Plan for S-RIP chapter on QBO and tropical variability
The scientific goal of the chapter is to evaluate and intercompare how different global reanalyses represent the quasi-biennial oscillation (QBO) and tropical stratospheric variability. Just as climate model intercomparisons are motivated by the large number of available climate models, the number of global reanalyses that are presently available (eight) or soon to be available (four) motivates an intercomparison of these datasets.
Reanalyses provide a best-guess estimate of the observed atmospheric state, yet differences between them exist due to the choice of observations that are assimilated, the assimilation method that is employed, and the characteristics of the forecast model. Characterizing these differences and understanding their origins is essential in order to make best use of reanalysis datasets, especially so that users can be warned away from inappropriate uses of them.
Specifically in the case of the QBO, the sparsity of tropical radiosonde wind observations, and the difficulty experienced by most atmospheric general circulation models (AGCMs) in representing the QBO, suggest that a careful assessment of the fidelity of reanalyses QBOs would be useful. Put simply: how well do we know the state of the QBO?
2. Overview of scientific issues
2.1 How well observed is the QBO?
We will characterize inter-reanalysis differences in the structure of the QBO in zonal-mean tropical winds and temperatures. Diagnostics will address the vertical and meridional structure of the QBO, its amplitude and period, and its partial synchronization with the annual cycle. Modellers trying to represent the QBO in free-running AGCMs may tune their models to match reanalyses, so it is of interest to determine how well constrained are the reanalysis QBOs. Direct comparison with observations will be made where possible. In particular we expect to use IGRA and IGRA2 tropical radiosondes, and SABER and HIRDLS satellite data.
2.2 Zonal asymmetry
Due to the sparse distribution of near-equatorial radiosonde stations, it is usually assumed that the FUB monthly-mean "Singapore winds" (http://www.geo.fu-berlin.de/en/met/ag/strat/produkte/qbo/) are representative of the zonal-mean QBO. This is a good assumption if zonal asymmetries (in the monthly means) are genuinely small. The FUB time series is compiled from the consecutive records of three near-equatorial stations. The availability of more tropical radiosonde records, especially from the soon-to-be-available IGRA2, should allow a closer examination of possible zonal asymmetries of the QBO.
2.3 Diagnosing model errors
Free-running (i.e. without data assimilation) AGCMs often have problems exhibiting spontaneous QBOs, and the forecast models used to produce reanalyses are likely to share similar systematic model errors. Examining the QBO momentum budget in reanalyses may give insight that is applicable free-running AGCMs, since the analysis increments act to correct the forecast model errors.
2.4 Extratropical influence
The QBO is observed to influence the winter stratospheric polar vortices of both hemispheres. Other teleconnections also affect the polar vortex, such as ENSO and the 11-year solar cycle. To the extent that these influences co-vary over the course of the observed record, diagnostic separation of their effects is difficult and their interactions may be important. We will compare these teleconnections across reanalyses, examining their robustness and hoping to gain insight into their interactions. One important aspect involves detemining how robust are these teleconnections to the choice of extratropical metric defining the polar vortex state - e.g. whether one focuses on major sudden warmings, annular mode anomalies, temperature at the pole, etc. Since the Stratosphere-troposphere coupling (STC) chapter of S-RIP will be producing a variety of such metrics, close collaboration with the STC chapter will facilitate a comprehensive analysis of teleconnections.
2.5 Tropical dynamics
The behaviour of tropical waves in the reanalyses is of interest: since reanalysis QBOs tend to be closer to observed tropical winds than free-running model QBOs, tropical waves propagate through more realistic background wind shears in the reanalyses than they do in most climate models. Since wave dissipation forces the zonal-mean QBO, spectral analysis of tropical waves may give insight into the partitioning of QBO forcing amongst different wave types. Effects of the QBO may also extend downward below the region of the stratosphere in which the QBO is the dominant mode of variability. There is evidence that QBO modulation of the tropical tropopause height affects deep convection, suggesting a QBO feedback on the generation of the waves that force it. If robust, this behaviour may be evident in reanalyses. Comparison of tropical waves in reanalyses and observations, such as the SABER and HIRDLS satellites, is also of interest.
3. Organization and schedule
The long-term goal of S-RIP is to produce a SPARC report, due in 2018. For the QBO and tropical variability chapter, we intend for analyses to be paper-driven - i.e. contributors will lead and co-author papers published in the peer-reviewed literature, and results from these papers will contribute to the final S-RIP report. However, the S-RIP report is intended to be a comprehensive reference for users of reanalysis products, and hence it may be feasible to include a greater level of detail in the report than in the papers. Nevertheless we anticipate that the most interesting analyses are driven by clear science questions, hence papers. Submitting papers prior to the publication of the S-RIP report may also prevent any potential future problems with publishing material that is already published in the S-RIP report.
The schedule of the chapter is to produce at least the basic QBO characterizations within the first 1-2 years, following the schedule of S-RIP chapters 1-4. More complex diagnostics can take slightly longer, consistent with the schedule of the other "advanced" S-RIP chapters. We hope to keep pace with developments in the stratosphere-troposphere coupling (STC) chapter, since this facillitates analysis of teleconnections as described above. Short progress reports are planned for each year, but the main source of up-to-date information re. this chapter is this website and the S-RIP wiki on the main S-RIP website (http://s-rip.ees.hokudai.ac.jp/internal/wiki/).
New contributors to the chapter are very welcome! If you are interested please get in touch with the chapter co-leads, James and Lesley (contact info on main page). Please first check the list of current projects (given below) to see how you could:
(1) Contribute to a current project
(2) Initiate a new one (i.e. convince us that there's something important missing from the list of current projects)
Regarding (1), inclusion of new contributors is at the discretion of those already taking the lead on those projects. Regarding (2), inclusion of new projects within the QBO chapter is at the discretion of the chapter co-leads. We hope to be as flexible and open as possible, but of course some coordination of efforts is needed to ensure that a coherent final chapter can be written, and also that contributors don't unwittingly duplicate or intrude on others' work.
4. Potential links with other chapters
Links with the following other S-RIP chapters seem clear at this early stage. Other links may emerge as the S-RIP project progresses.
Stratosphere-troposphere coupling (STC) chapter: Diagnosis of teleconnections will employ extratropical metrics provided by the STC chapter contributors, enabling a more comprehensive analysis than would be otherwise feasible.
Tropical tropopause layer (TTL) chapter: QBO modulations of tropical tropopause height, tropical deep convection, and the water vapour tape recorder are interesting aspects of the QBO.
Upper stratosphere and lower mesosphere (USLM) chapter: Above the QBO-dominated altitudes of the tropical stratosphere is found a semi-annual oscillation (SAO) near the stratopause. Since waves that force the SAO must first propagate through QBO wind shears, interpretation of inter-reanalysis SAO differences will likely require QBO diagnostics. The structure of stratospheric polar vortices at very high altitudes may also be influenced by the QBO (or other teleconnections, e.g. ENSO).
5. Current projects
This list describes papers/projects currently in progress by the chapter contributors. The following info is given (if available):
1) Who is involved
2) What diagnostics are used
3) Timeline (roughly)
4) Summary of most interesting results, and/or publication status
As analyses evolve, it may become unclear which chapter of the S-RIP report best includes a given analysis - e.g., analysis of QBO teleconnections with extratropical surface climate may be better placed in the stratosphere-troposphere coupling chapter than the QBO chapter. This should become clearer as S-RIP progresses
[This list is in rough form. We're still gathering this info.]
Regression analysis of solar cycle and other teleconnections
Lesley Gray, Dann Mitchell
Paper submitted to QJRMS, May 2014
Regression analysis of volcanic and other teleconnections
Presented at SPARC GA (Jan 2014) and Japan Geoscience Union (JpGU) 2014 Meeting
Manuscript in preparation
Gravity wave forcing of the QBO using HIRDLS and SABER data
Comparison of HIRDLS and SABER to gravity waves in reanalyses
Corwin Wright, James Anstey
Combined QBO-solar cycle modulation of the NH polar vortex
Here also is an indication of the contributors' interests, organized by person rather than by project.
Regression analysis of teleconnections, especially solar cycle
Regression analysis of teleconnections, especially volcanoes
HIRDLS + SABER comparison with reanalyses and models
Holtan-Tan effect, including mechanisms and interference with the solar cycle
Comparison of IGRA zonal wind, in addition to FUB zonal wind, with reanalysis data
Volcanic teleconnections (and their relation to other teleconnections)
Basic QBO characterization, seasonal synchronization of the QBO, teleconnections (Holton-Tan effect)
Teleconnections (Labitzke-van Loon effect, etc)
Latest update: 31 Jul 2014 by James Anstey