1. Introduction
The final warming (FW) of the polar stratosphere marks the transition from winter to summer circulation conditions and occurs every spring. The FW is forced radiatively, but wave–mean flow interactions play an important role and control its interannual variability. In the Southern Hemisphere (SH), a number of climate models predict that it occurs 1–2 weeks later than in observations (Eyring et al. 2010; Butchart et al. 2011; Wilcox and Charlton-Perez 2013), this systematic error being sometimes accompanied by a cold temperature bias in winter and spring. These biases have important implications on the stratospheric dynamics and chemistry, like a systematic late seasonal ozone recovery over Antarctica that affects simulated long-term ozone trends and Antarctic climate evolution (Perlwitz et al. 2008; Shaw et al. 2011; Sun et al. 2014; Barnes et al. 2014).
The general consensus to explain these late FW biases is that climate models underestimate the gravity wave (GW) forcing in the southern stratosphere, particularly around 60°S (McLandress et al. 2012), but the orographic or nonorographic origin of the missing GW drag is still controversial. In present-day climate models, this is unavoidably related to how GW parameterizations are constructed. On the one hand, parameterized orographic gravity waves (OGWs) usually break in the troposphere and stratosphere (e.g., Palmer et al. 1986; Lott et al. 2005). This is in contrast to nonorographic gravity waves (NGWs), which are usually treated in the parameterizations as small-amplitude waves, breaking at higher altitudes in the mesosphere to drive the upper branch of the Brewer–Dobson circulation and not interacting directly with the stratospheric flow (e.g., Alexander et al. 2010). In the atmosphere, OGWs are present around 60°S; although the underlying surface is an ocean, there are contributions due to small islands of the Southern Ocean (Alexander et al. 2009; Alexander and Grimsdell 2013) and due to lateral propagation from the Andes and the Antarctic Peninsula (Sato et al. 2009, 2012; Hindley et al. 2015). However, small islands are absent or dwarfed because of poor resolution in climate models, and horizontal propagation is absent from nearly all parameterizations by construction, leading to a gap in parameterized OGW drag around 60°S. Although this gap is unphysical, this does not imply that OGWs are solely responsible for the missing GW drag (GWD) in climate models, as suggested by McLandress et al. (2012). Another possibility is to modify the current representation of NGWs in parameterizations as small perturbations that propagate all the way up to the mesosphere, which also seems unrealistic according to recent observational studies that emphasize the intermittent character of the GW momentum fluxes entering the SH stratosphere (Hertzog et al. 2008, 2012; Plougonven et al. 2013; Wright et al. 2013; Alexander 2015). This intermittency is absent from nearly all GW parameterizations, and, if taken into account, it could increase the NGWs’ contribution to the missing GWD more substantially than usually believed.
From the observational side, some studies using satellite-derived products have demonstrated that GWs generated by flow over the small southern islands can carry a significant amount of momentum flux (e.g., Alexander et al. 2009; Alexander and Grimsdell 2013), but the expected contribution to the global-scale forcing is presumably modest. Hindley et al. (2015) observed latitudinal propagation of GWs originated above the mountainous southern Andes and Antarctic Peninsula toward the jet stream and also showed a long region leeward from this GW hotspot containing waves of orographic origin advected downwind, confirming the modeling results of Sato et al. (2009, 2012). Other works have pointed out that the stratospheric GWs observed over the ocean surrounding Antarctica likely have nonorographic sources. Hendricks et al. (2014) studied the source of the stratospheric GW belt (at around 60°S) in austral winter. They found a strong correlation between GW activity and midtropospheric maximum Eady growth rate, suggesting a nonorographic origin of the GWs. In a recent study, Jewtoukoff et al. (2015) showed quantitative evidences from in situ balloon observations and high-resolution ECMWF operational analysis that the momentum flux around 60°S in the lower stratosphere in spring is dominated by GWs from nonorographic sources (see also Hertzog et al. 2008; Plougonven et al. 2013). Using observations of the first mesosphere–stratosphere–troposphere radar in Antarctica, Shibuya et al. (2015) also stress the important contribution of nonorographic GWs to the total momentum flux in the austral lower stratosphere.
The goal of the present study is to contribute to the debate by analyzing the wave forcing during the final warming of the southern stratosphere in climate simulations with a general circulation model [Laboratoire de Météorologie Dynamique zoom model (LMDZ)] and in reanalysis products. LMDZ includes state-of-the-art stochastic parameterizations of nonorographic GWs tied to their tropospheric sources (Lott and Guez 2013; de la Cámara and Lott 2015), which generate lognormally distributed momentum fluxes in agreement with observations (de la Cámara et al. 2014; Jewtoukoff et al. 2015), as well as orographic gravity waves (Lott 1999; Lott et al. 2005). As a result, we will show that the contribution to the total GW drag in the stratosphere of nonorographic GWs is larger than that reported in previous studies with different parameterizations, and no significant bias on the FW date is found in our model. We will also show that the ratio of OGW drag (OGWD) to NGW drag (NGWD) parameterized in LMDZ is qualitatively realistic as compared to the GW-resolving ECMWF operational analysis.
The paper is organized as follows. Section 2 presents the LMDZ, the method used to infer the GWD from reanalysis fields, and the calculation of GWD from the ECMWF operational analysis. In section 3, we analyze the wave forcing during the final warming of the SH, with emphasis on the unresolved waves, and investigate the role of parameterized GW intermittency. The main conclusions are given in section 4.
2. Model and methodology
a. LMDZ
The LMDZ version that we use has a 3.75° × 1.875° longitude–latitude grid, 71 levels in the vertical with the top at 0.01 hPa, and a vertical resolution of around 1 km in the lower stratosphere. We show results from a control run of 20 yr, forced with climatological fields of sea surface temperature, sea ice, soil temperature, and composition over land. One important aspect determining the timing of the SH final warming is the ozone forcing. In fact, the southern polar stratosphere has cooled over the last decades of the twentieth century as a result of anthropogenic ozone depletion, which has delayed the FW by about 10 days (decade)−1 (e.g., Waugh et al. 1999; Black and McDaniel 2007). We have used a monthly climatology of stratospheric ozone for the period 1997–2006, obtained from the Chemistry–Climate Model Validation (CCMVal) REF-B1 run with the LMDZ–Reactive Processes Ruling the Ozone Budget in the Stratosphere (REPROBUS) Chemistry Climate Model (Eyring et al. 2008; Jourdain et al. 2008). Therefore, the massive reduction in ozone concentration during the SH spring that has occurred since the 1980s is included in the climatological radiative ozone forcing introduced in the model.
LMDZ uses three distinct GWD parameterizations, representing GWs generated by subgrid-scale orography (Lott 1999), by convection (Lott and Guez 2013), and by fronts and jet imbalances (de la Cámara and Lott 2015). The last two are stochastic and supposed to cover all the GWs of nonorographic origins. De la Cámara et al. (2014) and de la Cámara and Lott (2015) showed that the combination of a stochastic approach and the relation with the sources produce lognormally distributed momentum fluxes: that is, including large, rare events that account for much of the mean value and that potentially break at lower altitudes in the stratosphere.
b. Inferring gravity wave drag from (re)analysis data
To calculate the GW contribution to the middle-atmosphere dynamics, we next use two datasets that are independent from LMDZ. The first is ERA-Interim (ERAI), and it is used to produce climatological quantities, which is consistent, since the same data assimilation system and model is used in the whole reanalysis period (Dee et al. 2011). The second is the ECMWF operational analysis, which is used to obtain information on the GW field itself because of its high, GW-permitting resolution.
1) Residual term in ERA-Interim
2) Explicit evaluation from the ECMWF operational analysis
The ECMWF model used to prepare operational analysis four times a day and to make weather predictions has a spectral truncation of T1279 and 91 vertical levels, corresponding to a vertical grid spacing of around 500 m in the free troposphere and stratosphere. At these resolutions, it is expected that a significant fraction of the GWs is resolved, and we know from Ern et al. (2008), Shutts and Vosper (2011), and Preusse et al. (2014) that the GWs in the ECMWF operational analysis fairly compare to those observed by satellites, a result confirmed with in situ superpressure balloons (Jewtoukoff et al. 2015). Relevant for our work, Jewtoukoff et al. (2015) also show that the spatial distribution and wave statistics of the GWs in the analysis are realistic.
3. Results
a. The austral stratospheric final warming in LMDZ
Figures 1a and 1b illustrate the austral stratospheric final warming in ERAI through a climatological average of the altitude–time evolution of the zonal-mean zonal wind at 70°–50°S and temperature over the polar cap during the southern winter and spring. The latitude range for the wind corresponds to the approximate location of the jet maximum throughout the season. The time average spans the years 1992–2011, so the ozone-forcing period used in LMDZ lies in the middle (1997–2006; see section 2). During the winter months, the wind is eastward, with a maximum of ~70 m s−1 in July around 1 hPa and below. Starting in late September and from the highest altitudes, the winds decelerate and change to the westward direction, signaling the transition from winter to summer circulation conditions. The black contour in Fig. 1a represents the zero-wind line, and the gray contour represents the 10 m s−1 wind line, and these illustrate very clearly that the transition from eastward to westward winds happens at mesospheric levels first (in October above 1 hPa) and at stratospheric levels later in the season (in early December at 10 hPa). The transition in temperature appears lower down and about a month earlier, with a warming of several tens of kelvins, in agreement with early studies of the FW (e.g., Mechoso et al. 1985). Figures 1c and 1d show similar plots but for LMDZ. The maximum of the winter jet (80 m s−1) is stronger than in ERAI, and the very low temperatures in the winter lower stratosphere slightly expand to lower levels. Apart from these differences, the evolution of the zonal wind and temperature simulated by LMDZ compares well with that reproduced in ERAI.
Time–height evolution of (a) zonal-mean zonal wind (m s−1) averaged over 70°–50°S and (b) temperature (K) averaged over 85°–60°S during the southern winter and spring for ERAI. (c),(d) As in (a) and (b), respectively, but for LMDZ.
Citation: Journal of the Atmospheric Sciences 73, 8; 10.1175/JAS-D-15-0377.1
Figure 2 gives a complementary view of the zonal wind evolution over spring, displaying the monthly zonal-mean zonal wind for September, October, and November in ERAI and LMDZ (top and middle rows). There is reasonable agreement between the two datasets, in particular the position of the zero-wind line. We also see slight differences, such as a stratospheric jet that is 5 m s−1 stronger in LMDZ than in ERAI between 100 and 10 hPa in October and November and a subtropical jet in LMDZ that is systematically 5 m s−1 stronger than in ERAI. The right-hand column of Fig. 2 shows the December–February seasonal average of the zonal-mean zonal wind. Apart from the persistent bias in the strength of the summer westward jet (Lott et al. 2005), the model performs reasonably well in the northern winter.
Latitude–height cross sections of zonal-mean zonal wind (m s−1) in (a)–(d) ERAI, (e)–(h) LMDZ, and (i)–(l) LMDZ-CS (see section 3d) for (a),(e),(i) September, (b),(f),(j) October, (c),(g),(k) November, and (d),(h),(l) December–February. The black thick lines denote the zero and ±50 m s−1 isotachs.
Citation: Journal of the Atmospheric Sciences 73, 8; 10.1175/JAS-D-15-0377.1
Following Black and McDaniel (2007), we use 5-day running means of daily data to calculate the final warming date as the final time that the zonal-mean zonal wind at 60°S drops below 10 m s−1 until the following autumn. Figure 3 shows the mean FW date as a function of pressure level for ERAI (red curve) and LMDZ (blue curve). As expected from Fig. 1, the mean FW dates in ERAI and LMDZ agree reasonably well, with differences of 2–5 days later in LMDZ throughout the stratosphere. The shaded areas cover plus or minus one standard deviation, again with fairly good agreement. These results reveal that LMDZ does a good job in simulating the final warming of the SH. Since the ozone forcing in our model comes from a 1997–2006 monthly climatology, a period with maximum ozone loss over Antarctica in spring, we believe that the good model performance is not due to an excessively strong ozone forcing. In the next sections, we analyze the resolved and unresolved wave drag during the FW, focusing on the role of nonorographic GW parameterizations.
Final warming dates as a function of pressure level in ERAI (1992–2011; red), LMDZ (blue), and LMDZ-CS (green; see section 3d). The climatological means are given by the solid lines, and the shaded areas represent plus or minus one standard deviation.
Citation: Journal of the Atmospheric Sciences 73, 8; 10.1175/JAS-D-15-0377.1
b. Resolved and unresolved wave forcing
Figure 4 shows latitude–height cross sections of monthly mean resolved wave drag [i.e., divergence of the EP flux (DF)] in ERAI and LMDZ, for October and November from the midstratosphere to the lower mesosphere. In October, the magnitude and the extent of the negative wave forcing in ERAI resembles that in LMDZ (Figs. 4a,c). The main difference appears over the pole higher than 0.3 hPa, where the positive values are larger in ERAI. This positive EP flux divergence arises in a region of very weak positive and negative winds (Figs. 2b,e), where the waves tend to be refracted away, resulting in positive divergence of the EP flux. The difference in magnitude could be partly due to a weaker vertical shear in ERAI than in LMDZ, which favors refraction. In November as well, the forcing is similar in both datasets, with stronger negative forcing in LMDZ than in ERAI in the stratosphere (below ~1 hPa).
Latitude–height cross sections of resolved wave drag for (a) October and (b) November for ERAI. (c),(d) As in (a) and (b), respectively, but for LMDZ. Contour interval is 2 m s−1 day−1, starting at ±1 m s−1 day−1.
Citation: Journal of the Atmospheric Sciences 73, 8; 10.1175/JAS-D-15-0377.1
Figure 5 shows the corresponding latitude–height cross sections of monthly mean 
As in Fig. 4, but for unresolved (parameterized) gravity waves.
Citation: Journal of the Atmospheric Sciences 73, 8; 10.1175/JAS-D-15-0377.1
It is interesting that, despite the different horizontal resolutions, ERAI having a horizontal spacing of about 80 km (T255 spectral truncation) and LMDZ having a horizontal spacing of about 200 km, the resolved and unresolved wave forcings have a similar order of magnitude and latitude distribution in both datasets. The reason is probably that the resolved forcing in the stratosphere in both ERAI and LMDZ essentially comes from planetary-scale Rossby waves (i.e., waves with scales that can be resolved in both models) and is consistent with the fact that the synoptic disturbances play a small role in the middle-atmosphere dynamics (Andrews et al. 1987).
c. Orographic and nonorographic gravity wave drag
We next analyze the relative contribution of the nonorographic and orographic GWD parameterizations to the total GW forcing in LMDZ. Figure 6 shows the profiles of OGWD and NGWD for October, focusing on stratospheric levels from 100 to 0.1 hPa. The OGWD reaches its maximum at around 1 hPa and presents a minimum around 60°S, consistent with the absence of topography in that latitude band and the columnar approximation made in parameterizations. In contrast, the NGWD increases with increasing height and has a latitudinal maximum at 60°S, possibly because of the location of the tropospheric sources (Hendricks et al. 2014) and the presence of the stratospheric jet. Overall, we find that the parameterized GW drag in the stratosphere during the austral spring is not clearly dominated by orographic GWs, and this differs from most climate models. For example, in a study of the SH cold pole and strong jet biases in the Canadian Middle Atmosphere Model, McLandress et al. (2012) showed that the OGWD was much stronger than the NGWD in their model and found that the mentioned biases were reduced when including an extra forcing at 60°S in the OGWD scheme.
Latitude–height cross sections in the stratosphere of (a) orographic and (b) nonorographic gravity wave drag (m s−1 day−1) in LMDZ for October.
Citation: Journal of the Atmospheric Sciences 73, 8; 10.1175/JAS-D-15-0377.1
Recent observations indicate that the GW momentum flux in the springtime lower stratosphere over the Southern Ocean is dominated by nonorographic GWs (e.g., Hendricks et al. 2014; Jewtoukoff et al. 2015; Shibuya et al. 2015), pointing to the potential importance of these waves in forcing the stratospheric circulation in the region. However, this cannot be verified observationally since the derivation of the GW drag from global measurements remains a big challenge (Geller et al. 2013; Alexander 2015). Thus, to address whether the balance between orographic and nonorographic GWD in LMDZ is consistent, we next compare the GW drag in LMDZ with the drag obtained from the resolved spectrum of GWs in the ECMWF operational model (see section 2). Figure 7 shows the corresponding plots for ECMWF operational analysis data. We simply apply a geographical mask to discern between orographic and nonorographic GWs: all the GWs placed over the green areas in Fig. 7c will be considered to be most likely of orographic origin, and those outside the green areas will be considered to be almost surely of nonorographic origin. Using a GW-resolving climate model, Sato et al. (2012) showed that OGWs originating from the Andes and Antarctic Peninsula propagate very far leeward of the topographic obstacles. To account for this effect, our orographic region extends downstream of obstacles, as in Plougonven et al. (2013). The OGWD in the analysis (Fig. 7a) does not go to zero around 60°S, unlike in LMDZ (Fig. 6a). This is clearly due to the fact that we consider as orographic GWs those detected above small islands and over a vast region leeward of the Andes and Antarctic Peninsula (Fig. 7c). Apart from this difference, the magnitude and vertical extension of OGWD and NGWD agree reasonably well with parameterized data in LMDZ. And importantly, the ratio of OGWD to NGWD is similar in both datasets.
(a),(b) As in Fig. 6, but for resolved gravity waves in the ECMWF operational analysis (2006–10 period). Data are not displayed above 1 hPa (hatched area). (c) Map showing the continental mask (green) used to discriminate orographic and nonorographic GWs in the ECMWF operational analysis data.
Citation: Journal of the Atmospheric Sciences 73, 8; 10.1175/JAS-D-15-0377.1
d. The role of gravity wave intermittency
To clarify the significance of gravity wave intermittency, we next make offline tests using October daily fields from LMDZ and test different configurations of the NGWD schemes. Figure 8a presents the NGW drag averaged in time and longitude. It compares well with the online runs in Fig. 6b, demonstrating the potential of the offline calculations. First, these offline runs allow us to estimate the intermittency of the momentum flux predicted by our schemes, as Fig. 8b illustrates by showing the probability density functions of NGW absolute momentum flux at different levels in the stratosphere south of 40°S (Fig. 8b). As highlighted by de la Cámara et al. (2014) and de la Cámara and Lott (2015), the NGW sources included in these stochastic parameterizations naturally generate lognormally distributed momentum fluxes in agreement with observations (Hertzog et al. 2012; Jewtoukoff et al. 2015; Alexander 2015). We see in Fig. 8b that the larger, less frequent momentum fluxes are filtered out throughout the stratosphere and therefore are responsible for the NGW drag at stratospheric levels.
Gravity wave diagnostics produced offline using LMDZ fields for a given October: (a) nonorographic gravity wave drag (m s−1 day−1) and (b) probability density functions (histogram style) of NGW absolute momentum fluxes in the latitude band 90°–40°S at different altitudes.
Citation: Journal of the Atmospheric Sciences 73, 8; 10.1175/JAS-D-15-0377.1
To reveal more precisely the significance of intermittency, we next run our NGWs parameterization imposing a constant flux at the launching altitude, this constant flux being the averaged flux amplitude emitted when the sources are explicit. This corresponds to a value near 3 mPa, and the corresponding drag due to the westward component of the GW stress is shown in Fig. 9, where the bottom panels show the drag multiplied by a normalized density to highlight the values at stratospheric levels. The westward drag produced when considering a fixed emitted stress of 3 mPa is smaller in the stratosphere and larger in the mesosphere than when considering NGW sources (Figs. 9a–d).
Westward nonorographic gravity wave drag (m s−1 day−1) derived offline using (a) GW sources, (b) a fixed emitted stress of 3 mPa, and (c) a fixed emitted stress of 1.25 mPa. (d)–(f) To emphasize the drag at stratospheric levels, the drag scaled by a normalized density is displayed.
Citation: Journal of the Atmospheric Sciences 73, 8; 10.1175/JAS-D-15-0377.1
The differences at mesospheric levels are important. As mentioned in the introduction, NGW parameterizations were introduced in climate models to be active at high altitudes in order to close the mesospheric jets and to contribute to the upper branch of the Brewer–Dobson circulation. Therefore, as our model with source-related NGWs is quite realistic, it is likely that one should reduce the imposed fixed stress to reach comparable results online with fixed stress. This is therefore what is done in Figs. 9c and 9f, which show the westward wave drag for an offline run reducing the emitted fixed stress to 1.25 mPa. We obtain now a reasonable drag above 50 km, but at the cost of reducing significantly the drag in the stratosphere.
These results suggest that with schemes imposing fixed NGWs sources it will be difficult to predict the stratospheric GW drag requested to simulate the annual cycle of the westerly jet without altering the mesosphere. To further prove this point, we next run the model for 20 yr, suppressing the dependence to the sources intensity in the NGW schemes and fixing a constant momentum flux at the level of emission of 1 mPa. This run will be referred to as LMDZ-CS. The bottom panels in Fig. 2 show the zonal-mean zonal wind cross sections for different months in the SH spring and also a seasonal average for December–February. It is interesting to note that the northern winter winds are very similar to those in the reference LMDZ run (Fig. 2h) (the absence of an internally generated quasi-biennial oscillation in LMDZ-CS explains the unrealistic stratospheric winds in the tropics). However, the austral spring jet is much stronger in LMDZ-CS, especially in September and October, when the differences are larger than 20 m s−1. This figure points out that turning off the momentum flux intermittency in the NGW schemes has practically no effect in the NH winter, where strong planetary waves control the mean winds and variability, but it does have a negative impact on the SH spring winds, where GWs significantly contribute to the momentum forcing. Consistently, the mean final warming date in the LMDZ-CS run occurs around 10–15 days later than in LMDZ (Fig. 3; green line), revealing the dramatic impact of the NGW parameterization during the austral spring in our runs.
Figure 10 shows the resolved momentum forcing as well as the parameterized orographic and nonorographic GW drag for October in LMDZ-CS. While the orographic GW drag is practically identical in LMDZ-CS and LMDZ (Fig. 10b and 6a, respectively), we do see some interesting differences in the resolved wave forcing and nonorographic GW drag between LMDZ-CS and LMDZ. Regarding the NGW drag, in LMDZ-CS it is stronger than in LMDZ above 1 hPa and weaker below 1 hPa (see Figs. 10c and 6b). In particular, note that the drag at 30–10 hPa and 60°S is positive in LMDZ-CS (accelerating the eastward winds) and negative in LMDZ (decelerating the eastward winds). This is what one should expect: the run with intermittency in the NGW scheme is more efficient at exerting a decelerating drag in the lower levels of the middle atmosphere. The resolved wave forcing in LMDZ-CS (Fig. 10a) has the same vertical and latitudinal structure as in LMDZ (Fig. 4c), but it is almost twice as strong around 1 hPa and 60°S (and, importantly, twice as strong as in ERAI). Also, the drag is slightly weaker around 10 hPa and 60°S in LMDZ-CS. Since the only change between the two runs is the launched stress specification in the NGW schemes, this means that the planetary waves respond by trying to compensate the differential GW forcing (Lott et al. 2005; Cohen et al. 2013; Scheffler and Pulido 2015). It is not our purpose to evaluate the full response in LMDZ-CS, which would imply evaluating the residual-mean circulation; nonetheless, we clearly demonstrate that turning off the highly intermittent momentum fluxes at the launching level in the NGW parameterization, while keeping a reasonable northern winter climatology, fails at producing the momentum forcing needed for a good austral FW representation.
Latitude–height cross sections of (a) resolved wave forcing, (b) orographic gravity drag, and (c) nonorographic gravity wave drag (m s−1 day−1) in LMDZ-CS for October.
Citation: Journal of the Atmospheric Sciences 73, 8; 10.1175/JAS-D-15-0377.1
4. Summary and conclusions
Insufficient parameterized GW drag around 60°S is likely causing the delay in springtime breakdown of the austral polar vortex in a number of climate models (e.g., Wilcox and Charlton-Perez 2013). Yet, there is not a clear consensus on the origin of the “missing” GW drag (orographic vs nonorographic) (e.g., McLandress et al. 2012). Recent observational studies stress the significant contribution of nonorographic GWs to the total momentum flux in the lower stratosphere and highlight their intermittent behavior (e.g., Alexander 2015; Hertzog et al. 2012; Wright et al. 2013; Hendricks et al. 2014; Jewtoukoff et al. 2015; Shibuya et al. 2015). This intermittency decisively determines the altitude at which the waves break and is generally not modeled in NGW parameterizations.
We have shown that LMDZ does not present a significant delay of the stratospheric vortex breakdown and, consequently, can be used to analyze the wave forcing during the austral stratospheric final warming. We have found a good agreement in the zonal drag exerted by resolved and unresolved waves between LMDZ and ERAI. In LMDZ, the unresolved forcing comprises the parameterized GW drag (i.e., orographic, convective, and frontal GWs), while in ERAI it has been derived from the momentum balance in the transformed Eulerian-mean formalism.
Unlike in many climate models, where orographic GWs play a dominant role at stratospheric levels, the parameterized GW drag in LMDZ during the austral final warming is not larger for waves of orographic origin than for those of nonorographic origin. Furthermore, while the OGW drag presents a minimum at 60°S, the NGW drag presents a maximum at this latitude possibly related to the location of tropospheric sources (e.g., baroclinic activity and fronts) and favorable propagation conditions in the jet stream. Therefore, in LMDZ, nonorographic GWs make a significant contribution to the total wave forcing during the austral final warming. We have demonstrated that this significant contribution of NGWs at stratospheric levels is due to a qualitatively realistic representation of momentum flux intermittency in the NGW parameterizations used. The stochastic scheme, tied to convective and frontal GW sources (Lott and Guez 2013; de la Cámara and Lott 2015), naturally produces sporadic, high-amplitude GWs that tend to break and force the circulation at lower levels in the stratosphere. At the same time, the bulk of waves carrying small momentum flux produces a drag in the mesosphere that keeps simulated winds and temperature at those altitudes within reasonable limits.
Using resolved gravity waves from the high-resolution ECMWF operational analysis, we have shown that the balance between orographic and nonorographic GW drag is similar to the drag parameterized in LMDZ, which provides a physical justification for a fair representation of momentum flux intermittency in nonorographic GW parameterizations. We know that the ECMWF operational analysis underestimates by a factor of 5 the resolved GW momentum fluxes entering the stratosphere when compared to direct balloon measurements (Jewtoukoff et al. 2015). Although we have shown that the introduction of intermittency permits us to increase substantially the GW fluxes entering the model stratosphere without degrading the mesosphere, these quite-large observed values indicate that much more still needs to be understood concerning the drag exerted in the models’ stratosphere at lower levels. Also, we must not forget that the necessary simplifications made in parameterizations, such as instantaneous and purely vertical propagation, could be missing some fundamental dynamics that might explain the large quantitative deviations between the observed absolute momentum fluxes in the lower stratosphere (Jewtoukoff et al. 2015; Alexander 2015) and the parameterized values (see Fig. 8b).
Finally, our results do not rule out the potential role of misrepresented orographic GWs due to the absence of lateral propagation in the parameterizations (Kalisch et al. 2014). We rather argue that it is not the only cause of the GW drag deficit and that the missing drag in models can be, to a great extent, due to nonorographic GWs. Also, we argue that improving the NGW parameterizations by relating quantitatively the GW amplitudes to the resolved dynamics of the model help to simulate better the Antarctic stratospheric final warming.
Acknowledgments
We thank Anne Smith for valuable comments on an early version of the manuscript and Marta Abalos for kindly providing some of the ERA-Interim diagnostics. We also thank two anonymous reviewers for the constructive comments that helped improve the manuscript. This study was supported by the French ANR project Stratospheric Dynamics and Variability (Stradyvarius; ANR-13-BS06-0011-01) and the European project ARISE2 (Horizon 2020; GAN653980). AdlC has been partially supported by the Advanced Study Program at NCAR.
REFERENCES
Alexander, M. J., 2015: Global and seasonal variations in three-dimensional gravity wave momentum flux from satellite limb-sounding temperatures. Geophys. Res. Lett., 42, 6860–6867, doi:10.1002/2015GL065234.
Alexander, M. J., and K. H. Rosenlof, 2003: Gravity-wave forcing in the stratosphere: Observational constraints from the upper atmosphere research satellite and implications for parameterization in global models. J. Geophys. Res., 108, 4597, doi:10.1029/2003JD003373.
Alexander, M. J., and A. W. Grimsdell, 2013: Seasonal cycle of orographic gravity wave occurrence above small islands in the Southern Hemisphere: Implications for effects on the general circulation. J. Geophys. Res. Atmos., 118, 11 589–11 599, doi:10.1002/2013JD020526.
Alexander, M. J., S. D. Eckermann, D. Broutman, and J. Ma, 2009: Momentum flux estimates for South Georgia Island mountain waves in the stratosphere observed via satellite. Geophys. Res. Lett., 36, L12816, doi:10.1029/GL038587.
Alexander, M. J., and Coauthors, 2010: Recent developments in gravity-wave effects in climate models and the global distribution of gravity-wave momentum flux from observations and models. Quart. J. Roy. Meteor. Soc., 136, 1103–1124, doi:10.1002/qj.637.
Andrews, D. G., J. R. Holton, and C. B. Leovy, 1987: Middle Atmosphere Dynamics.International Geophysics Series, Vol. 40, Academic Press, 489 pp.
Barnes, E. A., N. W. Barnes, and L. M. Polvani, 2014: Delayed Southern Hemisphere climate change induced by stratospheric ozone recovery, as projected by the CMIP5 models. J. Climate, 27, 852–867, doi:10.1175/JCLI-D-13-00246.1.
Black, R. W., and B. A. McDaniel, 2007: Interannual variability in the Southern Hemisphere circulation organized by stratospheric final warming events. J. Atmos. Sci., 64, 2968–2974, doi:10.1175/JAS3979.1.
Butchart, N., and Coauthors, 2011: Multimodel climate and variability of the stratosphere. J. Geophys. Res., 116, D05102, doi:10.1029/2010JD014995.
Cohen, N. Y., E. P. Gerber, and O. Bühler, 2013: Compensation between resolved and unresolved wave driving in the stratosphere: Implications for downward control. J. Atmos. Sci., 70, 3780–3798, doi:10.1175/JAS-D-12-0346.1.
de la Cámara, A., and F. Lott, 2015: A stochastic parameterization of the gravity waves emitted by fronts and jets. Geophys. Res. Lett., 42, 2071–2078, doi:10.1002/2015GL063298.
de la Cámara, A., F. Lott, and A. Hertzog, 2014: Intermittency in a stochastic parameterization of nonorographic gravity waves. J. Geophys. Res. Atmos., 119, 11 905–11 919, doi:10.1002/2014JD022002.
Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553–597, doi:10.1002/qj.828.
Ern, M., P. Preusse, M. Krebsbach, M. G. Mlynczak, and J. M. Russell III, 2008: Equatorial wave analysis from SABER and ECMWF temperatures. Atmos. Chem. Phys., 8, 845–869, doi:10.5194/acp-8-845-2008.
Ern, M., and Coauthors, 2014: Interaction of gravity waves with the QBO: A satellite perspective. J. Geophys. Res. Atmos., 119, 2329–2355, doi:10.1002/2013JD020731.
Eyring, V., and Coauthors, 2008: Overview of the new CCMVal reference and sensitivity simulations in support of upcoming ozone and climate assessments and the planned SPARC CCMVal report. SPARC Newsletter, No. 30, SPARC International Project Office, Toronto, ON, Canada, 20–26.
Eyring, V., T. G. Shepherd, and D. W. Waugh, 2010: SPARC report on the evaluation of chemistry–climate models. SPARC Tech. Rep. 5, 425 pp.
Geller, M. A., and Coauthors, 2013: A comparison between gravity wave momentum fluxes in observations and climate models. J. Climate, 26, 6383–6405, doi:10.1175/JCLI-D-12-00545.1.
Hendricks, E. A., J. D. Doyle, S. D. Eckermann, Q. Jiang, and P. A. Reinecke, 2014: What is the source of the stratospheric gravity wave belt in austral winter? J. Atmos. Sci., 71, 1583–1592, doi:10.1175/JAS-D-13-0332.1.
Hertzog, A., G. Boccara, R. A. Vincent, F. Vial, and P. Cocquerez, 2008: Estimation of gravity wave momentum flux and phase speeds from quasi-Lagrangian stratospheric balloon flights. Part II: Results from the Vorcore campaign in Antarctica. J. Atmos. Sci., 65, 3056–3070, doi:10.1175/2008JAS2710.1.
Hertzog, A., M. J. Alexander, and R. Plougonven, 2012: On the intermittency of gravity wave momentum flux in the stratosphere. J. Atmos. Sci., 69, 3433–3448, doi:10.1175/JAS-D-12-09.1.
Hindley, N. P., C. J. Wright, N. D. Smith, and N. J. Mitchell, 2015: The southern stratospheric gravity wave hot spot: Individual waves and their momentum fluxes measured by COSMIC GPS-RO. Atmos. Chem. Phys., 15, 7797–7818, doi:10.5194/acp-15-7797-2015.
Jewtoukoff, V., A. Hertzog, R. Plougonven, A. de la Cámara, and F. Lott, 2015: Gravity waves in the Southern Hemisphere derived from balloon observations and the ECMWF analyses. J. Atmos. Sci., 72, 3449–3468, doi:10.1175/JAS-D-14-0324.1.
Jourdain, L., S. Bekki, F. Lott, and F. Lefevre, 2008: The coupled chemistry–climate model LMDz–REPROBUS: Description and evaluation of a transient simulation of the period 1980–1999. Ann. Geophys., 26, 1391–1413, doi:10.5194/angeo-26-1391-2008.
Kalisch, S., P. Preusse, M. Ern, S. D. Eckermann, and M. Riese, 2014: Differences in gravity wave drag between realistic oblique and assumed vertical propagation. J. Geophys. Res. Atmos., 119, 10 081–10 099, doi:10.1002/2014JD021779.
Lott, F., 1999: Alleviation of stationary biases in a GCM through a mountain drag parameterization scheme and a simple representation of mountain lift forces. Mon. Wea. Rev., 127, 788–801, doi:10.1175/1520-0493(1999)127<0788:AOSBIA>2.0.CO;2.
Lott, F., and L. Guez, 2013: A stochastic parameterization of the gravity waves due to convection and its impact on the equatorial stratosphere. J. Geophys. Res. Atmos., 118, 8897–8909, doi:10.1002/jgrd.50705.
Lott, F., L. Fairhead, F. Hourdin, and P. Levan, 2005: The stratospheric version of LMDz: Dynamical climatologies, Arctic Oscillation, and impact on the surface climate. Climate Dyn., 25, 851–868, doi:10.1007/s00382-005-0064-x.
McLandress, C., T. G. Shepherd, S. Polavarapu, and S. R. Beagley, 2012: Is missing orographic gravity wave drag near 60°S the cause of the stratospheric zonal wind biases in chemistry–climate models? J. Atmos. Sci., 69, 802–818, doi:10.1175/JAS-D-11-0159.1.
Mechoso, C. R., D. L. Hartmann, and J. D. Farrara, 1985: Climatology and interanual variability of wave, mean-flow interaction in the Southern Hemisphere. J. Atmos. Sci., 42, 2189–2206, doi:10.1175/1520-0469(1985)042<2189:CAIVOW>2.0.CO;2.
Palmer, T. N., G. J. Shutts, and R. Swinbank, 1986: Alleviation of a systematic westerly bias in general circulation and numerical weather prediction models through an orographic gravity wave drag parametrization. Quart. J. Roy. Meteor. Soc., 112, 1001–1039, doi:10.1002/qj.49711247406.
Perlwitz, J., S. Pawson, R. L. Fogt, J. E. Nielsen, and W. D. Neff, 2008: Impact of stratospheric ozone hole recovery on Antarctic climate. Geophys. Res. Lett., 35, L08714, doi:10.1029/2008GL033317.
Plougonven, R., A. Hertzog, and L. Guez, 2013: Gravity waves over Antarctica and the Southern Ocean: Consistent momentum fluxes in mesoscale simulations and stratospheric balloon observations. Quart. J. Roy. Meteor. Soc., 139, 101–118, doi:10.1002/qj.1965.
Preusse, P., M. Ern, P. Bechtold, S. D. Eckermann, S. Kalisch, Q. T. Trinh, and M. Riese, 2014: Characteristics of gravity waves resolved by ECMWF. Atmos. Chem. Phys., 14, 10 483–10 508, doi:10.5194/acp-14-10483-2014.
Sato, K., S. Watanabe, Y. Kawatani, Y. Tomikawa, K. Miyazaki, and M. Takahashi, 2009: On the origins of mesospheric gravity waves. Geophys. Res. Lett., 36, L19801, doi:10.1029/2009GL039908.
Sato, K., S. Tateno, S. Watanabe, and Y. Kawatani, 2012: Gravity wave characteristics in the Southern Hemisphere revealed by a high-resolution middle-atmosphere general circulation model. J. Atmos. Sci., 69, 1378–1396, doi:10.1175/JAS-D-11-0101.1.
Scheffler, G., and M. Pulido, 2015: Compensation between resolved and unresolved wave drag in the stratospheric final warmings of the Southern Hemisphere. J. Atmos. Sci., 72, 4393–4411, doi:10.1175/JAS-D-14-0270.1.
Shaw, T. A., J. Perlwitz, N. Harnik, P. A. Newman, and S. Pawson, 2011: The impact of stratospheric ozone changes on downward wave coupling in the Southern Hemisphere. J. Climate, 24, 4210–4229, doi:10.1175/2011JCLI4170.1.
Shibuya, R., K. Sato, Y. Tomikawa, M. Tsutsumi, and T. Sato, 2015: A study of multiple tropopause structures caused by inertia–gravity waves in the Antarctic. J. Atmos. Sci., 72, 2109–2130, doi:10.1175/JAS-D-14-0228.1.
Shutts, G. J., and S. B. Vosper, 2011: Stratospheric gravity waves revealed in NWP model forecasts. Quart. J. Roy. Meteor. Soc., 137, 303–317, doi:10.1002/qj.763.
Sun, L., G. Chen, and W. A. Robinson, 2014: The role of stratospheric polar vortex breakdown in Southern Hemisphere climate trends. J. Atmos. Sci., 71, 2335–2353, doi:10.1175/JAS-D-13-0290.1.
Waugh, D. W., W. J. Randel, S. Pawson, P. A. Newman, and E. R. Nash, 1999: Persistence of the lower stratospheric polar vortices. J. Geophys. Res., 104, 27 191–27 201, doi:10.1029/1999JD900795.
Wilcox, L. J., and A. J. Charlton-Perez, 2013: Final warming of the Southern Hemisphere polar vortex in high- and low-top CMIP5 models. J. Geophys. Res. Atmos., 118, 2535–2546, doi:10.1002/jgrd.50254.
Wright, C. J., S. M. Osprey, and J. C. Gille, 2013: Global observations of gravity wave intermittency and its impact on the observed momentum flux morphology. J. Geophys. Res. Atmos., 118, 10 980–10 993, doi:10.1002/jgrd.50869.
