Stratosphere-Troposphere Exchange (STE) is one of the key factors controlling the budgets of ozone, water vapor and other substances in both the troposphere and the lower stratosphere. It has traditionally been understood as the flux of air or trace constituents across the tropopause. However, this concept of STE is quite limited, since STE is also controlled by the rate at which tropospheric and stratospheric species are supplied to and removed from the tropopause region. Hence, this must be included into our picture of STE. Deep exchange events are more relevant for atmospheric chemistry as they lead to composition changes also in the lowermost troposphere and relatively high up in the stratosphere, whereas shallow exchange events only produce composition changes in the tropopause region and tend to be of a more reversible nature. A good example for the importance of deep exchange events is given in a recent study1 that has shown that aerosols originating from the earth´s surface may be found at altitudes up to 19 km (6 km above the tropopause) even in the middle latitudes. Another example are stratospheric intrusions reaching the earth´s surface, thereby directly contributing to the springtime ozone maximum at remote sites. Such processes are totally unaccounted for if only the fluxes across the tropopause are considered, as was done in most global and long-term studies of STE.


 1
We propose to develop a more complete global picture of STE by recording the physical and chemical fate of air parcels travelling through the tropopause region. The first objective of STACCATO is, thus,
to develop a new three-dimensional Lagrangian perspective of STE as an alternative to bulk exchange across the tropopause, with a focus on processes associated with "deep" exchange events in both directions.


2
In extratropical cyclones, air masses of very different chemical characteristics (boundary-layer air, free tropospheric air, stratospheric air) are stirred. An important, but poorly understood, topic for atmospheric chemistry is how and on what time-scales these air masses mix with each other. STACCATO will study three processes, representing a time sequence leading to mixing: filamentation of tracer structures by chaotic advection, diabatic decay of filaments due to radiative processes, and filament-destruction by turbulence. Because of non-linear chemical reactions, the timescale of mixing is of key importance for atmospheric chemistry. Idealized meteorological and chemical box models and global chemistry models and measurement data will be used to investigate in detail the mixing of stratospheric and tropospheric air.


 3
For model validation, but also for developing a model-independent view of STE, measurements are needed. Therefore, it is planned to do measurements in an extended network, suitable to estimate the strength of STE and its impact on tropospheric chemistry, and to validate the STACCATO models.
Radionuclide measurements provide especially valuable information on STE. A data set comprising beryllium 7 (7Be), other radionuclides and chemical and meteorological parameters is available from four high mountain sites. These measurements will be continued and supplemented with continuous measurements of 10Be at two of these sites, because the ratio of the concentrations of 7Be and 10Be provides an excellent measure of the strength of STE. Combining data from mountain stations, new high-resolution ozone lidar measurements and ozone soundings from Central, Southern and Southeastern Europe, with existing data (satellite products, aircraft measurements, ozone soundings and radionuclide measurements), process studies will be made to deepen our understanding of intrusions of stratospheric air into the troposphere and to identify shortcomings of the models.


 4
Model testing is an integral part of STACCATO. The STACCATO team will use seven different models and methods to diagnose stratosphere-troposphere exchange. An important task to reduce uncertainties in current estimates of STE is thus to intercompare methods and models used to calculate STE, to find the strengths and weaknesses of each approach and to identify the reasons for discrepancies.


 5
There are indications for ongoing climatic changes in the tropopause region like an increase in the average tropopause height, and climate models predict even more pronounced changes for the future. Therefore, another focal point of STACCATO is to study the variability and possible trends of STE during the past few decades and to investigate STE under scenarios of climate change in future times.
Several Eulerian and Lagrangian diagnostics will be applied on consistent high-resolution global meteorological re-analysis datasets for a period of at least fifteen years to derive a unique picture of the past and present spatio-temporal variability and possible trends of STE. To examine possible future changes in STE, simulations with two climate models, one with high vertical resolution in the middle atmosphere, will be performed. Online tracers and Lagrangian postprocessing of the model output will be used to study STE. The impact of changes in STE on tropospheric chemistry will be studied with chemistry-climate models.


 6
It is recognized that STE has a large impact on the oxidizing capacity of the troposphere. Nevertheless, its contribution to the tropospheric and lower stratospheric ozone budget relative to photo-chemical ozone formation from natural and anthropogenic precursor emissions, including those from aircraft, is still uncertain. Using three Eulerian and Lagrangian global chemistry models based on the output of climate models and meteorological re-analyses, it is aimed at studying the relative impact of STE processes on factors controlling the oxidation capacity of the troposphere and lower stratosphere.





Although stratosphere-troposphere exchange (STE) and its effect on atmospheric chemistry have been studied for a long time now, there are still large unknowns that have to be addressed. For instance, it is still heavily debated to what extent STE contributes to the springtime ozone maximum observed at remote surface sites. We believe that the continuing uncertainties are due to several factors, the most important being:

  • Measurements were mostly done within coordinated campaigns; suitable observations over longer periods are relatively scarce.
  • Model studies were done either using high-resolution limited-area models over short time periods, or (too) coarse-resolution global model output over periods of a few years at most, often applying the isentropic assumption. No studies used high-resolution global data over long time periods.
  • There are two approaches to study STE in midlatitudes: the application of the "global" downward control principle, which describes transport at higher levels in the stratosphere, and the consideration of small-scale processes in the tropopause region (which finally lead to observable exchange events). These approaches have not yet been merged into one quantitative picture of STE.
  • Practical implementations of Wei´s formula, the most popular tool to determine STE, often yield unreliable results.
  • There has been little consideration of the depth and irreversibility of STE events. For atmospheric chemistry deep exchange events are more significant than shallow ones, since they influence not only the tropopause region, but also the surface and higher levels in the stratosphere. And irreversibility (i.e. mixing of an air parcel with air from the other sphere before being transferred back to its origin) is important for an STE event´s impact on the atmosphere´s chemical composition.
  • The tropopause region was poorly resolved in many chemical models studying the impact of STE on atmospheric chemistry.
  • Contour advection techniques were successful in producing fine-scale tracer structures. However, there is a lack of understanding on how different air masses mix, once they are brought together by STE processes, and how that impacts on non-linear chemistry.


Within STACCATO, we want to make important advances beyond the state of the art, consequently applying more accurate models and methods and conducting long-term measurements and model simulations. The most important steps forward are:


 1. First continuous monitoring of 7-Be and 10-Be at two sites
Radionuclides, for instance 7-Be, have often been used to diagnose STE. A major problem, however, is that 7-Be attaches to aerosols which are susceptible to washout. Taking the ratio of the two radionuclides 7-Be and 10-Be, which are equally affected by washout, removes this difficulty and gives a much improved estimate of STE. However, 10-Be measurements need accelerator mass spectroscopy, and thus few data are available globally. We plan to continuously monitor 10-Be at two mountain sites with a time resolution of 1 week, and with higher time resolution also at two other sites during a few interesting periods. A number of stratospheric samples will also be analyzed. These activities will be supported by long-term high-resolution lidar monitoring and ozonesounding campaigns.


2. Development of a new three-dimensional Lagrangian perspective of STE
To overcome difficulties with Eulerian methods to diagnose STE, we will use Lagrangian methods to create histories of air parcels crossing the tropopause region. First steps in this direction were the introduction of concepts like the age (or age spectrum) of air. However, what we aim at is to record detailed statistics of the paths of air masses and the physical and chemical environments encountered. Special focus will be on the important deep exchange events.


 3. First systematic intercomparison of different models (coarse-resolution climate models, high-resolution weather prediction model, limited area model) and diagnostics (Lagrangian, Eulerian, on-line tracers) for STE, and their evaluation with measurement data
At least seven different models and methods based on different datasets will be used within STACCATO to diagnose STE. We will make a systematic intercomparison of these methods and evaluate them with measurement data. The climate models involved, which usually do not use analyzed meteorological data, will be nudged towards analyses for this exercise.


 4. First comprehensive study of STE on climatic timescales in the past, present time and future
The improved diagnostics will be used to study past (for at least the last 15 years, some diagnostics for the last 40 years) and present STE based on high-resolution (1°) meteorological re-analyses and analyses. Future scenarios will be studied based on model predictions with two climate models (one of them a high-resolution middle-atmosphere version).


 5. Study of the influence of STE on the oxidation capacity of the troposphere; study of the relevance of mixing processes for atmospheric chemistry
The influence of STE on tropospheric chemistry will be studied with a Eulerian chemistry-climate model and with an off-line fully Lagrangian global chemistry model. The processes leading to mixing of different air masses and their effect on tropospheric and stratospheric chemistry will be studied using idealized meteorological and chemical models.


 6. First study of the effect of changes in STE due to a changing climate on atmospheric chemistry
The above chemistry models, applied for scenarios of climate change, will be used to identify possible effects of changes in STE on tropospheric chemistry.