Dynamic Couplings and their Impact on Climate and Weather

Stratospheric circulation influences near-surface regional climate and seasonal weather via dynamical couplings with the troposphere. Improved climate and weather predictions therefore require a realistic model representation of the stratospheric circulation response to anthropogenic and natural climate change. Prominent stratospheric circulation patterns in this context are the circumpolar polar jet and variations of its strength, the tropical quasi-biennial oscillation (QBO) of stratospheric tropical winds and its teleconnections into the extra-tropical troposphere, as well as the variability of strength of the meridional Brewer-Dobson circulation.

One important example of the dynamic coupling of the atmosphere with the near-surface climate is the stratospheric quasi-biennial oscillation (QBO). According to recent analyses, the QBO makes a difference of up to 2 °C to winter temperatures in Europe, for example. Ern and Preusse [1] have shown that small-scale gravity waves contribute most of the QBO forcing (approx. 70 %).
Atmospheric waves play a decisive role in atmospheric couplings (e.g. due to their interaction with the average wind field). In particular, gravity waves with a horizontal length of a few tens of kilometres to several hundred kilometres are too short to be directly resolved by climate and weather models. The effect of gravity waves is therefore allowed for by greatly simplified parametrizations. However, more and more experts regard these simplifications in the gravity wave parametrizations as clearly too imprecise. Various efforts are therefore being undertaken to make gravity wave parametrization more realistic based on global measurements of the gravity wave momentum flux. This can be achieved by restricting the freely adjustable parameters by means of measurements, for example. IEK-7 has carried out pioneering work in this field on the basis of satellite observations (CRISTA, SABER, HIRDLS, AIRS). These studies play an important role in the “gravity wave initiative” – Stratospheric Processes And their Role in Climate (SPARC) – which is part of the World Climate Research Programme (WCRP). The European Centre for Medium-Range Weather Forecasts (ECMWF) has integrated a parametrization for non-orographic gravity waves into their predictive model in order to improve weather forecasts [2]. This parametrization is based on an IEK-7 analysis of CRISTA satellite observations [3]. Using a gravity wave parametrization optimized in this study, the ECMWF model now represents global circulation patterns more realistically. The future improvement of gravity wave parametrization in chemistry climate models (CCMs) will mean that the predictive value of current climate models will take on a whole new quality.

Interaction of large-scale circulation patterns with gravity-waves (H): By considering the total momentum balance from the large scales (meteorological analysis) together with the GW drag values derived from satellite temperature observations (SABER, HIRDLS), the relative role of GWs for driving global atmospheric wind patterns such as the quasi-two-day planetary waves [7], sudden stratospheric warmings (SSWs) [8] and the tropical semiannual oscillation (SAO) [9] were investigated by our group.
A particular interesting example is the quantification of gravity-wave driving of the QBO [10].The QBO couples to surface climate and can thus be inferred to exist for at least a century. The QBO arises from a balance of drag exerted by dissipating waves, which tend to pull the QBO phase downward, and of advection in the Brewer-Dobson circulation, which tends to advect the QBO phase upward. The momentum balance is presented in Figure 1.2.7. Closure of the momentum balance from the ERA-Interim reanalysis requires additional drag, which is in good agreement with gravity-wave drag derived from SABER and HIRDLS (see Figure, lower panel). Deviations are seen in periods of westward shear. Potential explanations include problems in the ERA momentum-balance (either by incorrect vertical winds or resolved waves such as Mixed-Rossby-GWs) or a stronger activity of short horizontal wavelength GWs not observable by HIRDLS / SABER. According to our study [10] mesoscale GWs provide more than 50 % of the driving of the QBO.
Gravity waves resolved in numerical weather prediction (NWP) data: IEK-7 developed methods to derive GWMF from 3D temperature fields to be observed by proposed satellite-borne infrared limb-imagers [11] or infrared nadir sounder such as AIRS [12].These methods can also be used to analyze gravity-waves explicitly resolved by ECMWF. While GWs excited in the mid- and high latitudes by orography and instabilities in the tropospheric jets compare favorably to observations from, for example, limb sounders, there are deviations in the tropics and subtropics. At these latitudes deep convection is the most important source [13]. Compared to observed spectra from satellite data the GWs resolved in the ECMWF model develop less pronounced hot spots of GWMF, have slower phase speeds and are generally weaker. The reason presumably is that convection needs to be parameterized at the resolution achievable by a global model. The parametrization, however, does not fully feed back to the dynamical core, hence the differences in wave excitation. At mid and high latitudes where ECMWF-resolved GWs are largely realistic, the ECMWF data allow the short-term variability of GWMF to be investigated. It is found that even a hemispheric mean value may vary by more than a factor of 3 on a time scale of 1 day [11]. A further advantage of identifying localized wave packets is that the results can be used for the initialization of GW ray tracing. In this way, GW sources can be identified, although they may be thousands of kilometers distant from a stratospheric observation location.

Improved GW parametrizations (H): The fact that convective GWs are underrepresented in the ECMWF model demonstrates that increasing resolution does not guarantee correct representation of GWs. Therefore, even for future high-resolution models, part of the GW spectrum will need to be represented by GW parametrization schemes, in particular at very short scales. Most climate and NWP models employ GW parameterizations which subsume all sources except orography in a non-orographic GW parametrization. This method, however, has the disadvantage that it feeds back to the actual meteorological situation and the actual climate state only via the interaction with the background wind, but not via the source! In order to overcome this limitation, we need physics-based parametrizations of GW sources. One of the most important sources is convection. A convective GW source scheme (CGWS) was developed at Yonsei University (Seoul, Korea). However, such a CGWS needs to rely on simplified physics. In particular, the temporal and spatial scales of the convection are free parameters in the Yonsei CGWS. At IEK-7 we developed an observational filter to simulate the effects of the observational technique of HIRDLS and SABER [14]. By varying the free parameters of the CGWS and comparing the model results to GW spectra from HIRDLS, we could determine these free parameters of the CGWS for subtropical convection [15]. The results are in good agreement with direct observations of the size of mesoscale convective complexes (MCCs), frequently occurring in the summer subtropics. Together with a general background, most of the global variations of GWMF can be successfully described [16]. Our modelling efforts include oblique propagation of GWs, which is of importance for global momentum balances. On their way up the mesopause, GWs propagate 10° to 20° latitude poleward on average [17] and may thus avoid critical levels. The latter may also be important in rebuilding the polar vortex after sudden stratospheric warmings [8].

Satellite climatology: A climatology of gravity wave momentum flux from infrared limb sounder tempearture measurements is published in open access [18,19]

References:

1. Ern, M.et al. 2009, Geophysical Research Letters, 36, L21801, doi:10.1029/2009GL040493.
2. Orr, A. et al. 2010, Journal of Climate, 23, 5905 – 5926.
3. Ern, M. et al. 2006, Atmospheric Chemistry and Physics, 6 , 4361 – 4381.
4. Preusse, P. et al. 2014, Atmospheric chemistry and physics / Discussions 14(8), 11961 - 12018 (2014) [10.5194/acpd-14-11961-2014].
5. Preusse, P. et al 2009, Journal of Geophysical Research D: Atmospheres, 114, D08126.
6. Schröder, S. et al. 2009, Geophysical Research Letters, 36, L10805.
7. Ern, M. et al. 2013, J. Geophys. Res.-A. 118: 3467-3485, doi:10.1029/2012JD018208.
8. Ern, M. et al. 2016, Atmos. Chem. Phys. 16: (15), 9983-10019, doi:10.5194/acp-16-9983-2016.
9. Ern, M. et al.. 2015., Ann. Geophys. 33: 483-504, doi:10.5194/angeo-33-483-2015.
10. Ern, M. et al. 2014, J. Geophys. Res.-A. 119: 1-27, doi:10.1002/2013JD020731.
11. Preusse, P. et al. 2014, Atmos. Chem. Phys. 14: 10483-10508, doi:10.5194/acp-14-10483-2014.
12. Ern, M. et al. 2016, Geophys. Res. Lett. 43: n/a, doi:10.1002/2016GL072007.
13. Jia, J.Y. et al. 2014, Ann. Geophys. 32: 1373-1394, doi:10.5194/angeo-32-1373-2014.
14. Trinh, T. et al. 2015, Atmos Meas. Tech. 8: 1491-1517, doi:10.5194/amt-8-1491-2015.
15. Trinh, Q.T. et al. 2016, Atmos. Chem. Phys. 16: (11), 7335-7356, doi:10.5194/acp-16-7335-2016.
16. Kalisch, S. et al. 2016, J. Geophys. Res.-A. 121: (22), 13474-13492, doi:10.1002/2016JD025235.
17. Kalisch, S. et al. 2014, J. Geophys. Res.-A. 119: 10,081-10,099, doi:10.1002/2014JD021779.
18. Ern, M. et. al. 2017, PANGAEA.Data Publisher for Earth & Environmental Science
19. Ern, M. et. al. 2018, Earth Syst. Sci. Data