be used to test separate radiation codes and cirrus cloud
 6   models. These models include detailed treatments of microphysics, and several
 7   parameterized approaches for GCM’s. The expected result is a better understanding of
 8   the tradeoffs between complexity and accuracy in GCM microphysics schemes in the
 9   UTLS. This will hopefully improve the validity of the UTLS climate for studying climate
10   changes and changes in stratospheric water vapor. The project will be a collaboration
11   between MMM, ACD and CGD, along with several outside collaborators.

12   IV.1.3.      Airborne Experiments

13   Additional aircraft observations in the tropical UTLS would be valuable to better
14   characterize the properties of UTLS ice cloud layers and their role in UT water vapor
15   dehydration. We would seek to supplement the satellite observations, particularly to
16   examine the water budgets of ice cloud layers formed through gravity waves, and the
17   evolution of ice in anvil and in-situ generated cirrus cloud layers. Size distributions in
18   deep convective anvils and thin cirrus in particular would be valuable to help validate
19   models, and cannot be derived from satellites.

20   To acquire this data, we propose to coordinate with other planned efforts in the tropics.
21   Although HIAPER does not reach the altitude of the tropical tropopause, it has the
22   capability to sample the level of mean convective outflow. We envisage that a planned
23   campaign to the tropics would include HIAPER along with a high altitude aircraft. There
24   are several opportunities for such collaborations with NASA aircraft in 2005 through
25   2007

26   IV.2. Extratropical Stratosphere-Troposphere Exchange

27   Despite the progress of recent years, large uncertainties remain in the quantitative
28   diagnoses of STE. Due to the large gradient in ozone concentration across the
29   extratropical tropopause and the long lifetime of ozone in the UTLS region, the
30   contribution of stratosphere to troposphere transport (STT) to upper troposphere ozone is
31   significant. The large uncertainty of STE across the extratropical tropopause is a key
32   factor in the closure of the UT ozone budget. Similar issues exist for the lower
33   stratospheric water vapor budget and trend due to the uncertainty of troposphere to
34   stratosphere transport (TST). TST affects global change by transporting water vapor,
35   biomass burning products and other anthropogenic emissions into the stratosphere.

36   IV.2.1. Objectives

37   There are four major objectives for this project:

     NCAR UTLS White Paper DRAFT v1a                                                        14
 1   IV.2.1.1. To establish an improved STE climatology with emphasis on validating the
 2             results from idealized models using observations

 3   Recent studies of STE climatology using Lagrangian approaches (Seo and Bowman
 4   2002; Stohl et al., 2003; Wernli and Bourqui, 2002; Sprenger and Wernli, 2003; James et
 5   al., 2003) have provided much improved information on the seasonality as well as the
 6   spatial distribution of STE activities. However, results from these idealized model studies
 7   need to be verified by observations. For example, initial results from the European
 8   STACCATO project show that, although many of the model results are qualitatively
 9   consistent with each other, large discrepancies exist quantitatively (e.g. Stohl et al.,
10   2003). Validation by observations is a key step to help establish reliable methods of
11   characterizing and quantifying STT and TST.

12   In addition to the seasonality of exchange activities, it has been shown using trajectory
13   model studies that anthropogenic emissions near the eastern seaboard of North America
14   and Asia are more likely to be transported to the upper troposphere and lowermost
15   stratosphere due to the distribution of the starting point of warm conveyor belts (Stohl,
16   2001). These results also need to be verified by observations. How in general the
17   troposphere airstreams, including warm conveyor belts, cold conveyor belts and dry
18   intrusions, impact the preferred exchange locations needs to be investigated using
19   coupled global modeling and observations.

20   IV.2.1.2. To examine the definitions of the extratropical tropopause and to
21             characterize the transition region, especially in the vicinity of the
22             subtropical jet

23   Uncertainty in the transport boundary is a major contributor to the uncertainty in
24   quantifying STE. Should the extratropical tropopause be characterized as a surface, or a
25   transition layer? Analyses of tracer relationships derived from aircraft measurements
26   (Fischer et al., 2000, Zahn et al., 2000, Hoor et al., 2002; Pan et al., 2003) suggest the
27   existence of a transition layer in the extratropical UTLS region where maximum STE
28   occurs (somewhat analogous to the transition across the TTL, or to the polar vortex
29   boundary). What controls the thickness of the transition layer? How do we locate this
30   layer? These questions need to be answered with multi-scale analyses using observations.

31   IV.2.1.3. To better characterize the spatial extent and the seasonal variability of
32             mixing between stratospheric and tropospheric air, to investigate how
33             multiple scale processes are coupled to induce mixing in the tropopause
34             region, and to investigate the effect of gravity wave breaking in particular

35   Mixing is an important part of irreversible STE. Yet when, where and how mixing occurs
36   between the stratosphere and troposphere is poorly understood. The processes that cause
37   mixing are not well represented in the models. This is a multi-scale problem, with large-
38   scale mixing occurring on spatial scales of 10-1000 km, in which filaments of

     NCAR UTLS White Paper DRAFT v1a                                                         15
 1   stratospheric air are sheared away from upper level troughs and stirred into the
 2   troposphere. Recent high-resolution radiosonde data (e.g. Birner et al., 2002), as well as
 3   aircraft measurements in the vicinity of the tropopause (e.g. Pavelin et al, 2002), have
 4   documented the role of small-scale mixing by turbulence in the vicinity of the
 5   tropopause. Turbulence in this stably-stratified region is characterized by intermittency
 6   and is associated with a gradient Richardson number of 1/4 or less. The small
 7   Richardson numbers are primarily the result of wind shear generated by gravity waves, or
 8   associated with the jet stream or deep convection in the troposphere. The tropopause
 9   structure is somewhat analogous to that occurring at the top of the planetary boundary
10   layer (PBL) where turbulence generated in the PBL by shear and convection engulfs free
11   tropospheric (FT) air and sharpens the interface between the PBL and the FT. Birner et
12   al. (2002) have documented similar structure at the mid-latitude tropopause. The
13   objective is to investigate how these processes are coupled with multiple scale
14   observations and models.

15   IV.2.1.4. To investigate the mechanisms that produce deep intrusion into the
16             stratosphere, especially the impact of deep convection

17   Although it is recognized that extratropical TST can be due to both adiabatic and diabatic
18   processes, isentropic transport associated with large-scale Rossby waves has been
19   considered the main contributor (e.g., Holton et al., 1995, Bradshaw et al., 2002 a,b).
20   Recent observations (Fromm and Servranckx, 2003) and model studies (e.g. Wang 2003)
21   have suggested that mid-latitude summertime convection can inject significant amounts
22   of water vapor and boreal fire smoke particles deep into the lower stratosphere. In some
23   cases, the transport is via a combination of vertical lifting due to penetrating convective
24   updrafts, nonlinear cloud-top instabilities, and gravity wave breaking. Gravity wave
25   breaking above deep convection may play an important role in creating turbulent mixing
26   and consequent irreversibility. The relative importance of these transport mechanisms
27   needs to be better characterized and quantified using models and observations. How often
28   does the deep intrusion occur quasi-laterally? To what extent does deep convection pump
29   boundary layer material, including boreal fire products and aerosols, from the lower
30   troposphere into the UTLS? How important is midlatitude deep convection to TST?
31   What is the contribution to the water vapor budget in the stratosphere?
32   IV.2.2.      Proposed Studies

33   Global satellite measurements of multiple tracers with the upcoming AURA satellite,
34   HIRDLS in particular, together with the high altitude capability of HIAPER, present an
35   unprecedented opportunity to create an improved STE climatology. The collaboration
36   between investigators working with global scale and mesoscale models presents a new
37   possibility of characterizing mixing processes and their controlling mechanisms.

38   IV.2.2.1. UTLS climatology of chemical tracers and their relationship

39   UTLS climatologies of chemical tracers, including O3, H2O and CO, can be compiled
40   from the upcoming AURA data, HIRDLS and TES data in particular. The temporal

     NCAR UTLS White Paper DRAFT v1a                                                         16
 1   variation in the seasonal scale and the spatial variations in the extratropics will provide
 2   important information for establishing and validating the STE climatology. Climatology
 3   of tracer relationships can further provide unprecedented information that can help
 4   validate the modeled STE climatologies. Comparison of satellite measurements and in
 5   situ airborne measurements will provide a connection between the tracers and tracer
 6   relationships observed on different scales.

 7   IV.2.2.2. Mixing in the extratropical tropopause region, location, and the
 8             variability of the transition layer

 9   The occurrence of mixing in the tropopause region and its effect on chemical trace gas
10   transport can be quantified using tracer relationships, for tracers with tropospheric or
11   stratospheric origin (eg. CO and ozone). An example is given in Figure 5.

12   Case studies such as this one will provide insight into the observed mixing layer, its
13   location, its transport effect and the associated event. Statistical studies using airborne
14   and spaceborne observations will be used to create a climatology of the transition layer,
15   including its latitudinal distribution and seasonal variability. Satellite constituent
16   measurements (eg., HIRDLS) will provide the large-scale perspective and allow
17   interpretation of model results for global behavior.

18   Such detailed observations for individual case studies will also provide information for
19   parallel model studies. We plan to use models of different scale to complement the
20   aircraft observations. The models planned to be involved are MOZART3, a global CTM,
21   the mesoscale WRF model, a cloud-scale model configured to examine gravity waves
22   (e.g., the Clark-Hall model, Clark et al. 1996), and Lagrangian chemical transport model
23   (ClaMS). They can be used together to study the mechanisms that control mixing and
24   produce irreversible exchange.

25   IV.2.2.3. Investigate turbulent mixing due to gravity wave breaking

26   The role of inertia-gravity waves (IGWs) in the UTLS region has been investigated by
27   O'Sullivan and Dunkerton (1995), who simulated their generation during the lifecycle of
28   a growing baroclinic wave using a GCM. The IGWs propagate into the lower
29   stratosphere, break near a critical level, and contribute to cross-isentropic mixing where
30   they break. O'Sullivan and Dunkerton speculated that, even when the waves do not break,
31   they will be strained rapidly to smaller scales due to the background flow, leading to
32   irreversible isentropic transport. Current computing capabilities allow robust
33   representations of the gravity waves generated by such systems (e.g., the recent results
34   from Zhang, 2003) and make a quantitative investigation of the contribution of IGWs to
35   mixing possible.

36   Deep convective clouds, whose tops reach the upper troposphere, are capable of
37   generating large amplitude gravity waves. These waves are capable of transporting air

     NCAR UTLS White Paper DRAFT v1a                                                         17
 1   masses vertically through laminar lifting (e.g., Wang 2003), which may often be
 2   irreversible. Also, in some conditions, gravity waves above convection become unstable
 3   and break down, inducing deep layers of turbulent mixing of momentum and constituent
 4   species. Figure 6 gives an example from recent high resolution numerical modeling
 5   studies of Lane et al. (2003), which indicate that gravity waves generated by deep
 6   midlatitude convection induced a turbulent layer that was about 3 km deep, directly
 7   above the tropopause and the cloud top. Lane et al. attributed the cause of the wave
 8   breakdown to interactions between the waves and a wind-shear induced critical level.
 9   However, the processes causing the breaking and the details of the turbulence are far
10   from properly understood. Further research is necessary to understand the conditions that
11   are conducive to wave breaking, and to characterize the turbulence and mixing associated
12   with the breakdown.

13   Both mountain waves and waves generated by jet/front systems are also capable of
14   breaking down and causing turbulent mixing (e.g., Ralph 1997). The waves generated by
15   convection, mountains, and fronts possess different spectral characteristics and therefore
16   the flow conditions that control their breakdown, are different. Also, because of the
17   different horizontal scales of the forcing, the turbulent dissipation and mixing may
18   possess different properties and relative contributions to mixing and STE in the UTLS. It
19   will be important to determine the processes that control the breakdown of gravity waves
20   generated by convection, mountains, and jet/front systems respectively, and to investigate
21   the character of the turbulence and mixing generated during the breakdown.
22   It is important to have case studies using aircraft missions guided by mesoscale model
23   forecasts, with further analysis with multi-scale models. Analyses of the case study
24   results can give insight to the utility of the satellite data in producing global scale
25   statistics.

26   IV.2.2.4. Deep intrusion into the troposphere

27   A recent Lagrangian-based 15-year climatology of cross-tropopause exchange shows that
28   STT occurs predominantly over the Pacific and Atlantic storm track regions during
29   winter, spring and autumn, and also over the Mediterranean in winter and spring.

30   Studies have shown that stratospheric intrusions are diluted by turbulent mixing (Shapiro,
31   1980) or by large scale stirring processes that shear them into streamers or filaments
32   (Appenzeller et al., 1996; Bithell et al., 2000; Cooper et al., 2001). Cooper et al. (2003)
33   using a Lagrangian transport model have recently shown how a deep stratospheric
34   intrusion decays over several days and experiences large-scale mixing with adjacent
35   polluted warm conveyor belts. To date, no study has provided clear observations of the
36   long term evolution of a deep intrusion, its decay and dilution and the impact of its
37   remnants on surface ozone mixing ratios. Furthermore, it is unlikely that Eulerian or
38   Lagrangian models realistically simulate the small-scale mixing processes that dilute
39   deep STT events.

40   To clearly understand the life-cycle of a deep stratospheric intrusion and the mixing

     NCAR UTLS White Paper DRAFT v1a                                                         18
 1   processes that irreversibly entrain it into the troposphere, a carefully planned aircraft-
 2   based experiment is required. A high resolution Lagrangian transport model such as
 3   FLEXPART (Stohl et al., 2003) can be used to forecast the location of a stratospheric
 4   intrusion and the filaments that shear off over several days. A particular region of the
 5   intrusion can be targeted by the HIAPER aircraft at an early stage of its life-cycle. Once
 6   inside the intrusion the aircraft will release an inert chemical tracer that will be sampled
 7   over the next several days by the HIAPER and C-130 aircraft as the intrusion decays and
 8   descends into the atmospheric boundary layer. In addition to exploring large-scale
 9   stirring and turbulent mixing this study will also explore the impact of convection on the
10   erosion of the intrusion. A recent study by Griffiths et al. (2000) has shown that the
11   potential vorticity inversion associated with a stratospheric intrusion can trigger
12   convection, but to date the impact of this process on trace gas re-distribution has received
13   little attention. Concentrations of the inert tracer will reveal the rate and spatial extent of
14   the dilution processes, with particular attention given to the impact the event has on
15   surface ozone mixing ratios. In addition, instruments on-board the AURA satellite will
16   help to track the spatial extent of the intrusion and its break-away features, and data
17   collected over several years can be used to verify the deep stratospheric intrusion
18   climatology developed by Sprenger and Wernli (2003).
20   IV.2.2.5. Deep intrusion into the stratosphere including deep convection, especially
21             the transport of Boreal fire product into UTLS

22   Two types of processes have been identified to produce deep intrusion into the
23   stratosphere: large scale Rossby wave breaking and deep convection at the mid latitudes.
24   The processes from these two mechanisms have very different effect on stratospheric
25   composition. To establish the relative importance of the two mechanisms, measurements
26   of tracers in the lowermost stratosphere that have different lifetimes, combined with
27   satellite and modeling data analyses will be effective. A new generation of satellite data
28   will be used to investigate the statistical occurrence of these events and their impact.

29   New evidence is mounting that the contribution of mid-latitude deep convection to TST
30   cannot be neglected (Fromm and Servranckx, 2003; Wang, 2003). In particular, Fromm
31   et al. (2000, 2003) have shown that large values of extinction, at temperatures above that
32   of the ice frost point, are observed in the lower stratosphere during spring, summer, and
33   fall, and that this aerosol is likely due to intense Boreal fires in Siberia and Canada. There
34   are approximately 30,000 Boreal fires per year. Sometimes, the fires last over a four
35   months period. Figure 7 highlights the issue using TOMS optical depth data. The most
36   intense fires can generate their own particular dynamics, which needs to be studied. It is
37   important to quantify the frequency of occurrence and the physical conditions by which
38   the smoke is injected into the UTLS.

39   We propose to combine airborne and satellite data sets, in addition to modeling efforts, to
40   produce a coherent picture of how Boreal smoke is transported from the surface into the
41   UTLS. Satellite data from various platforms will provide the following information:

     NCAR UTLS White Paper DRAFT v1a                                                             19
 1   MODIS (and/or TRMM) fire count data, plus true color MODIS images will reveal the
 2   surface sources of the Boreal fires. TOMS aerosol index and MODIS aerosol optical
 3   depth data will specify the geographical location of smoke that is in the troposphere.
 4   MOPITT observations will indicate at several pressure regions in the middle troposphere
 5   enhanced mixing ratios of CO. HIRDLS and several solar occultation experiments (e.g.
 6   POAM, SAGE, ILAS, HALOE) will observe enhanced aerosol extinction in the upper
 7   troposphere and lower stratosphere. MLS and TES measures of HCN, CH3CN will also
 8   provide good biomass burning signatures. Airborne measurements of smoke in the UTLS
 9   will quantify the particle size distribution of the smoke, and chemical analysis of the
10   particles will identify the composition of the smoke.

11   Modeling efforts, ideally using a convective cloud model, nested into a global scale
12   model (such as MOZART), will follow the smoke from the surface into the UTLS. Of
13   particular interest is the prediction of the evolution of the smoke particles (position and
14   particle size) and that of other gases that are transported into the UTLS (CO, NO2, HCN,
15   NH3, etc). The model calculations will quantify the influence upon UTLS chemistry, e.g.
16   the perturbations of photolysis rates and changes in O3 concentrations.


18   IV.2.3.      Airborne Experiment

19   An airborne experiment using HIAPER and the C-130 is being planned to investigate the
20   extratropical tropopause region, with a focus on the region near the subtropical jet.

21   Objectives:

22   There are several objectives for the planned airborne experiment: (1) to examine mixing
23   away and in the vicinity of the subtropical jet using tracer relationships and how it is
24   associated with the thermal and dynamical structure of the tropopause region; (2) to
25   investigate turbulent mixing as shown in the tracers and its connection to gravity wave
26   breaking, (3) to investigate the contribution of tropospheric airstreams to transport into
27   the lowermost stratosphere; (4) to investigate the effect of deep convection; and (5) to
28   contribute to AURA validation.

29   Experiment strategy:

30   The experiment strategy will be (1) to use the C-130 as a remote sensing platform with
31   uplooking LIDAR to map out the tropopause region. Use HIAPER as the in situ platform
32   to sample the tropopause region; (2) to establish partnership with AURA validation
33   experiment (AVE) and team up with the WB57 in planning the flights; (3) Work with
34   HIRDLS team to coordinate the flights with AURA orbit overpass. For the well planned
35   events, HIRDLS can turn on a special zoom mode to focus on the HIAPER flight path.
36   (4) Since the events that contribute to the STE in the extratropics are episodic, model

     NCAR UTLS White Paper DRAFT v1a                                                          20
 1   guided flights will potentially be more effective. Use the WRF model to predict
 2   mesoscale events of interest. Using Lagrangian models to predict airstreams for air
 3   masses of particular interest.

 4   Measurement priority:

 5   In addition to the standard suite of state parameters, the following measurements are
 6   required:

 7       •   In situ measurement of tracers, O3, CO, CO2, H2O, CH4, N2O. Additional tracers
 8           desirable include HCl, Beryllium isotopes, organic halogens and methylnitrate.

 9       •   Remote sensing LIDAR ozone and aerosol and water vapor, although in the case
10           the detection limit needs to be improved.

11       •   High resolution temperature and wind (MTP like instrument and dropsondes,
12           currently existing wind profile LIDARs are of limited utility), to identify the onset
13           of gravity wave and the breaking events.

14   Comparisons of measurements from redundant instruments are an important part of the
15   initial experiments.

16   IV.3. Ozone and Radical Budget Chemistry in the UTLS

17   The chemical composition, including water, of the UT determines the boundary condition
18   for air entering the tropical stratosphere via the Hadley circulation. In the extratropics the
19   composition is determined in part by two-way exchange across the tropopause and input
20   from deep convection or synoptic scale disturbances. Understanding the photochemical
21   processing of this variable composition throughout the global UTLS region is required
22   for model improvement and for their prediction of trends in ozone and other constituents
23   in both the troposphere and stratosphere. Changes in ozone ultimately impact the earth’s
24   radiation budget, the UV flux to the surface, and the production of radical species (e.g.,
25   OH) that are responsible for removal of primary pollutants.

26   The tropopause region is characterized by strong gradients in the mixing ratio of
27   numerous chemical and aerosol constituents and represents a transition between two
28   different photochemical regimes. In the UT, chemistry involving HOx and NOx radicals
29   in the presence of CO and CH4 and a possible wide variety of volatile organic carbon
30   compounds (VOCs) usually results in net ozone production rates, while in the lower
31   stratosphere, catalytic cycles involving HOx, NOx, ClOx, BrOx, and perhaps IOx radicals
32   result in net ozone depletion rates. Thus the overall chemical processing in models of the
33   UTLS region is particularly sensitive to identifying all of the chemical constituents and
34   processes that influence radical production and loss rates.

     NCAR UTLS White Paper DRAFT v1a                                                            21
 1   With the high-altitude HIAPER aircraft and the launching of HIRDLS, TES and other
 2   satellite-borne instruments, measurements of chemical species spanning the UTLS region
 3   will yield important data to test our current understanding of transformation mechanisms
 4   within and between chemical families. These measurements coupled with modeling
 5   studies will reduce major gaps in our understanding.

 6   IV.3.1.      Objectives

 7   The overall objective of this project is to investigate the processes controlling the budgets
 8   of radicals and ozone in the tropopause region, particularly emphasizing the impacts of
 9   perturbations on chemical composition due to transport by deep convection, frontal
10   lifting, and two-way exchange across the tropopause.

11   IV.3.1.1.     Determine the contribution of convective processing on radical (HOx, ROx,
12                NOx, ClOX, BrOX,…) abundance and radical precursor concentrations and
13                their impact on O3 production and loss rates in the UTLS

14   Deep convection plays a vital role in the transport of species from the lower troposphere
15   to the upper troposphere, in the uptake and rainout of species, with aqueous and ice
16   chemistry, and in its connection with lightning. Figure 8 highlights the effect of
17   electrically active convection to NOx. The transport of chemical constituents into the
18   UTLS via convection is not necessarily similar to the transport of heat or water. Thus,
19   the mass flux of key species through detailed measurement and modeling studies for a
20   variety of convective systems must be done so that global-scale models can reliably
21   represent the transport of species. A better understanding of what species survive or are
22   produced in the convective process and their fate after deposition into the UTLS is
23   needed.

24   Although there have been many studies examining the transport of passive tracers, e.g.
25   CO, CO2, and O3, in convection, better understanding of the processing of soluble and
26   reactive species, e.g. peroxides, aldehydes, and VOCs, is required (e.g., Prather and
27   Jacob, 1997). These soluble, reactive species may be transformed to other species in the
28   aqueous or ice phase and/or rained out of the storm via precipitation. Recent studies
29   (Crutzen and Lawrence, 2000; Barth et al., 2001) have shown the importance of the
30   formation of precipitation to gases with a spectrum of solubilities. In- situ measurements
31   and detailed model studies need to be continued for a variety of convection such that the
32   results can be generalized and implemented in global-scale models. Process studies will
33   require constituent measurements from the surface layer to the LS including the anvil
34   outflow.

35   Both the total production of reactive nitrogen by lightning activity and its distribution
36   with altitude are quite uncertain. Global estimates range from 2-20 Tg(N)/yr versus, for
37   example, that from the major global near-surface source fossil fuel combustion of 25-35
38   Tg(N)/yr. On the scale of storms, lightning production has been observed to increase

     NCAR UTLS White Paper DRAFT v1a                                                           22
 1   NOx in the UT over background conditions by an order of magnitude or more. With the
 2   ~5 day lifetime of NOx in the UT the impact on ozone production and HOx partitioning
 3   can cover much larger downwind scales. In current global chemistry transport models
 4   deep convection and lightning production are described by sub-grid parameterizations
 5   and the lightning production/distribution can be essentially a “tuning knob” in
 6   simulations. Further in-situ and remote (ground-based and satellite) observations of the
 7   amount and vertical distribution of NOx from a variety of storms are required for model
 8   refinement.

 9   Halogen species (including halogen radicals and halocarbons) are critical to controlling
10   the abundance of ozone in the LS. A major source region for these species (e.g., CH3I,
11   CH3Br) is the ocean, and therefore the transport of the halogen species from its oceanic
12   source to the UTLS must be quantified. Whether cirrus particles can activate halogen
13   radicals requires further investigation. In particular, determining the contribution of
14   convective transport compared to isentropic transport of halogen species should be
15   determined.

16   IV.3.1.2.     Reconcile the HOx abundance with HOx precursors in air masses in the
17                unperturbed UT, and air masses perturbed by high NOx and high solar
18                zenith angles

19   There has been a general tendency for photochemical box models to underestimate HOx
20   levels in the UT not recently influenced by convective input, a discrepancy that has been
21   linked to the presence of numerous oxygenated VOCs and peroxides (Brune et al. 1998;
22   Faloona et al., 2000). The photolysis products of gases such as acetone, CH2O, CH3OOH,
23   H2O2, etc., provide important sources of HOx radicals. Figure 9 highlights these issues
24   (Wennberg et al., 1998). Discrepancies in the HOx budget, varying in both sign and
25   magnitude, have also been observed at high NOx, in the presence of cirrus clouds, at high
26   solar zenith angles, and in convective outflow (Jaeglé et al., 2000). In contrast, the
27   abundance and partitioning of HOx in the unperturbed LS is reasonably well documented
28   and is in basic agreement with modeling exercises (Lanzendorf et al., 2001). Thus in situ
29   investigations contrasting the HOx abundance and partitioning across the tropopause and
30   through sunrise and sunset would provide important tests of both instrument and model
31   capability. It is also likely that all of the precursor VOC’s have not been identified and/or
32   quantified, especially in cases of convective outflow influence. Studies designed to
33   investigate Objective 1 would provide tests of the radical behavior under the conditions
34   of high NOx produced in outflow from thunderstorms.

35   IV.3.1.3. Determine the effect of multiphase and heterogeneous chemistry on the
36             abundance of radicals, O3, and their precursors

37   Cirrus from both convective cloud outflow and formation by large-scale lifting or gravity
38   waves can play a role in UT chemistry via heterogeneous chemistry and by altering the
39   actinic flux. Such processes are only beginning to be investigated experimentally. Is ice

     NCAR UTLS White Paper DRAFT v1a                                                           23
 1   particle formation/evaporation affected by local gaseous uptake including organic
 2   constituents or by incorporation at lower altitudes into the liquid phase? Do cirrus
 3   particles affect radical loss rates? The uptake of HNO3 and possibly other species onto ice
 4   crystals and any chemical reactions occurring on or in the crystals must be quantified to
 5   assess the role cirrus have in the redistribution of chemical species as larger crystals fall.
 6   For example, Lawrence and Crutzen (1998) suggested via global model simulations that
 7   settling of soluble trace gases in ice resulted in a notable redistribution of these species.

 8   Studies investigating the role of the soluble gas on ice formation must also be pursued.
 9   Recent evidence has shown that the concentration of the highly soluble gas HNO3
10   contributes to the rate of formation of ice (Hienola et al., 2003). The importance of these
11   processes needs to be further assessed along with their impact on the regional and global
12   scales.

13   IV.3.1.4.    Reduce the uncertainty of NOx and HOx chemistry in the UT

14   The coupling between the HOx and NOx families dominates O3 production rates. In the
15   remote upper troposphere recent studies have indicated that cycling between NOx and its
16   principal reservoir, HNO3 is poorly understood: photochemical box models consistently
17   underestimate NOx/HNO3 observations, and this has lead to speculation regarding the
18   importance of various heterogeneous reactions that may convert NOy species back to
19   NOx. The difference is less but remains significant in some global-scale models where
20   “fresh” source advection is automatically considered (Staudt et al., 2002). Furthermore,
21   knowledge of the photolysis rate constants and kinetic rate constants at low temperatures
22   and pressures representative of the UTLS has generally been determined by extrapolating
23   data taken from laboratory measurements at high temperatures and pressures. Gaining
24   accurate kinetic data and exploring previously undiscovered reactions that are favored in
25   UTLS conditions could reduce the uncertainty of NOx and HOx chemistry. Measurements
26   of HNO4 which can be a nighttime source of HOx radicals (similar to NO3 formation at
27   night) and HONO are needed to compare with model expectations. Do halogen radicals
28   (e.g. BrO) significantly perturb the partitioning of NOx in the UT? How do short-lived
29   VOCs transported rapidly in convective disturbances affect the production of organic
30   nitrates (e.g., PAN or other alkyl and peroxynitrates) and the sequestration of active NOx?

31   IV.3.2.      Proposed Studies

32   IV.3.2.1. Contribution of convection on distributions of ozone and free radicals in
33             the UT

34   Activities are being pursued to examine the effect of convection on chemistry on both the
35   mesoscale and global scale. On the mesoscale analysis and modeling of STERAO deep
36   convective storms is being conducted, while on the global scale comparison of MOZART
37   results to satellite data is being accomplished (see Figure 10). However to verify the
38   modeling and satellite analysis, an aircraft field campaign needs to be conducted to

     NCAR UTLS White Paper DRAFT v1a                                                            24
 1   examine the effect of convection on chemistry in the upper troposphere.

 2   Much analysis has already been accomplished for the 10 July 1996 STERAO storm and
 3   future work will aim toward analyzing the 12 July storm. These two storms differ in their
 4   evolution and lightning activity, height of the convective outflow, and location of some
 5   of the aircraft measurements relative to the storm. These contrasting features will enable
 6   us to determine how the distribution of soluble, reactive species and passive tracers differ
 7   from one storm to another via measurement analysis and cloud-scale modeling.

 8   We will attempt to answer the following questions when analyzing the cloud chemistry
 9   mesoscale simulations of the 12 July 1996 STERAO storm.

10       •   How well do results from the convective cloud model compare to observations of
11           passive tracers? This analysis will give credence to the transport simulation.

12       •   How well do results from the convective cloud model compare to observations of
13           soluble reactive species? One particular comparison to be done will be the
14           formaldehyde concentrations. Formaldehyde is soluble and is reactive in the
15           aqueous phase. Thus, comparisons of model results and observations will
16           illuminate how well we understand the cloud processing of reactive, soluble
17           tracers.

18       •   How do model results and observations of chemical species from the 12 July
19           storm contrast with those from the 10 July storm? Can we begin to generalize the
20           effect of convection on chemical species from one type of thunderstorm to the
21           next?

22       •   Can we estimate the flux of HOx precursors into the upper troposphere?

23   Satellite observations of deep convection, chemical species concentration, and lightning
24   flashes used in combination with global model results can provide an analysis of the
25   contribution of convection to some species in the UTLS. Although the satellite data is
26   available from a particular time of day (fly over time of the satellite) and the contribution
27   of convection from a full day cannot be estimated, these analyses will provide an
28   indication of what the magnitude of the convective flux is to species abundances in the
29   UTLS. Another limitation of the satellite data to address is retrieving concentrations of
30   the chemical species in the vicinity of clouds. Nevertheless, convective outflow may still
31   be detected after the anvil has evaporated.

32   As part of the analysis effort of MOPITT and HIRDLS satellite data, studies are planned
33   to examine the contribution of the abundance of various species from convection. These
34   studies will need to be performed in conjunction with aircraft measurements in outflow
35   regions so that the satellite data can be validated. Once the validity of the satellite data is
36   established, subsequent studies analyzing satellite data and global model results can

     NCAR UTLS White Paper DRAFT v1a                                                             25
 1   reveal on a global scale the role convection has on the concentration of key species in the
 2   UT.

 3   Lightning production of NOx measured in Florida thunderstorms during the NASA
 4   CRYSTAL FACE experiment is being compared with the parameterization used in the
 5   MOZART model.

 6   IV.3.2.2. Effect of Multiphase and Heterogeneous Chemistry on Chemical
 7             Composition of the UT

 8   During the 2002 NASA CRYSTAL-FACE program, measurements of the uptake of NOy
 9   constituents on ice crystals in both convectively-formed cirrus and cirrus formed by
10   large-scale lifting were obtained. These data are being compared with expectations based
11   on laboratory studies. Analysis of the ACD NOy and NO measurements along with the
12   HNO3 measurements made by scientists at the NOAA Aeronomy Laboratory will
13   quantify the uptake of NOy constituents other than HNO3.

14   Modeling studies are also planned to investigate the importance of HNO3 uptake on ice,
15   of ice-phase chemistry, and of aqueous-phase chemistry. These studies will be
16   accomplished as part of the convective cloud chemistry model simulations and with
17   large-scale chemistry transport models (either WRFchem or MOZART) to assess the
18   effect of ice crystal sedimentation on HNO3 redistribution.

19   IV.3.3.      Airborne Experiments

20   In-situ observations using HIAPER and other aircrafts are vital to improving our
21   understanding of chemical and aerosol processing in the UTLS. Satellite instruments
22   cannot yet measure key constituents. They will also contribute to satellite instrument
23   validation. A study of deep convection will require comprehensive sampling from the
24   boundary layer to the LS thus likely necessitating simultaneous deployment of HIAPER
25   and the C-130 or cooperation with other agencies in planned experiments like the NASA
26   AVE or TC4 projects. Short-lived constituents measurements will need to be
27   complemented by measurement of a variety of longer-lived tracers to assess air mass
28   origins and dilution/mixing processes.

29   There are a number of possible projects centered on HIAPER capabilities, for example,
30   projects focused on halogen radicals, HOx-NOx-VOC photochemistry, heterogeneous
31   studies of cirrus outflow, or investigations of polar stratospheric clouds. In practice
32   projects will likely be a combination of microphysics, convection, STE, and
33   photochemistry. A rate-limiting step to a specific project will be the modification and/or
34   new development of the large suite of instruments that will be required. Flights for tests
35   and evaluation of the instrumentation will be required.

36   The following is a synopsis of one particular project that is in the planning stage. Further

     NCAR UTLS White Paper DRAFT v1a                                                           26
 1   details are given by Fried et al. (2003), and can be found at
 2   http://www.acd.ucar.edu/UTLS/ScienceIndex.htm

 3   Objectives:

 4       •   Determine the radical production rates from the most important radical precursors
 5           and contrast these processes between the upper troposphere and lower
 6           stratosphere and across the sunrise and sunset transitions.

 7       •   Compare measurement and model calculations of HOx and precursor abundance
 8           in air masses affected by convection, by cirrus, and by high NOx levels.

 9       •   Determine the concentrations of NOy constituents including HONO, HNO4 and
10           organic and aerosol nitrates in the upper troposphere in the perturbed and
11           unperturbed UTLS.

12       •   Determine the impact of chemical processes on the local and larger scale ozone
13           budget via model analyses.

14   Experimental strategy: (1) Use the C-130 to provide information on lower and middle
15   troposphere composition and as a remote sensing platform with an upward-looking
16   LIDAR. Use HIAPER to obtain in-situ and remote sensing measurements near the
17   tropopause of radiation, species concentrations, and aerosols. (2) Use ground-based
18   radars and interferometers, rain collectors, and other sounding instruments to obtain
19   information on transport, clouds, and chemical composition. (3) Coordinate with satellite
20   overpasses, e.g. HIRDLS, TES. (4) Use global and regional-scale models, e.g. WRF to
21   guide flights and predict convective and frontal lifting events. The global model
22   MOZART will be used for long-range campaign planning, for next day forecasting and
23   flight planning, and for post-campaign analysis. (4) Conduct studies over the relatively
24   clean lower troposphere, e.g. the upper Great Plains, and over more polluted regions, e.g.
25   the southeastern U.S.

26   Measurement priority in addition to the standard suite of parameters on HIAPER: (1) in-
27   situ measurements of O3, CO, H2O, CH4, OH, HO2, RO2, jCH2O, VOCs especially CH2O,
28   H2O2, CH3OOH for CH2O studies. (2) PANs, NO, NO2, OH and aldehydes for aldehyde
29   studies. (3) For HOx studies, OH, HO2, CH3OOH, H2O2, HNO4, O3, j-values, CH2O,
30   acetone, and H2O. (4) For NOy studies, NO, NO2, N2O5, PANs, RONO2, HNO4, HNO3,
31   NOy, aerosol surface area, H2O.


33   IV.4. Aerosol and Cloud Particle Composition in the UTLS
34   Our understanding of the origins, transport pathways, and chemical effects, of aerosol

     NCAR UTLS White Paper DRAFT v1a                                                          27
 1   and cloud particles in the UTLS is far from complete. The tropopause region is a
 2   crossroads region, with influences from above and below. Volcanic sulfate particles are
 3   transported downward into the UTLS. Cirrus particles form near the tropopause on
 4   aerosol and insoluble particles, some of which originate from the lower troposphere.
 5   Boreal forest fires and high altitude aircraft inject particles into the UTLS. A firm
 6   understanding of UTLS aerosol, cirrus, and Polar Stratospheric Cloud (PSC) composition
 7   and temporal variations has not been established.

 8   IV.4.1.      Objectives

 9   The main objectives of this project are to address the following issues:

10   IV.4.1.1. What is the chemical composition and role of aerosol in UTLS chemistry
11             and in the generation of cirrus particles in the UTLS?

12   While the composition of aerosol in the stratosphere at mid-latitudes is basically
13   understood – the particles are liquid droplets composed of sulfuric acid (i.e. sulfate
14   particles), the composition of particles in the UTLS is more complicated. As discussed by
15   Murphy et al. (1998), in-situ measurements of the chemical composition of individual
16   aerosol particles at altitudes between 5 and 19 kilometers reveal that upper tropospheric
17   aerosols often contain more organic material than sulfate, and at least 45 elements have
18   been detected in the aerosol particles. Further measurements of the composition of
19   aerosol particles in the UTLS are needed to better understand their origins .

20   Laboratory measurements of heterogeneous rates of reaction on particles that reflect the
21   chemical make-up of UTLS aerosol and cirrus are needed to quantify the degree to which
22   chemical processing occurs near the tropopause. The extent to which cirrus particles
23   process gas phase species, in a manner similar to that of the PSCs in the polar regions,
24   has not been quantified. Hervig and McHugh (2002) present evidence of nitric acid
25   clouds near the tropical tropopause, based upon multi-wavelength Halogen Occultation
26   Experiment (HALOE) extinction data. Measurements of the HNO3 content of cirrus at
27   mid-latitudes (i.e. at altitudes accessible by HIAPER) is an important task.

28   The role of heterogeneous ice nucleation (i.e. nucleation of ice on insoluble or partially
29   insoluble particles) has been an open question due to a shortage of information on
30   concentrations and properties of ice nuclei (IN) in the upper troposphere. DeMott et al.
31   (1998) measured the fraction of total aerosol particles that were active as ice nuclei
32   during three airborne flights in the upper troposphere. These results show that the fraction
33   varies by a factor of ten at any given temperature. It is an important task to have a better
34   understanding of these variations, and relate these variations to differences in IN
35   composition.

36   IV.4.1.2. Are there fundamental aerosol composition differences between the
37             northern and southern hemisphere UTLS regions and does this impact the
38             water vapor density of the air entering the stratosphere?

39   Recent observations of UT cloud layers have been conducted during survey flights in the

     NCAR UTLS White Paper DRAFT v1a                                                          28
 1   northern and southern hemispheres during the European INCA (Interhemispheric
 2   differences in Cirrus properties from Anthropogenic emissions) project. The results have
 3   shown distinct differences in northern and southern hemisphere aerosols, implying a
 4   significant role for pollution. These differences are thought to lead to major differences in
 5   the cirrus cloud formation mechanisms: heterogeneous nucleation (in the northern
 6   hemisphere) versus homogeneous nucleation (in the southern hemisphere), with
 7   potentially large differences in the humidity of the air entering the stratosphere.

 8   IV.4.1.3. What is the composition of aerosol and PSCs in the polar regions?

 9   Rates of ozone depletion in the northern and southern polar regions in the lower
10   stratosphere are dependent upon the intricacies of PSC microphysics. The chemical
11   composition of PSCs has influence upon the rates at which inactive chlorine is
12   transformed into active chlorine. The details of the formation mechanisms of PSCs of
13   different composition types (e.g. solid NAT and NAD hydrates, liquid ternary solution
14   droplets, and ice particles), however, are not adequately understood. No microphysical
15   model can successfully predict the time evolution of PSC composition types or predict
16   the altitudinal distribution of co-existing compositional types. No three dimensional
17   transport model can successfully predict the complexities of stratospheric dynamics that
18   produce the temperature fields that allow PSCs to exist.


20   IV.4.2.      Proposed Studies

21   IV.4.2.1. Chemical composition of aerosol and cloud particles in the UTLS

22   We propose to deploy chemical composition instrumentation to measure the composition
23   of aerosol and cloud particles in the UTLS. Of particular interest is to characterize the
24   organic and HNO3 contents, and the particle size distributions of the aerosol and cloud
25   particles. Individual single particle, filter collection, mass spectrometry, and particle size
26   instrumentation would be deployed.

27   With regard to cirrus measurements, attention would focus upon the evolution of ice
28   particle size distributions in the upper regions of convective storm anvils, downwind of
29   the anvil, and in regions in which the uplift of humid layers produces cirrus in the upper
30   troposphere. Of particular interest is to combine chemical analysis of the CNN and IN
31   particles with analysis of the crystal growth characteristics, and the ambient motion
32   fields, to determine how differences in CNN and IN influence the crystal growth rates.

33   IV.4.2.2. Examine aerosol composition differences between the northern and
34             southern hemisphere UTLS regions and the impact on cirrus formation.

35   Drawing upon the approach suggested by Heymsfield and Miloshevich (1995) and for the
36   INCA observations (Haag et al. 2003), an examination of the relative humidities in and
37   out of cirrus can provide insight into the mechanisms of cirrus formation: peak humidities
38   between 130 or 140% and 160% with respect to ice suggest homogeneous ice nucleation

     NCAR UTLS White Paper DRAFT v1a                                                             29
 1   whereas peak humidities of 120-130% suggest heterogeneous nucleation. We propose to
 2   characterize the peak humidities in and around UTLS cirrus with accurate water vapor
 3   sensors to characterize the cirrus formation mechanisms in terms of the aerosol
 4   composition and sizes.

 5   IV.4.2.3. Distribution and composition of aerosol and PSCs in the polar regions.

 6   To make progress in our understanding of PSCs and ozone loss, we propose to utilize
 7   airborne and satellite instrumentation to measure PSC composition type and its evolution,
 8   and associated parameters (such as temperature, and gas phase mixing ratios) over
 9   periods of time in which PSC formation, chlorine transformation, and ozone recovery
10   phases can be studied in detail. Airborne instruments will be deployed 1) to identify the
11   chemical composition and approximate particle size distributions of PSCs, 2) to map their
12   locations and geographical extent, 3) to observe the PSCs under a number of conditions
13   in order to comment on formation mechanisms, 4) to measure the effects of PSCs on the
14   chemistry of several species (e.g. O3, NO2, HNO3), and 5) to place the aircraft
15   observations into a wider context of observations made by satellite-borne instruments.

16   Satellite instruments will be used to provide additional and complementary information.
17   The HIRDLS and TES instruments on the Aura platform will be used to study the spatial
18   and temporal evolution of PSCs with particular reference to their formation temperatures,
19   locations and 3D morphology. The Aura-MLS measurements of enhanced ClO along
20   isentropic trajectories downstream from sunlit PSCs will be analyzed along with
21   measurements of Cl reservoir species and PSC particle size and temperature in order to
22   investigate the conversion to reactive Cl. Box model calculations will be run along
23   trajectories to validate the current understanding of chlorine partitioning and activation.

24   In addition, the process of PSC formation will be studied using Lagrangian analyses of
25   air parcels during the formation of large-scale PSCs. Time-histories of air parcels will
26   allow calculation of cooling rates and sedimentation velocities. Statistics on the
27   frequency of PSC occurrence, spatial extent, longevity and correlation with orographic
28   features will be compiled from the wide-area coverage of the satellite data. Classification
29   of PSC types will be attempted from infrared spectral measurements of the infrared
30   aerosol extinction and compared to theoretical spectra.

31   IV.4.3.      Airborne Experiments

32   Measurement strategies of the airborne experiments are as follows.

33   IV.4.3.1. Composition of aerosol and PSCs in the polar regions

34   Airborne instrumentation would include a lidar, a microwave temperature profiler (MTP),
35   a sun photometer, and a Fourier Transform Infrared Interferometer (FTIR). Lidar
36   depolarization data would provide physical phase (e.g. indicate the location of liquid and
37   solid particles), the sun photometer and FTIR data would provide size distribution
38   information, and the FTIR would provide additional composition information, since
39   different PSC composition types have different extinction wavelength variations. The

     NCAR UTLS White Paper DRAFT v1a                                                          30
 1   FTIR would also measure columns along the line of sight of species such as O3, HCl,
 2   HNO3, NO2. Proposed HIAPER flight paths will allow the airborne instruments to view
 3   local noon for about four hours while crossing regions of varying potential vorticity and
 4   stratospheric temperature. Scandanavia would be the base of operation, with a flight
 5   along a latitude circle at constant noon for about four hours. HIAPER would turn back
 6   east and move approximately 200 km south, so that the instruments would be directly
 7   beneath the region of the stratosphere at ~18 km which were observed with the solar
 8   instruments on the westbound leg. As HIAPER flies east, the sun quickly sets, and the
 9   sky is dark for optimum lidar observations. On the return flight the zenith instruments
10   would observe the same air parcels. Multiple flights with similar flight paths over the
11   course of the winter would trace the evolution of the PSCs and the chemistry of the
12   stratosphere. Initial observations would be made in early December, before temperatures
13   are cold enough for PSC formation, and observations would be made during the coldest
14   part of winter, and at the end of winter to evaluate the amount of ozone loss. For a
15   complete discussion of aircraft observation strategy in the polar regions, please see
16   Coffey et al., at http://www.acd.ucar.edu/UTLS/ScienceIndex.htm.

17   IV.4.3.2. Aerosol and cirrus particles in the UTLS.

18   The high altitude capability of HIAPER allow new experiments to be done in the vicinity
19   of cirrus and anvil clouds near the extratropical tropopause. HIAPER could also sample
20   Boreal fire smoke in the UTLS guided by trajectory calculations, and HIRDLS
21   observations of aerosol in the lower stratosphere.


23   V.      Concluding Remarks

24   To achieve the above outlined objectives, it is essential to have collaborations and
25   expertise on multi-scale processes across dynamics, chemistry, microphysics and
26   radiation. NCAR is particularly well place to facilitate this kind of collaboration, because
27   of the broad range of expertise that exists among NCAR divisions. Scientists from ACD
28   are contributing expertise in the chemistry (including multiphase chemistry) and large
29   scale dynamics, using both observations and models. Scientists from MMM division and
30   RAP are contributing their expertise in cloud microphysics, mesoscale dynamics and
31   turbulence studies. ATD participants contribute their experience with cloud microphysics
32   and expertise in the planning of airborne experiments in general. Scientists from CGD
33   lead the chemistry-climate coupling. This white paper is written not only to outline
34   science issues in the UTLS, to identify the synergy between observations and models, to
35   plan the use of results from a new generation observation capabilities and multi-scale
36   models, but also to facilitate collaborations between NCAR scientists and colleagues in
37   the wider community.


     NCAR UTLS White Paper DRAFT v1a                                                          31
 1   References

 2   Alexander, M. J., T. Tsuda, and R. A. Vincent, Latitude variations observed in gravity
 3       waves with short vertical wavelengths, J. Atmos. Sci., 59, 1394–1404, 2002.
 4   Appenzeller, C., H. C. Davies and W. A. Norton, Fragmentation of stratospheric
 5       intrusions, J. Geophys. Res., 101, 1435-1456, 1996.
 6   Barth, M. C., A. L. Stuart, and W. C. Skamarock, Numerical simulations of the July 10
 7        Stratospheric-Tropospheric Experiment: Radiation, Aerosols and Ozone/Deep
 8        Convection storm: Redistribution of soluble tracers, J. Geophys. Res., 106, 12,381-
 9        12,400, 2001.
10   Birner, T. and A. Dornbrack and U. Schumann, How sharp is the tropopause at
11       midlatitudes? Geophys. Res. Lett., 2002, 29(14), 10.1029/2002GL015142.

12   Bithell, M., G. Vaughan, and L. J. Gray, Persistence of stratospheric ozone layers in the
13        troposphere, Atmos. Environ., 34, 2563-2570, 2000.
14   Boehm, M. T., and J. Verlinde, Stratospheric influence on upper tropospheric tropical
15       cirrus, Geophys. Res. Lett., 27, 3209– 3212, 2000.
16   Bradshaw, N. G. , G. Vaughan and G. Ancellet. Generation of layering in the lower
17       stratosphere by a breaking Rossby wave. J. Geophys. Res., 107, D2, ACL 7-1 to 7-
18       10, 2002.
19   Bradshaw, N. G. , G. Vaughan, R. Busen, S. Garcelon, R. L. Jones, T. Gardiner and J.
20       Hacker. Tracer filamentation generated by small-scale Rossby wave breaking in the
21       lower stratosphere. J. Geophys. Res., 107 (D23), doi: 10.1029/2002JD002086, ACL
22       12-1 to 12-14, 4689, 2002.
23   Brune, W.H., et al., Airborne In-Situ OH and HO2 Observations in the Cloud-Free
24       Troposphere and Lower Stratosphere During SUCCESS, Geophys. Res. Lett., 25
25       (10), 1701-1704, 1998.
26   Clark, T.L., W.D. Hall, and J.L. Coen, Source code documentation for the Clark-Hall
27        cloud-scale model: Code version G3CH01. NCAR technical note, NCAR/TN-
28        426+STR, 1996.
29   Cooper, O. R., J. L. Moody, D. D. Parrish, M. Trainer, J. S. Holloway, T. B. Ryerson, G.
30       Hübler, F. C. Fehsenfeld, S. J. Oltmans and M. J. Evans, Trace gas signatures of the
31       airstreams within North Atlantic cyclones - Case studies from the NARE'97 aircraft
32       intensive, J. Geophys. Res., 106, 5437-5456, 2001.
33   Cooper, O. R., C. Forster, D. Parrish, E. Dunlea, G. Hübler, F. Fehsenfeld, J. Holloway,
34       S. Oltmans, B. Johnson, A. Wimmers, and L. Horowitz, On the life-cycle of a
35       stratospheric intrusion and its large-scale mixing with polluted warm conveyor belts,
36       J. Geophys. Res., in-review, 2003.
37   Crutzen, P. J. and M. G. Lawrence, The impact of precipitation scavenging on the
38       transport of trace gases: A 3-dimensional model sensitivity study, J. Atmos. Chem.,
39       37, 81-112, 2000.

     NCAR UTLS White Paper DRAFT v1a                                                          32
 1   DeMott, P. J., D. C. Rogers, S. M. Krqeidenweis, Y. Chen, C. H. Twohy, D.
 2      Baumgardner, A. J. Heymsfield, and K. R. Chan, The role of heterogeneous freezing
 3      nucleation in upper tropospheric clouds: Inferences from SUCCESS, Geophys. Res.
 4      Lett., 25, 1387-1390, 1998.
 5   Emmons, L.K., et al., Budget of tropospheric ozone during TOPSE from two chemical
 6      transport models, J. Geophys. Res., 108, 8372, doi:10.1029/2002JD002665, 2003.
 7   Esler, J. G., D. G. H. Tan, P. H. Haynes, M.J. Evans, K. S. Law, P.-H. Plantevin, and J.
 8        A. Pyle, Stratosphere-troposphere exchange: Chemical sensitivity to mixing, J.
 9        Geophys. Res., 106, 4717-4731, 2001.
10   Faloona, I., et al., Observations of HOx and its Relationship with NOx in the Upper
11       Troposphere During SONEX, J. Geophys. Res., 105, (D3), 3771 – 3783, 2000.
12   Fischer, H., F. G. Wienhold, P. Hoor, O. Bujok, C. Schiller, P. Siegmund, M. Ambaum,
13        H. A. Scheeren, and J. Lelieveld, Tracer correlations in the northern high latitude
14        lowermost stratosphere: Influence of cross-tropopause mass exchange, Geophys.
15        Res. Lett., 27, 97–100, 2000.
16   Folkins, I., Origin of lapse rate changes in the upper tropical troposphere, J. Atmos. Sci.,
17        59, 992–1005, 2002.
18   Fromm, M. and R. Servranckx, Transport of Forest Fire Smoke Above the Tropopause
19       by Supercell Convection, GRL, in press 2003.
20   Fromm, M., J. Alfred, K. Hoppel, J. Hornstein, R. Bevilacqua, E. Shettle, R. Servanckx,
21       Z. Li, and B. Stocks, Observations of boreal forest fire smoke in the stratosphere by
22       POAM III, SAGE II, and lidar in 1998, Geophys. Res. Lett., 27, 1407-1410, 2000.
23   Gettelman A., P. M.D. Forster, A Climatology of the Tropical Tropopause Layer, J. Met.,
24        Soc. Japan, 80, 911-924, 2002.
25   Gettelman, A., A. R. Douglass, and J. R. Holton. Simulations of water vapor in the upper
26        troposphere and lower stratosphere. J. Geophys. Res., 105:9003-9023, 2000.
27   Griffiths, M., A. J. Thorpe, and K. A. Browning, Convective destabilization by a
28        tropopause fold diagnosed using potential-vorticity inversion, Q. J. R. Meteorol.
29        Soc. 126, 125-144, 2000.Harries, J. E., The greenhouse Earth: A view from space,
30        Q. J. Roy. Met., Soc., 122, 799, 1996.
31   Haag, W., B. Karcher, J. Strom, A. Minikin, U. Lohmann, J. Ovarlez, and A. Stohl,
32       Freezing thresholds and cirrus cloud formation mechanisms inferred from in-situ
33       measurements of relative humidity. Atmos. Chem. Phys. Discuss., 3, 3267—3299,
34       2003.
35   Hartmann, D. L. and K. Larson, An important constraint on tropical cloud-climate
36       feedback, Geophys. Res. Lett. 29:20, doi:10.1029/2002GL015835, 2002
37   Hartmann, D. L., J. R. Holton, and Q. Fu, The heat balance of the tropical tropopause,
38       cirrus, and stratospheric dehydration, Geophy. Res. Lett., 28, 1969-1972, 2001.
39   Hervig, M., and M. McHugh, Tropical nitric acid clouds, Geophys. Res. Lett., 29(No.7),
40       doi:10.1029/2001GL014271,2002.

     NCAR UTLS White Paper DRAFT v1a                                                           33
 1   Heymsfield, A. J., Ice particles observed in a cirriform cloud at --83C and implications
 2      for polar stratospheric clouds. J. Atmos. Sci., 43, 851—855, 1986.
 3   Heymsfield, A. J., and L. M. Miloshevich, Relative humidity and temperature influences
 4      on cirrus formation and evolution: Observations from wave clouds and FIRE-II. J.
 5      Atmos. Sci., 52, 4302—4326, 1995.
 6   Hienola, J., M. Kulmala, A. Laaksonen, Model studies on the effect of nitric acid vapour
 7       on cirrus cloud activation, Atmos. Res. 65, no. 3-4, 235-250, Jan 2003.
 8   Holton , J. R., and A. Gettelman. Horizontal transport and dehydration in the
 9       stratosphere. Geophys. Res. Lett., 28(14):2799-2802, 2001.
10   Holton, J. R., P. H. Haynes, A. R. Douglass, R. B. Rood, and L. Pfister. Stratosphere-
11       troposphere exchange. Rev. Geophys., 33(4):403-439, 1995.
12   Hoor, P., H. Fischer, L. Lange, J. Lelieveld, and D. Brunner, Seasonal variations of a
13       mixing layer in the lowermost stratosphere as identified by the CO-O3 correlation
14       from in situ measurements, J. Geophys. Res., 107(D5), 4044,
15       doi:10.1029/2000JD000289, 2002.
16   Jaeglé, L, et al., Photochemistry of HOx in the Upper Troposphere at Northern
17       Midlatitudes, Geophys. Res., 105 (D3), 3877 –3892, 2000.
18   James, P., A. Stohl, C. Forster, S. Eckhardt, P. Seibert, and A. Frank, A 15-year
19       climatology of stratosphere–troposphere exchange with a Lagrangian particle
20       dispersion model, 2, Mean climate and seasonal variability, J. Geophys. Res.,
21       108(D12), 8522, doi:10.1029/2002JD002639, 2003.
22   Jensen, E. J., O. B. Toon, H. B. Selkirk, J. D. Spinhirne, and M. R. Schoeberl, On the
23        formation and persistence of subvisible cirrus clouds near the tropical tropopause, J.
24        Geophys. Res., 101, 21,361 – 21,375,1996.
25   Kiladis, G. N., and K. M. Weickmann, Circulation anomalies associated with tropical
26        convection during northern winter. Mon. Wea. Rev., 120, 1900–1923, 1992.
27   Lane, T.P., M.J. Reeder, and T.L. Clark, Numerical modeling of gravity wave generation
28       by deep tropical convection. J. Atmos. Sci., 58, 1249-1274, 2001.
29   Lane, T.P., R.D. Sharman, T.L. Clark, and H.-M. Hsu, An investigation of turbulence
30       generation mechanisms above deep convection. J . Atmos. Sci., 60, 1297-1321,
31       2003.
32   Lanzendorf, E. J., T. F. Hanisco, P. O. Wennberg, R. C. Cohen, R. M. Stimpfle, and J. G.
33       Anderson, Comparing atmospheric (HO2)/(OH) to modeled (HO2)/(OH): Identifying
34       discrepancies with reaction rates, Geophys. Res. Lett., 28, 967-970, 2001.
35   Lau, K. Km., T. Nakazawa, and C. H. Sui, Observations of cloud cluster hierarchies over
36        the tropical western Pacific, J. Geophys. Res., 96, 3857-3208, 1991.
37   Lawrence, M. G. and P. J. Crutzen, The impact of cloud particle gravitational settling on
38       soluble trace gas distributions Tellus, 50B, 263-289, 1998.
39   Massie, S., P. Lowe, X. X. Tie, M. Hervig, G. Thomas, and J. M. Russell III. The effect

     NCAR UTLS White Paper DRAFT v1a                                                            34
 1        of the 1997 El Ni\ no on the distribution of upper tropospheric cirrus. J. Geophys.
 2        Res., 105(D18):22,725-22,741, 2000.
 3   Massie, S. T., A. Getelman, W. Randel, and D. Baumgardner, Distribution of tropical
 4       cirrus in realation to convection, J. Geophys. Res.,107,doi:10.1029/2001JG2001293,
 5       2002.
 6   McFarquhar, G. M., A. J. Heymsfield, J. Spinhirne, and B. Hart, Thin and Subvisual
 7      Tropopause Tropical Cirrus: Observations and Radiative Impacts, J. Atm. Sci., 57,
 8      1841-1853, 2000.
 9   McFarquhar, G.M., A.J. Heymsfield, J. Spinhirne, and W. Hart, Subvisible tropopause
10      tropical cirrus: Observations and radiative impacts, J. Atmos. Sci., 57, 1841-1853,
11      2000.
12   McKenna, D. S., P. Konopka, J.-U. Grooß, G. Günther, R. Müller, R. Spang, D.
13      Offermann, and Y. Orsolini, A new Chemical Lagrangian Model of the Stratosphere
14      (CLaMS) 1. Formulation of advection and mixing, J. Geophys. Res., 107(D16),
15      4309, doi:10.1029/2000JD000114, 2002.
16   Murphy, D. M., D. S. Thomson, M. J. Mahoney, In Situ Measurements of Organics,
17       Meteoritic Material, Mercury, and Other Elements in Aerosols at 5 to 19 Kilometers,
18       Science, 282, 1664-1669, 1998.
19   Oltmans, S. J. , H. Voemel, D. J. Hofmann, K. H. Rosenlof, and D. Kley. The increase in
20       stratospheric water vapor from balloon-borne, frostpoint hygrometer measurements
21       at Washington, D.C. and Boulder, Colorado. Geophys. Res. Lett., 27(21):3453-
22       3456, 2000.
23   O'sullivan, D., Dunkerton, T. J. 1995: Generation of Inertia–Gravity Waves in a
24        Simulated Life Cycle of Baroclinic Instability. . J. Atmos. Sci Vol. 52, No. 21, pp.
25        3695-3716.
26   Pan, L. L., W. J. Randel, E. Browell, B. J. Gary, M. J. Mahoney, and E. J. Hintsa,
27        Definitions and the Sharpness of the Extratropical Tropopause: A Trace Gas
28        Perspective, submitted to J. Geophys. Res, 2003.

29   Park, M., W. J. Randel, D. E. Kinnison, R. R. Garcia, and W. Choi, Seasonal variation of
30        Methan, water vapor, and nitrogen oxides near the tropopause: satellite observations
31        and model simulations, submitted to J. Geophys. Res, 2003.

32   Pavelin, E. and J. A. Whiteway and R. Busen and J. Hacker, Airborne observations of
33       turbulence, mixing, and gravity waves in the tropopause region J. Geophys. Res.,
34       2002, 1029/2001JD000775.

35   Pfister, L., et al., Aircraft observations of thin cirrus clouds near the tropical tropopause,
36        J. Geophys. Res., 106, 9765-9786, 2001.

37   Prather, M.J., and D.J. Jacob, A Persistent Imbalance in HOx and NOx Photochemistry of
38        the Upper Troposphere Driven by Deep Tropical Convection, Geophys. Res. Lett.,
39        24 (24), 3189 – 3192, 1997.

     NCAR UTLS White Paper DRAFT v1a                                                              35
 1   Ralph, F.M., P. J. Neiman, and D. Levinson, 1997: Lidar observations of a breaking
 2       mountain wave associated with extreme turbulence. Geophys. Res. Lett., 24, 663-
 3       666.
 4   Randel, W. J., F. Wu, and D. J. Gaffen. Interannual variability of the tropical tropopause
 5       derived from radiosonde data and NCEP reanalyses. J. Geophys. Res.,
 6       105(D12):15,509-15,523, 2000.
 7   Randel, W. J., A. Gettelman, F. Wu, J. M. Russell III, J. Zawodny, and S. Oltmans.
 8       Seasonal variation of water vapor in the lower stratosphere observed in Halogen
 9       Occultation Experiment data. J. Geophys. Res., 106:14313-14325, 2001.
10   Randel, W. J., F. Wu, and W. Rivera Rı´os, Thermal variability of the tropical tropopause
11       region derived from GPS/MET observations, J. Geophys. Res., 108(D1), 4024,
12       doi:10.1029/2002JD002595, 2003.
13   Seidel, D. J. , R. J. Ross, J. K. Angell, and G. C. Reid. Climatological characteristics of
14       the tropical tropopause as revealed by radiosondes. J. Geophys. Res., 106:7857-
15       7878, 2001.
16   Seo, K.-H., and K. P. Bowman, Lagrangian estimate of global stratosphere-troposphere
17        mass exchange, J. Geophys. Res., 107(D21), 4555, doi:10.1029/2002JD002441,
18        2002.
19   Shapiro, M. A., Turbulent mixing within tropopause folds as a mechanism for the
20       exchange of chemical constituents between the stratosphere and troposphere, J.
21       Atmos. Sci., 37, 994-1004, 1980.
22   Sherwood , S. C., and A. E. Dessler. On the control of stratospheric humidity. Geophys.
23       Res. Lett., 27(16):2513-2516, 2000.
24   Sherwood, S. C. and A. E. Dessler. A model for transport across the tropical tropopause.
25       J. Atmos. Sci., 58:765-779, 2001.
26   Sherwood, S. C., Maintenance of the free-tropospheric tropical water vapor distribution,
27       Part II, simulation by large scale advection. J. Clim., 9(11):2919-2934, 1996.
28   Sprenger, M., and H. Wernli, A northern hemispheric climatology of cross-tropopause
29       exchange for the ERA15 time period (1979–1993), J. Geophys. Res., 108(D12),
30       8521, doi:10.1029/2002JD002636, 2003.
31   Staudt, A. C., D. J. Jacob, J. A. Logan, D. Bachiochi, T. N. Krishnamurti, and N. Poisson,
32       Global chemical model analysis of biomass burning and lightning influences over
33       the south Pacific in austral spring, J. Geophys. Res., 107, (D14),
34       10.1029/2000JD00296, 2002.
35   Stohl, A., A 1-year Lagrangian ‘‘climatology’’ of airstreams in the Northern Hemisphere
36        troposphere and lowermost stratosphere, J. Geophys. Res.,106, 7263–7279, 2001.
37   Stohl, A., Stratosphere-troposphere exchange: A review, and what we have learned form
38        STACCATO, J. Geophys. Res., 108(D12), 8516, doi:10.1029/2002JD002490, 2003.
39   Thuburn, J., and G. C. Craig, On the temperature structure of thetropical substratosphere,
40       J. Geophys. Res., 107(D1), doi:10.1029/2001JD000448, 2002.

     NCAR UTLS White Paper DRAFT v1a                                                               36
 1   W. J. Randel, A. Gettelman, F. Wu, J. M. Russell III, J. Zawodny, and S. Oltmans.
 2        Seasonal variation of water vapor in the lower stratosphere observed in Halogen
 3        Occultation Experiment data. J. Geophys. Res., 106:14313-14325, 2001.
 4   Wang, P. K., Moisture plumes above thunderstorm anvils and their contributions to cross-
 5      tropopause transport of water vapor in midlatitudes, J. Geophys. Res., 108(D6),
 6      4194, doi:10.1029/2002JD002581, 2003.
 7   Wang, P.-H., P. Minnis, M. P. McCormack, G. S. Kent, and K. M. Skeens, A 6-year
 8      climatology of cloud occurrence frequency from Stratospheric Aerosol and Gas
 9      Experiment II observations (1985-1990), J. Geophys. Res., 101, 29407-29429, 1996.
10   Waugh, D. W., and L. M. Polvani. Climatology of intrusions into the tropical upper
11      troposphere. Geophys. Res. Lett., 27(23):3857-3860, 2000.
12   Wennberg, et al., Hydrogen radicals, nitrogen radicals, and the production of ozone in the
13      upper troposphere, Science 279, 49-53, 1998.
14   Wernli, H., and M. Bourqui, A Lagrangian “1-year climatology” of (deep) cross-
15       tropopauseexchange in the extratropical Northern Hemisphere, J. Geophys. Res.,
16       107, 10.1029/2001JD000812, 2002.
17   Wheeler, M., G. N. Kiladis, and P. J. Webster, Large-scale dynamical fields associated
18      with convectively coupled equatorial waves, J. Atmos. Sci., 57, 613– 640, 2000.
19   Winkler, D. M., and C. R. Trepte, Laminar cirrus observed near the tropical tropopause
20       by LITE, Geophys. Res. Lett., 25, 3351–3354, 1998.
21   Zahn, A., et al., Identification of extratropical two-way troposphere-stratosphere mixing
22       based on CARIBIC measurements of O3, CO, and ultrafine particles, J. Geophys.
23       Res., 105, 1527–1535, 2000.
24   Zhang, F., Generation of mesoscale gravity waves in upper-tropospheric jet-front
25       systems. J. Atmos. Sci., submitted, 2003.

     NCAR UTLS White Paper DRAFT v1a                                                          37
 6   Figure 1. This schematic figure highlights the important processes coupling dynamics,
 7   chemistry and cloud microphysics in the UTLS region. The green line denotes the time
 8   average tropopause. In the tropics, maximum outflow from deep convection occurs near
 9   ~12-14 km, while the cold point tropopause occurs near 17 km. The intervening region
10   has characteristics intermediate between the troposphere and stratosphere, and is termed
11   the tropical transition layer (TTL). Extratropical stratosphere-troposphere exchange
12   occurs in tropopause folds and intrusions linked with synoptic weather systems; these
13   events transport stratospheric ozone into the troposphere. In addition, convection brings
14   near-surface pollutants (from biomass burning or anthropogenic emissions) into the upper
15   troposphere, strongly influencing global-scale chemistry.

     NCAR UTLS White Paper DRAFT v1a                                                       38

 5   Figure 2: An AIRS “granule” of nighttime observations of September 6, 2002. The upper
 6   left panel shows brightness temperature differences between 980 and 811 cm-1 (related to
 7   cloud characteristics), while the upper three right panels show retrieved cloud fraction,
 8   temperature and water vapor at 250mb. The bottom panels show retrieved temperature
 9   and water mass mixing ratio profiles on the (nine) locations inside the boxes drawn
10   above.

     NCAR UTLS White Paper DRAFT v1a                                                       39
32   Figure 3. UTLS ozone cross-section from MOZART3 model (using the ECMWF wind).
33   Position of the subtropical jet is indicated by the zonal wind maximum (white contour).
34   This figure demonstrates the large gradient of ozone cross the tropopause and the
35   intrusion into the troposphere in the vicinity of the subtropical jet. The thermal
36   tropopause is marked by the white dash line. The dynamical tropopause is indicated by
37   the PV contours. 2 PVU and 3.5 PVU contours are shown (in red). Isentropes (shown as
38   black lines) intercept the tropopause in the vicinity of the subtropical jet.

     NCAR UTLS White Paper DRAFT v1a                                                      40

 6   Figure 4. Vertical profile of temperature anomalies near the equator, measured by GPS
 7   radio occultation on 28 January 2002. Each profile shows the temperature anomalies
 8   calculated by subtracting a constant background structure; all measurements occurred
 9   over latitudes 7 N-S, at the longitudes noted by the light dashed lines. The scale for the
10   amonalies is shown at the upper right. Note the global-scale coherence, and the eastward
11   phase tilt with altitude of the maxima and minima, which are signatures of an
12   equatorially-centered Kelvin wave.

     NCAR UTLS White Paper DRAFT v1a                                                         41
29   Figure 5. Effect of mixing of stratospheric and tropospheric air observed in the vicinity
30   of the STJ. The tracers and temperature measurements are made onboard DC-8 on
31   October 29th, 1997 during SONEX. O3-CO and O3-H2O for the flight segment between
32   17 UTC and 18.7 UTC, with the chemical and thermal background information given by
33   the LIDAR measured ozone curtain (lower left) and potential temperature lapse rate
34   curtain (lower right) derived from the MTP temperature profiles. In all four panels,
35   letters A, B, C, and D help identify the segment of flight where the tracer mixing ratios
36   form “mixing lines”. (Pan et al., 2003)

     NCAR UTLS White Paper DRAFT v1a                                                        42
 2   (a)
19   (b)

35   Figure 6. Images from a high resolution numerical simulation of deep midlatitude
36   convection (derived from Lane et al. 2003). (a) illustrates the breakdown of vertically
37   propagating gravity waves inthe lower stratosphere, and the resultant turbulent mixing.
38   Contours are potential temperature at 2K intervals, cloudy air (blue) and regions of
39   convective instability outside cloud (red) are shaded. (b) shows a close-up view at the
40   cloud top, highlighting intrusions of stratospheric air downwards into the cloudy
41   troposphere, which are subsequently irreversibly mixed. Contours are potential
42   temperature at 1K intervals and cloudy air is shaded (blue).

     NCAR UTLS White Paper DRAFT v1a                                                           43
4   Figure 7. Total Ozone Mapping Spectrometer (TOMS) 0.55 micron aerosol optical
5   depths for the month of July 1998. The high value of optical depth over the Eastern
6   Siberia indicates the intense aerosol production due to the Boreal forest fires.

    NCAR UTLS White Paper DRAFT v1a                                                       44
2   Figure 8. Convectively transport of NO. Nitric oxide (NO) measured with the North
3   Dakota Citation aircraft within the anvils of 10 thunderstorms over the northeastern
4   plains of Colorado. The apparent spikes are actually enhanced NO over scales of 30-50
5   km as the aircraft transected an anvil at constant altitude. These large mixing ratios will
6   have a large impact on the production of ozone downwind of the storms. Only data
7   above ~7 km applies to the storm anvils.

    NCAR UTLS White Paper DRAFT v1a                                                           45

 4   Figure 9. The production rate of HOx (A) and the concentration of OH (B) on 7
 5   November 1995. (A) As shown in blue, the HOx production rate from the reaction of
 6   O(1D) with H2O (process 11) drops by orders of magnitude between 7 km and the
 7   tropopause following the drop in the mixing ratio of H2O. Shown in red is an estimate of
 8   the HOx production rate from photolysis of acetone, which recent measurements have
 9   shown is ubiquitous in the upper troposphere. Both the instantaneous production rates
10   (solid lines) and the 24-hour average rates (dashed lines) are shown. (B) Without the
11   acetone source, the measured [OH], shown here filtered with a 30-s running median, and
12   [HO2] (not shown) are underpredicted by about a factor of 2 between 12 km and the
13   tropopause. Even with acetone, [HOx] is often underpredicted. For example, at the
14   bottom of this profile, measured OH concentrations are 20 to 100% larger than
15   calculated. Typical of all the observations, the agreement between calculated and
16   measured [OH] is excellent in the stratosphere (Wennberg, 1998).

     NCAR UTLS White Paper DRAFT v1a                                                       46
27   Figure 10. Meridional cross sections of NOx in the South Asian monsoon region (40-100
28   E) in September. The left panel shows satellite measurements from the Halogen
29   Occultation Experiment (HALOE), and extend down only to the 100 hPa level. The
30   right panel shows results from a MOZART simulation. The heavy dashed line is the
31   tropopause. Note the upper tropospheric NOx maximum in the MOZART result (due to
32   parameterized lightning NOx generation), which extends into the lower stratosphere.
33   [Paket al., 2003]

     NCAR UTLS White Paper DRAFT v1a                                                   47

To top