Climatology of the middle atmosphere in LMDz: Impact of source‐related parameterizations of gravity wave drag

A. de la Cámara; F. Lott; M. Abalos

doi:10.1002/2016MS000753

Climatology of the middle atmosphere in LMDz: Impact of source‐related parameterizations of gravity wave drag

de la Cámara, A.; Lott, F.; Abalos, M. 2016-12-01 00:00:00 PUBLICATIONS Journal of Advances in Modeling Earth Systems RESEARCH ARTICLE Climatology of the middle atmosphere in LMDz: Impact of 10.1002/2016MS000753 source-related parameterizations of gravity wave drag 1 2 1 Key Points: A. de la Camara , F. Lott , and M. Abalos Good stratospheric climatology using source-related GW parameterizations 1 2 National Center for Atmospheric Research, Boulder Colorado, USA, Laboratoire de Meteorologie Dynamique, IPSL and Gravity wave sources have an impact the annual cycle in the middle CNRS, Ecole Normale Superieure, Paris, France atmosphere Climate change is not signiﬁcantly affected by changes in GW sources Abstract Gravity wave (GW) parameterizations control the mean state and variability of the middle atmosphere in present-day climate models. The most recent parameterizations relate the GWs to their non- Supporting Information: orographic sources (fronts and convection), which impacts the annual cycle of the GW drag, and makes the Figure S1 GWs respond to the changing climate. These issues are addressed using the Laboratoire de Meteorologie Dynamique Zoom (LMDz) climate model, showing ﬁrst a climatology of the middle atmosphere in the pres- Correspondence to: ence of nonorographic GW sources. The model performance is comparable with that documented in earlier A. de la C amara, [email protected] model versions, illustrating that there are no major difﬁculties in including nonorographic GW sources in models. A twin experiment where the parameterization of GWs has no link with the nonorographic sources Citation: is also performed. Provided that in the twin experiment the launched GW stress is very intermittent, its cli- de la C amara, A., F. Lott, and M. Abalos matology compares reasonably well with the experiment with sources. This illustrates that GW intermittency (2016), Climatology of the middle is a key factor in GW dynamics, but also that the dynamical ﬁltering of the waves by the background ﬂow atmosphere in LMDz: Impact of source-related parameterizations of strongly modulates the signiﬁcance of the sources. Some impacts of having GW sources on the annual cycle gravity wave drag, J. Adv. Model. Earth of the zonal mean circulation of the middle atmosphere are nevertheless evident. In a changing climate, the Syst., 8, 1507–1525, doi:10.1002/ impact of introducing GW sources also seems to be substantially mitigated by the dynamical ﬁltering. The 2016MS000753. experiments and diagnostics are nevertheless limited in time and to the averaged climatology, respectively, calling for longer tests to measure the impacts on the atmospheric low frequency variability. Received 4 JUL 2016 Accepted 10 SEP 2016 Accepted article online 19 SEP 2016 Published online 6 OCT 2016 1. Introduction The representation of gravity waves (GW) is critical for the proper representation of the circulations of both the troposphere and the middle atmosphere in general circulation models (GCM). Orographic GWs were the ﬁrst to be parameterized, their effects helping to reduce biases in the upper tropospheric and lower stratospheric jets [e.g., Palmer et al., 1986; McFarlane, 1987; Lott, 1999]. Nonorographic GWs have been incor- porated thereafter, aiming at reducing very large biases in the stratosphere and mesosphere [e.g., Manzini et al., 1997; Sassi et al., 2002; Lott et al., 2005]. Contrarily to the orographic GWs, for which the source mechanisms are relatively well understood, the mechanisms exciting the nonorographic GWs are less evident [Fritts and Alexander, 2003]. For this reason, early parameterizations of the nonorographic GWs have no relation with their sources. Among these param- eterizations, the so-called ‘‘globally spectral’’ ones [Hines, 1997; Warner and McIntyre, 1996] assume that the GWs follow a saturated spectra, somehow in agreement with observations [Fritts, 1989, and references therein]. The good performance of the models that use these schemes [e.g., Lott et al., 2005, and references therein] witnesses that, for gravity waves, the dynamical ﬁltering due to the air density decrease with alti- tude and the vertical variations of the large scale winds play a central role determining the GW drag (GWD). V C 2016. The Authors. The globally spectral schemes are also used for the practical reason that they permit the treatment of a This is an open access article under the large ensemble of waves at a reasonable numerical cost. Nevertheless, the absence of sources in GW param- terms of the Creative Commons eterizations limit their potential calibration with the growing number of in situ and satellite observations, Attribution-NonCommercial-NoDerivs and is maybe a cause for systematic errors, at least in the Southern Hemisphere spring [McLandress et al., License, which permits use and 2012; de la Camara et al., 2016]. distribution in any medium, provided the original work is properly cited, the Consequently, many efforts have been made from theoretical, observational and modeling perspectives to use is non-commercial and no understand the mechanisms generating nonorographic GWs. As a result, many climate models now include modiﬁcations or adaptations are made. parameterizations of GWs generated by convection [Beres, 2005; Song and Chun, 2005; Lott and Guez, 2013; DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1507 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 Schirber et al., 2014; Bushell et al., 2015] and by fronts [Rind et al., 1988; Charron and Manzini, 2002; Richter et al., 2010] or planetary wave breaking [Zulicke and Peters, 2008]. Many of these parameterizations prefer to adopt a ‘‘multiwave’’ approach rather than a globally spectral one to treat the GWs. For the convective waves this is because it is quite easy to include a diabatic heating into a GW linear equation, and for the fronts an ageostrophic momentum forcing could play a similar role. This is the approach followed by the parameterizations that evaluate a frontogenesis function to identify where the GWs are excited [Rind et al., 1988; Charron and Manzini, 2002; Richter et al., 2010], a step that is quite demanding technically. For this rea- son, and also because there are no closed theories relating such an ‘‘ageostrophic forcing’’ to the GWs potentially produced, de la Camara and Lott [2015] use a simple relation between GWs and fronts that is based on theoretical results on GW emission from potential vorticity anomalies in sheared ﬂows [Lott et al., 2010, 2012a]. Interestingly, de la Camara et al. [2016] have recently demonstrated that the GW intermittency resulting from the introduction of sources (convection and fronts) is signiﬁcant to predict well the timing of the Southern Hemisphere stratospheric ﬁnal warming. It is also worthwhile to recall that by using stochastic techniques, the multiwave methods can be made much more computationally efﬁcient than initially thought (see discussions in Eckermann [2011] and Lott et al. [2012b]). A fundamental motivation to relate the GWs to their potential sources is that these sources can have an annual cycle and change when the climate changes. It is therefore important to test if this can affect the model annual cycle in the middle atmosphere and to analyze if the changes in the GW sources impact the prediction of the future climate. To have a more thorough understanding of their impact, a longer term objective is to include the GW sources in some of the models participating in the next climate model inter- comparison project (CMIP6). This is the approach followed by the Laboratoire de M et eorologie Dynamique Zoom (LMDz) GCM where all the parameterized GWs will be related to their sources, e.g., mountains, con- vection and fronts. This is in opposition with CMIP5 [Lott et al., 2005], where the LMDz model used the Hines [1997] globally spectral scheme to parameterize the nonorographic GWs. The ﬁrst purpose of this paper is therefore to carefully analyze the model middle atmospheric climate and variability, and to see if the frontal and convective GWs can do at least as well as the Hines [1997] uniform background of waves. The second is to start testing if having time and space varying sources inﬂuences the troposphere and middle atmosphere climate. The paper is organized as follows. Section 2 presents the LMDz model and summarizes the source-related GW parameterizations. Section 3 validates the model climatology and variability of the middle atmosphere against the European Centre for Middle-range Weather Forecast Interim Reanalysis (ERA-Interim), while sec- tion 4 addresses the impact of introducing nonorographic GW sources in the parameterizations. The main conclusions are given in section 5. 2. Model Description The version of LMDz we use has a 3.758 3 1.8758 longitude-latitude grid, and 71 levels in the vertical with the top at 0.01 hPa. Its vertical resolution is around 1 km in the lower stratosphere. The results shown are from a 20 year simulation (referred to as CONTROL), forced with climatological ﬁelds of sea surface tempera- ture, sea ice, soil temperature and composition over land. These climatological ﬁelds are averages over the period 1979–2005, as are the ozone ﬁelds, which are those predicted by the LMDz-Reprobus coupled climate-chemistry model [Jourdain et al., 2008]. 2.1. Gravity Wave Drag Parameterizations LMDz uses three distinct GWD parameterizations that account for GWs generated by topography [Lott, 1999], convection [Lott and Guez, 2013], and fronts [de la Camara and Lott, 2015]. The parameterizations of nonorographic GWs are based on the stochastic approach introduced by Eckermann [2011] and Lott et al. [2012b], and consists in sampling randomly the GW spectrum by launching 8 monochromatic waves at each grid point and ‘‘physical’’ time step (e.g., every 30 min). The waves chosen are purely zonal, and their zonal wavenumber k is chosen randomly between two extrema corresponding to wavelength between 1 km and 300 km and using a uniform statistics. The phase speed is chosen randomly according to a Gauss- –1 ian distribution with 40 ms standard deviation and centered on the wind velocity at the emission level (500 hPa for convective waves, 900 hPa for frontal waves). DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1508 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 As the sensitivity to the nonorographic sources is our ﬁrst objective, we next recall how this is done in LMDz (for the other aspects, like the treatment of the breaking, or the statistical superposition of the waves see Lott and Guez [2013] and de la Camara and Lott [2015]). For convective waves, the emitted GW stress at the launching altitude (z ) is: 2 2 2 2 2m Dz ~ ~ z RL jkj e k l W F 5q G P ; (1) c0 conv 3 q Hc NX r p jjkjj where q is the density at a reference level, G is a tunable, dimensionless parameter of order 1 (we take r c0 ; G 51:75), Dz a tunable characteristic depth of the heating source (we take Dz51 km), R is the ideal gas c0 constant, L is the latent heat of condensation, H5 7 km is the stratospheric scale height, c is the speciﬁc W p heat at constant pressure, k is the horizontal wavenumber vector, m is the vertical wavenumber 2 2 2 2 ~ ~ ~ (m 5N jkj =X ), N is the buoyancy frequency, X5x2k U is the intrinsic frequency (with x the ground- based frequency, and U the horizontal wind vector), and P is the grid-scale precipitation. Therefore, equa- tion (1) translates the gridscale precipitation into a subgrid scale GW stress. For frontal waves, the emitted GW stress is [see de la Camara and Lott, 2015]: z 0 ~ top Nðz Þ z dz k l 2p 0 0 2 0 0 0 ~ Uz ðz Þ F 5G q ðz ÞNðz Þf ðz Þe dz ; (2) f 0 fron 0 4f jjkjj where z is the top of the model, G is a tunable, dimensionless parameter of order 1 (we take G 52), dz top f 0 f 0 2z=H is a typical vertical depth of the vorticity anomaly (set to 1 km), f is the Coriolis parameter, q 5q e is the 0 r reference state density, f is the grid-scale relative vorticity, and U is the vertical shear. As we see equation (2) translates the resolved dynamics (gridscale vorticity and stability conditions) into a subgrid scale GW stress. 2.2. Characteristics of the Nonorographic gravity waves Figure 1 shows the annual cycle of eastward and westward momentum ﬂux (MF) at the launching altitude (i.e., convective GW stress at 500 hPa plus frontal GW stress at 900 hPa), 100 hPa and 1 hPa. At the level of emission (Figures 1e and 1f), the band of high MF in the tropics is due to the convectively generated GWs. The bands in the midlatitudes of both hemispheres are mainly due to frontal GWs, although convective GWs also contribute. The emitted MF is almost similar in amplitude for eastward and westward MF, and both exhibit a pronounced annual cycle. The tropical band migrates northward and gets stronger during the northern summer, consistent with the behavior of precipitation in the model (not shown). In the midlati- tudes, the emitted MF is weaker than in the tropics and presents higher values in winter of both hemi- spheres, consistent with stronger baroclinicity. The effect of wind ﬁltering on the GW propagation is evident, the MF getting smaller and smaller when entering the stratosphere at 100 hPa and the mesosphere at 1 hPa. At 100 hPa (Figures 1c and 1d), the MF still has some pattern similarities with the emitted MF, although the wind ﬁltering modulates the annual cycle. At 1 hPa (Figures 1a and 1b) the tropical band has been ﬁltered out to a large extent. In the extra- tropics, while the annual cycle pattern of westward MF somehow resembles that of the emitted ﬂux, the pattern of eastward ﬂuxes is almost in phase opposition with that emitted, presenting larger values during the summer months in both hemispheres. It is interesting to compare the GW stress at 100 hPa with the results of Richter et al. [2010], where the authors show similar plots for convective and frontal GW stresses separately at 100 hPa in WACCM3.5 (their Figures 2 and 3). The annual cycle of the GW stress in the tropics (Figures 1c and 1d) resembles the convec- tive GW stress in Richter et al. [2010], but the magnitude in our model is smaller by a factor of 2. At mid-to- high latitudes, the GW stress at 100 hPa is also qualitatively similar to the frontal GW stress in Richter et al. [2010], and this time the magnitude is smaller in our scheme by a factor of 1.5. Such differences in stress amplitude are not surprising given that the two models are very different, and most of all the WACCM mod- el top (0.0001 hPa) is much higher than the LMDz model top (0.01 hPa), meaning that a given stress can give much larger drag near the top in the ﬁrst model than in the second. This can yield modelers to tune the launched GW stress to control the drag amplitude near the model top. DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1509 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 b) Westward MF 1 hPa CONTROL a) Eastward MF 1 hPa CONTROL 1.5 1.35 1.2 1.05 30 30 0.9 0 0.75 0.6 −30 −30 0.45 0.3 −60 −60 0.15 J F M A M J J A S O N D J F M A M J J A S O N D c) Eastward MF 100 hPa CONTROL d) Westward MF 100 hPa CONTROL 2.7 60 60 2.4 2.1 1.8 0 0 1.5 1.2 −30 −30 0.9 0.6 −60 −60 0.3 J F M A M J J A S O N D J F M A M J J A S O N D f) Westward MF emitted CONTROL e) Eastward MF emitted CONTROL 5.4 60 60 4.8 4.2 3.6 0 0 3 2.4 −30 −30 1.8 1.2 −60 −60 0.6 J F M A M J J A S O N D J F M A M J J A S O N D Figure 1. Total eastward (left column) and westward (right column) momentum ﬂux at the level of emission, 100 hPa, and 1 hPa as indicat- ed, from nonorographic gravity waves as a function of latitude and time of the year (in mPa) in CONTROL. Figure 2 presents the drag imposed on the mean ﬂow by the frontal, convective, and orographic GW parameterizations for DJF and JJA. Frontal GWs are the main contributor to the total GWD in the southern 21 –1 extratropics in both seasons, with peak values larger than 621 ms d near 608S at mesospheric levels above 0.1 hPa. Convective GWD is weaker than the frontal drag in the extratropics, but it presents relative –1 maxima (about 63–6 ms d ) near 508 latitude in both hemispheres at the highest altitudes of the mod- el, presumably associated with the location of the storm tracks. The strong dissipation of MF in the tropics between 100 and 1 hPa described in Figure 1 is not evident here due to density effects, i.e., the drag is pro- portional to the the vertical divergence of the momentum ﬂux and inversely proportional to density. Oro- –1 graphic GWD is mainly active in northern winter extratropical stratosphere, reaching 29ms d at 0.1 hPa. 3. Mean Climate and Variability of the Middle Atmosphere 3.1. Zonal Mean Climate As this paper focuses on the impact of including sources in the nonorographic GWD schemes, we have tuned these parameterizations to ensure that LMDz has a climatology at least comparable to that docu- mented in its previous stratospheric version [Lott et al., 2005]. As we shall see, the improvements in some places are obvious, like in the QBO region, whereas in the midlatitudes the effects are more neutral. Note DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1510 latitude latitude latitude Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) DJF FGWD b) DJF CGWD c) DJF OGWD 0.01 0.1 10 18 100 12 e) JJA CGWD f) JJA OGWD d) JJA FGWD 0.01 −3 −6 0.1 −9 −12 1 −15 −18 −21 −24 −27 −30 −90 −45 0 45 90 −90 −45 0 45 90 −90 −45 0 45 90 latitude latitude latitude 21 –1 Figure 2. Longitudinally averaged drag (in ms d ) from the (left column) frontal (FGWD), (middle column) convective (CGWD) and (right column) orographic (OGWD) gravity waves for DJF and JJA in CONTROL. * * a) DJF U, Ψ CONTROL b) MAM U, Ψ CONTROL res res 0.01 0.01 0.1 0.1 95 1 1 10 10 100 100 1000 1000 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 15 * d) SON U, Ψ CONTROL −5 c) JJA U, Ψ CONTROL res res −15 0.01 0.01 −25 0.1 −35 0.1 −45 1 −55 −65 −75 −85 −95 1000 1000 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 latitude latitude –1 Figure 3. Zonally averaged zonal wind proﬁles (in ms , shaded), and stream function of the residual mean meridional circulation (con- tours) in CONTROL. Magenta contours represent positive values (i.e., clockwise circulation), and cyan contours represent negative values (i.e., counter-clockwise circulation). DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1511 pressure (hPa) pressure (hPa) pressure (hPa) pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 * * a) DJF U, Ψ ERAI b) MAM U, Ψ ERAI res res 0.01 0.01 0.1 0.1 95 1 1 10 10 100 100 1000 1000 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 15 d) SON U, Ψ ERAI −5 c) JJA U, Ψ ERAI res res −15 0.01 0.01 −25 0.1 −35 0.1 −45 1 −55 −65 −75 −85 100 100 −95 1000 1000 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 latitude latitude Figure 4. As in Figure 3 but for ERA-Interim. that having a model version with GW sources and a QBO but without degrading the model in other places was an implicit objective of the present study. To illustrate this, Figure 3 shows the seasonal averages of zonal-mean zonal wind proﬁles. It shows well- –1 deﬁned polar night jets in the solstices with values up to 40 and 85 ms in the boreal and austral jet cores, –1 respectively. The summer easterly jets present maximum values of 270 ms in the subtropics at around 1 hPa, and the winds show transition conditions in the equinoxes. These zonal mean winds compare well with those corresponding to an earlier model version [Lott et al., 2005, Figure 3], but some biases that were present in the previous version of the model remain. When compared to ERAI in Figure 4 we see that the –1 largest biases in the model are in the summer easterly jets, with winds 20 ms stronger in LMDz than in ERAI. Also, the SH easterly jet in DJF splits into two parts (Figure 3a). The strength of the polar night jet is comparable in the two data sets, although the boreal jet in LMDz is weaker than in ERAI in the upper strato- sphere and lower mesosphere. This is more clearly seen in Figure 5, which speciﬁcally shows the wind speed in the jet core and its latitudinal position as a function of height during the northern and southern winters (note that Figures 5 and 6 include results for GWLOG, which will be discussed in section 4). The lati- tudinal tilt of the jets with altitude is well captured, with the exception of the southern jet in JJA. This bias is common to most climate models [Butchart et al., 2011]. The model performance in MAM and SON shows good agreement with ERAI (Figures 3 and 4). To complete the description of the zonal mean circulation, the contours in Figure 3 display the mass streamfunction in CONTROL, representing the residual mean meridional circulation W : res @W res 52q cos / v ; (3) @z where q 5q ðzÞ is the background density, / is latitude, and v is the latitudinal component of the residual 0 0 circulation in the Transformed Eulerian Mean (TEM) formalism [Andrews et al., 1987]: DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1512 pressure (hPa) pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) Strength NH Jet (DJF) b) Latitude NH Jet (DJF) 0.01 0.01 0.03 CONTROL 0.03 0.1 0.1 ERAI 0.3 0.3 GWLOG 1 1 3 3 10 10 30 30 100 100 300 300 0 25 50 75 100 20 40 60 80 c) Strength SH Jet (JJA) d) Latitude SH Jet (JJA) 0.01 0.01 0.03 0.03 0.1 0.1 0.3 0.3 1 1 3 3 10 10 30 30 100 100 300 300 0 25 50 75 100 −80 −60 −40 −20 zonal wind (m/s) latitude Figure 5. Zonal wind speed and latitude and latitude of the jet maxima (top) of the NH DJF climatology and (bottom) of the SH JJA clima- tology, for CONTROL (blue line), ERAI (green line), GWLOG (red line, see section 4). 1 @ q v h v v2 (4) q @z @h=@z In DJF and JJA (Figure 3c), the main circulation cell presents upward motions in the tropics, extending to the summer hemisphere, that reach mesospheric altitudes, and downward motions in the winter high lati- tudes. The meridional motion in the mesosphere above 1 hPa is responsible for the dynamical maintenance of winter pole temperatures much warmer than summer pole temperatures in the mesosphere (not shown). A secondary, shallow circulation cell can also be seen in the summer hemisphere lower stratosphere. All these features compare well with ERAI (Figure 4). During MAM and SON the circulation cells grow deeper in the autumn and shallower in the spring hemispheres, in good agreement with the reanalysis. a) NH Max variability (DJF) b) NH Latitude max variability (DJF) 0.01 0.01 0.03 0.03 CONTROL 0.1 0.1 ERAI 0.3 0.3 GWLOG 1 1 3 3 10 10 30 30 100 100 300 300 0 5 10 15 20 25 20 40 60 80 c) SH Max variability (JJA) d) SH Latitude max variability (JJA) 0.01 0.01 0.03 0.03 0.1 0.1 0.3 0.3 1 1 3 3 10 10 30 30 100 100 300 300 0 5 10 15 20 25 −80 −60 −40 −20 zonal wind (m/s) latitude Figure 6. Location and amplitude of the maximum interannual standard deviation of the zonal mean zonal wind (top) in the NH in DJF poleward of 458N and (bottom) in the SH in JJA poleward of 308S, for CONTROL (blue line), ERAI (green line), GWLOG (red line, see section 4). DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1513 pressure (hPa) pressure (hPa) pressure (hPa) pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) NP T 10hPa CONTROL b) NP T 10hPa, ERAI c) NP T 10hPa GWLOG J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J d) SP T 10hPa CONTROL e) SP T 10hPa, ERAI f) SP T 10hPa GWLOG J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D Figure 7. Polar temperatures (in K) at 10 hPa and 858 latitude in CONTROL, ERA-Interim, and GWLOG (see section 4) for (a, b, c) the North- ern Hemisphere, and (d, e, f) the Southern Hemisphere as a function of time of the year. 3.2. Interannual Variability To analyze the simulated variability, Figure 6 shows the amplitude and latitudinal location of maximum interannual variability of the polar night jets as a function of height, for CONTROL and ERAI. For the North- ern Hemisphere (NH) winter, the amplitude of the maximum variability is well represented below 1 hPa as compared to ERAI, but the model underestimates it in the lower mesosphere (Figure 6a). The latitudinal location of this variability is not so well represented (Figure 6b). While in ERAI the position tilts equatorward with height between 30 and 1 hPa, in LMDz it tilts poleward. This bias is common to many climate models, and needs further investigation [Butchart et al., 2011]. For the Southern Hemisphere (SH) winter, the com- parison to ERAI provides similar conclusions. A slight equatorward tilt with height of the maximum variabili- ty does appear in CONTROL below 1 hPa, but with a much steeper slope than in ERAI. The maximum variability is 10–208 poleward in CONTROL as compared to ERAI, possibly due to differences in the locations of the jet core (Figure 5d). A complementary view of the interannual variability is given by the time series of polar temperature in both hemispheres (at 858 latitude) at 10 hPa in Figure 7. The seasonal evolution and variability, as well as the inter- hemispheric contrasts, are generally well captured. However the model presents too much variability, as evi- denced by sporadic warmings in 0.5 October and November in the North CONTROL (0.76/year) Pole (Figure 7a), or the spread in ERAI (0.67/year) 0.4 temperatures in the South Pole dur- GWLOG (0.58/year) ing the austral winter/spring that are not present in the ERAI data. Figure 0.3 8 further shows an histogram of fre- quencies of major sudden warmings 0.2 (MSW) in the NH for both CONTROL and ERAI, sorted by winter month. 0.1 We have followed the method by Charlton and Polvani [2007] to iden- tify these events. The frequency of Nov Dec Jan Feb Mar events is higher in CONTROL than in ERAI (0.76/year versus 0.67/year), Figure 8. Frequency of major stratospheric sudden warmings (number of MSW per conﬁrming larger simulated vari- year) in the NH for CONTROL (blue bars), ERAI (green bars), and GWLOG (red bars, see section 4), sorted by month. The total frequency is also indicated in the ﬁgure legend. ability. Importantly, the intraseasonal DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1514 frequency of major sudden warmings Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) QBO CONTROL 1 2 3 4 5 6 7 8 9 10 b) QBO ERAI −5 −15 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 c) QBO GWLOG −25 −35 −45 1 2 3 4 5 6 7 8 9 10 year Figure 9. Zonal mean zonal wind averaged over the equatorial band 58S–58N for (a) CONTROL, (b) ERAI, and (c) GWLOG (see section 4). –1 Contour interval: 10 m s , bold grey contour indicates the zero-wind line. distribution of major warmings consistently presents higher frequencies as the winter season progresses, peaking up in February. The differences against ERAI include too high frequencies in November and February, and too low in December. Given the multiple factors inﬂuencing the occurrence of MSWs, we consider that the performance of LMDz compares well with that of the previous model version (not shown here, but see Lott et al. [2005, Figure 13]). Additionally, the model does not present a signiﬁcant delay in the simulation of the stratospheric ﬁnal warming in the SH [de la Cam ara et al., 2016], a bias that most climate models still have [e.g., Butchart et al., 2011; McLandress et al., 2012; Wilcox and Charlton-Perez, 2013]. In the tropical lower stratosphere, the QBO dominates the interannual variability of the zonal winds. Figure 9 shows the zonal winds at the Equator as a function of time and height. The model internally generates a QBO with an average period of 28 months that closely matches that in ERAI (27 months). Yet there are some discrepancies between the model and the reanalysis in Figure 9, such as wind velocities that are up –1 to 10 ms weaker in the model, especially during the westward phase. Also the QBO in CONTROL does not descend as low as it does in ERAI, and it lacks the westerlies stalling that often occurs below 30 hPa (see e.g., the years 2009–2010 in Figure 9b). The causes are multiple, but we suspect that the underestimation of the slow Kelvin waves in LMDz might play a signiﬁcant role [Maury et al., 2013]. A ﬁner vertical resolution in the lower stratosphere (1 km in this model version) might also contribute to improve the QBO simulation especially at lower levels [see Anstey et al., 2016]. The reader is referred to Lott and Guez [2013] for further details on the simulation of the QBO in LMDz and the comparison with observations. For completeness, we recall here that the QBO was absent in Lott et al. [2005]. 4. Impact of Source-Varying GWD Parameterizations In this section we evaluate the impact of including sources of nonorographic gravity waves (NGW). First, we describe the twin experiments performed, and then we will present the results, with the focus on DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1515 Pressure (hPa) Pressure (hPa) Pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 the simulated annual cycle in the middle atmosphere, and on possible impacts under future climate conditions. 4.1. Model Experiments Recent studies have shown that linking the parameterized GW amplitudes to their nonorographic sources naturally produces intermittent MFs, the probability density function of absolute momentum ﬂuxes ﬁtting a lognormal distribution [de la Camara et al., 2014; de la Camara and Lott, 2015; Stephan and Alexander, 2015]. These studies also suggest that the NGW intermittency can help reduce model biases, simply because for a given averaged launched momentum ﬂux, few large amplitude waves break at lower altitude than a large number of small amplitude waves. Therefore, to evaluate the role of the NGW sources speciﬁcally, we next replace the source terms in the convective and frontal schemes (i.e., the P and f terms in equations (1) and (2)) by random numbers produced by a lognormal distribution. The characteristics of the distribution are tuned to obtain a reasonable zonal mean climatology (see next section). This run is referred to as GWLOG. We also apply a latitudinal weighting in the modiﬁed convective GW scheme to launch larger stress in the tropics and help generate a QBO. Speciﬁcally, the latitudinal weighting function chosen is 2 30 f ð/Þ5ð0:15sin 2/11:1cos /Þ. This function has a narrow maximum at the equator and two secondary peaks at 458 latitude, qualitatively mimicking the averaged latitudinal distribution of precipitation [e.g., Lott and Guez, 2013]. The magnitudes of the maxima have been chosen ad hoc to obtain a reasonable climatolo- gy (see next section). A different potential impact of having source-related NGW schemes is that parameterized wave amplitudes will change if climate changes. To investigate this point, we perform two additional experiments. First we make a 20 year experiment, named 4xCONTROL, where the GW speciﬁcations are as in CONTROL, but increasing the CO concentrations by a factor of four, and by adding everywhere 4 K to the prescribed SST. Second we make another 20 year experiment, named 4xGWLOG, similar to 4xCONTROL but using the GW speciﬁcations of GWLOG. 4.2. Climatology of the Simulation Without GW Sources To make a fair comparison between the simulations with and without NGW sources we have tried to make them as close as possible in terms of the GW drag in the midlatitudes, the middle atmosphere jets in the midlatitudes and subtropics, and the QBO in the tropical lower stratosphere. Figure 10 displays the nonoro- graphic GWD for DJF and JJA in the CONTROL and GWLOG runs. The lognormal distributions of emitted GW stress used in GWLOG provide GWD proﬁles qualitatively similar to those in CONTROL. Gravity wave drag –1 values larger than 63ms d are found above 1 hPa in the mesosphere of the two runs. Quantitatively, the drag in CONTROL is slightly weaker than in GWLOG, particularly in the summer hemisphere. Concerning the impacts on the mean climate, we return to Figure 5 that shows the strength and location of the wintertime polar jets in GWLOG. The zonal mean climate of GWLOG is comparable to that of CON- TROL during the solstices. Although the panels in Figure 5 focus on the winter westerly jets, essentially because the GW parameterizations are ﬁrst intended to improve them, it is important to say that similari- ties are found in the midlatitudes during other seasons. Beyond the zonal means, it is much more difﬁcult to control the variability, as illustrated in Figure 6 where the variability of the jet in GWLOG is also shown. The most notorious differences appear in the SH, such as larger variability of the jet in GWLOG (Figure 6c), and the absence of the equatorward tilt with height in the jet variability below 1 hPa (Figure 6d). Still concerning the variability but coming back to the NH, Figure 8 also shows the MSW statistics for GWLOG. The mean winter frequency is reasonable (0.58 per year), but GWLOG fails in capturing the intra- seasonal distribution of the major warmings. This result may be due to chance, but it is interesting that removing the relation with the NGW sources degrades the SSW seasonality. We nevertheless need to test it with longer model simulations. On the other hand, there is some improvement in MSW frequency in December, where most climate models do not get nearly enough warmings just like the control run. Finally, and concerning the tropical region, Figure 9 shows that GWLOG also has an internally generated QBO, its period is slightly longer than in CONTROL. Above the QBO region, the semi-annual signal seems more pronounced in GWLOG. DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1516 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) DJF CONTROL b) DJF GWLOG 0.01 0.1 10 24 100 16 d) JJA GWLOG 0 c) JJA CONTROL 0.01 −4 −8 −12 0.1 −16 −20 −24 −28 −32 −36 −40 −90 −45 0 45 90 −90 −45 0 45 90 latitude latitude 21 –1 Figure 10. Zonally averaged drag from the NGW parameterizations (in ms d ), for (a, b) DJF and (c, d) JJA, from the CONTROL and 21 –1 GWLOG runs as indicated. The magenta and cyan lines indicate 10.1 and 20.1 ms d contours, respectively. 4.3. Impact on the Annual Cycle We showed in Figure 1 that the GW stress emitted in CONTROL presents a strong annual cycle, presumably due to the annual cycle of the GW sources activity. Figure 11 presents the eastward and westward NGW stress as a function of latitude and time of the year for GWLOG. As expected, a very weak seasonality appears at the altitude of emission since NGW sources are not considered in this run. At 100 hPa, seasonal differences start to show up, and at 1 hPa a strong annual cycle is present due to momentum ﬂux dissipa- tion. We can now compare this performance in GWLOG with that in CONTROL (Figure 1). At 1 hPa both east- ward and westward momentum ﬂuxes are very similar in magnitude and seasonal evolution in both runs. This contrasts with the stress at 100 hPa, where the annual cycle is much stronger and peak values are much larger for CONTROL than for GWLOG, specially for the westward direction (e.g., 2.1 versus 1.2 mPa at 508S in August). It can be interpreted then that the GW stress entering the mesosphere in our simulations is only weakly dependent on the seasonal cycle of the stress at lower altitudes, and in particular on the sea- sonal cycle introduced by the GW sources. On the other hand, this implies a distinct momentum ﬂux dissi- pation in the stratosphere between these two runs, which may result in differences of GW drag in the stratosphere. In terms of nonorographic GW drag, the difference between CONTROL and GWLOG is also signiﬁcant as illustrate the Figures 12a and 12b where the annual cycle of the drag averaged for the northern (508N– 808N) and southern (508S–808S) high latitudes, are shown. In the NH (Figure 12a), there is a band of negative differences in the lowermost stratosphere during the whole year, perhaps pointing to larger westward net stress emitted (i.e., producing a negative drag) in CONTROL than in GWLOG. Above 50 hPa, a marked sea- sonal cycle appears, with positive differences during summer and negative during winter, changing sign in the mesosphere above 0.5 hPa. In the SH (Figure 12b), there is a noticeable annual cycle in the drag up to 0.1 hPa, with positive differences in summer and negative in winter. The magnitude is also small, reaching 21 –1 up to 20.3 ms d in the upper stratosphere in JJA. The negative differences during the summer months descend throughout the season and reach the lower stratosphere by September. The impact of the GW seasonality on the annual cycle of the zonal winds at midlatitudes is not very signiﬁ- cant, consistently with the fact that we tuned GWLOG with this objective (see supporting information). The situation is somehow different if we look at the Brewer-Dobson circulation, as we show below. To evaluate DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1517 pressure (hPa) pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 b) Westward MF 1 hPa GWLOG a) Eastward MF 1 hPa GWLOG 1.5 1.35 1.2 1.05 30 30 0.9 0 0 0.75 0.6 −30 −30 0.45 0.3 −60 −60 0.15 J F M A M J J A S O N D J F M A M J J A S O N D c) Eastward MF 100 hPa GWLOG d) Westward MF 100 hPa GWLOG 2.7 60 60 2.4 2.1 1.8 0 0 1.5 1.2 −30 −30 0.9 0.6 −60 −60 0.3 J F M A M J J A S O N D J F M A M J J A S O N D f) Westward MF emitted GWLOG e) Eastward MF emitted GWLOG 5.4 4.8 4.2 3.6 0 0 3 2.4 −30 −30 1.8 1.2 −60 −60 0.6 J F M A M J J A S O N D J F M A M J J A S O N D Figure 11. As in Figure 1 but for LMDz-GWLOG. the impact of the GWs on the Brewer-Dobson circulation we use the TEM formalism, where the zonal momentum equation is given by [Andrews et al., 1987]: @u @u 2 v f1w 5DF1X (5) @t @z ^ 1 and where ðv ; w Þ are the meridional and vertical components of the TEM residual circulation, f5f2 acos / @ðu cos /Þ ~~ rF with f the Coriolis parameter, DF5 is the force applied by the resolved waves with F the @/ q a cos / Eliassen-Palm (EP) ﬂux, and X is the force applied by unresolved processes (in our case the parameterized gravity wave drag). In equation (5), the left-hand terms represent the circulation response to the forcing applied by the right-hand terms. We next evaluate the vertical motion over the high latitudes (i.e., downwel- ling), focusing on the possible response of the mean meridional circulation to the NGW drag differences between CONTROL and GWLOG. Following Randel et al. [2002] and Abalos et al. [2012], we combine equa- tion (5) and the TEM continuity equation [Andrews et al., 1987] to derive the vertical component of the resid- ual circulation: () 0 2 z=H 1 2z =H 0 2e e cos / @u ð/; z Þ 0 0 0 w ð/; zÞ5 DFð/; z Þ1Xð/; z Þ2 dz : (6) 2 0 @t z fð/; z Þ a cos /; d/ / / 1 1 We take / 560 N, / 580 N for the NH, and / 580 S, / 560 S for the SH. Note that the forcing from the 1 2 1 2 total GW drag (orographic plus nonorographic, i.e., X ) is explicitly taken into account in equation (6). DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1518 latitude latitude latitude −0.01 0.03 0.03 −0.1 −0.01 −0.03 −0.03 0.03 −0.1 0.1 −1 −3 −1 1 1 −0.3 0.3 −0 0.03 0 0.3 0.01 .01 .1 3 −0.1 − 0.1 0.0 0.01 0.01 3 −0.03 −0.01 0.01 0.03 0.1 −0.01 −0.3 0.01 0.01 − −0 0.03 .01 0.01 0.03 −0.1 −0.3 0.3 0.3 0.03 − −0.0 0.01 3 −0.1 0.1 0.01 0.01 0.1 0.03 0.03 0.01 −0.01 0.03 − −0 0.0 .01 3 −0.3 −0.1 −0.3 0.3 0.1 0.03 0.01 0.03 0.01 0.03 −1 0.1 0.3 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) NGWD NH CONTROL−GWLOG b) NGWD SH CONTROL−GWLOG 0.01 0.01 0.1 0.1 1 −3 0.3 0.3 −0.03 −0.3 0.10.3 0.3 0.01 −0.1 0.1 −1 −0.01 −0.01 1 1 0.03 0.01 0.01 −0.03 −0.01 −0.1 0.03 10 10 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * c) w NH CONTROL−GWLOG d) w SH CONTROL−GWLOG m m 0.01 0.01 0.1 0.03 −0.3 −1 0.1 0.1 0.3 −0.01 0.01 −0.3 0.3 −0.03 −0.1 0.3 1 −1 0.1 0.1 0.1 −0.1 1 1 −0.03 −0.01 0.3 10 10 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * e) w NH CONTROL−GWLOG f) w SH CONTROL−GWLOG m,NGWD m,NGWD 0.01 0.01 0.01 −0.03 −0.3 0.01 −0.1 −1 0.1 0.1 0.1 0.3 −0.01 −1 −0.1 0.03 −0.1 1 0.3 0.3 1 −0.03 −0.03 0.1 0.01 0.1 −0.01 0.03 0.03 10 10 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * g) w NH CONTROL−GWLOG h) w SH CONTROL−GWLOG m,DF m,DF 0.01 0.01 0.3 −0.01 0.1 0.1 −0.3 −0.1 −0.01 0.01 −0.1 1 −0.03 −0.01 0.03 −0.03 −0.1 0.1 −0.01 0.01 0.3 0.1 1 1 0.01 0.3 10 10 100 100 J F M A M J J A S O N D J F M A M J J A S O N D Figure 12. Differences in the (a, b) NGWD and (c–h) vertical component of the residual mean meridional circulation derived from the TEM momentum balance equation, between the CONTROL and GWLOG runs as a function of height and time of the year. The data are longitu- –1 dinally averaged over the 508–808 latitude band in both hemispheres. Contours are at 60.01, 60.03, 60.1, 60.3, 61, 63ms d for –1 NGWD, and mms for w . Light red and blue shading indicate positive and negative statistically signiﬁcant differences, respectively (Student t-test, a50.01). The panels in the second row of Figure 12 show the annual evolution of the differences of w in the north- ern and southern high-latitudes. In the NH (Figure 12c), the pattern is similar to that of the NGWD (Figure 12a). The fact that the patterns in w are found at lower altitudes than those in the NGWD is consistent with equation (6), which links the vertical motion to the drag at that level and above. We see positive differ- ences in the winter mesosphere and negative differences in the summer mesosphere (note that the regions of statistical signiﬁcant differences are somewhat limited). This means that the amplitude of the annual cycle of w is around 10% weaker in CONTROL than in GWLOG in the mesosphere, and around 10% stron- ger in the lower stratosphere (note that in the lower stratosphere the value is not statistically signiﬁcant). In the SH, the differences in w do not present a clear pattern and are barely signiﬁcant. To address whether the w differences in the NH between CONTROL and GWLOG emerge from the NGWD differences, we evaluate separately the contributions from the NGWD and from the resolved forcing (i.e., DF in equation (5)) to the vertical component of the residual circulation [Haynes et al., 1991]. We do so by com- puting w using NGWD alone (i.e., w ): m m;NGWD DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1519 0.01 −0.03 0.01 −0.01 0.01 0.01 0.1 −0.01 0.3 0.1 −0.03 −0.01 0.03 0.03 0.1 0.1 −0 01 −0.03 0.01 0.01 −0.01 −0.01 −0.1 0.01 −0.03 0 01 −0.01 −0.03 Pressure (hPa) Pressure (hPa) Pressure (hPa) Pressure (hPa) 7 −1 −9 −7 −7 −9 60 E 120 W −1 120 W 120 E 60 W −1 −1 −1 −9 −7 −1 −7 −5 −5 −9 −3 −1 −1 −1 −1 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) DJF SLP 4xCONTROL−CONTROL b) SON SLP 4xCONTROL−CONTROL o o 180 W 0 o o 0 180 W c) DJF SLP 4xGWLOG−GWLOG d) SON SLP 4xGWLOG−GWLOG o o 180 W 0 o o 0 180 W Figure 13. Differences of sea level pressure between 4xCONTROL and CONTROL, and between 4xGWLOG and GWLOG, for (a, c) DJF (in the NH), and (b, d) SON (in the SH). Contours start at 61 hPa, with an interval of 2 hPa. Light red and blue shadings indicate positive and negative statistically signiﬁcant differences, respectively (Student t-test, a50.01). () ð 2 z=H 2z =H 2e e cos / 0 0 w ð/; zÞ5 ; X ð/; z Þ; dz ; (7) NGWD m;NGWD 2 ^ 0 a cos /; d/ z fð/; z Þ / / 1 1 and using the divergence of the EP ﬂux alone (i.e., w ): m;DF () ð 0 2 z=H 2z =H 2e e cos / 0 0 w ð/; zÞ5 ; DFð/; z Þ; dz ; (8) m;DF / 2 ^ 0 a cos /; d/ z fð/; z Þ 1 1 The corresponding plots for w and w are shown in the third and bottom rows, respectively, of m;NGWD m;DF Figure 12. In the NH, the main contribution to the change in vertical motion is due to the changes in NGWD (Figure 12e). The differences in w strongly resemble in both magnitude and evolution those in w , m;NGWD m while no clear pattern is observed for w . m;DF In the SH, the w pattern agrees with the pattern in the forcing (Figures 12b and 12f). Interestingly, m;NGWD the residual circulation induced by the resolved forcing opposes almost exactly (Figure 12h) that induced by the NGWD, resulting in the insigniﬁcant w differences in the SH. We interpret that in the NH the ampli- tude and variability of the resolved waves are sufﬁciently large not to be sensitive to the rather small differ- ences in the annual cycle of the NGWD. In contrast, in the SH the amplitude and variability of the resolved waves are not as large, and they respond compensating the forcing from the parameterized NGWs. 4.4. Impact on a Warmer Climate In this section we analyze the potential impact of NGW with source-depending amplitudes on a warmer cli- mate. Figure 13 displays sea level pressure (SLP) differences 4xCONTROL-CONTROL and 4xGWLOG-GWLOG, in DJF (NH) and SON (SH). The tropospheric circulation response to warmer conditions reinforces the DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1520 −1 −1 −9 −1 −3 −1 −5 −7 −1 −3 −1 −7 −7 −5 −3 −1 −5 −3 −13 −3 −1 −9 −3 −11 −3 −11 −5 −9 120 W 120 W 60 E 60 W 60 E 60 W 120 E 120 E −1 −5 −7 −9 −7 −3 −11 −5 −13 −3 −5 −7 −13 −11 −5 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) East 4xCONTROL−CONTROL 1 hPa b) West 4xCONTROL−CONTROL 1 hPa 0.05 0.05 0.15 60 60 30 30 −0.05 0 0 −30 −0.05 −30 −0.05 0.25 0.05 0.15 0.05 −60 −60 0.45 0.35 0.15 0.25 J F M A M J J A S O N D J F M A M J J A S O N D c) East 4xCONTROL−CONTROL 100 hPa d) West 4xCONTROL−CONTROL 100 hPa 0.05 0.15 0.15 0.25 0.35 0.25 0.45 0.55 0.35 0.65 0.45 0.05 60 60 0.55 −0.05 −0.15 0.15 0.15 0.05 −0.15 0.05 −0.05 30 30 −0.15 0.05 −0.05 −0.05 −0.05 −0.05 0.15 0.05 0 0 0.05 −0.05 −0.05 −0.15 0.05 −0.05 0.15 −0.05 −0.15 −30 −30 0.05 −0.05 −0.15 0.05 0.05 0.15 0.45 0.15 0.55 0.25 −0.15 0.25 0.65 0.35 0.75 −0.05 0.350.35 −60 −60 0.75 0.65 0.25 0.55 0.35 0.45 0.15 0.25 0.15 J F M A M J J A S O N D J F M A M J J A S O N D e) East 4xCONTROL−CONTROL emitted f) West 4xCONTROL−CONTROL emitted 0.3 0.3 0.3 0.3 0.5 0.5 0.5 0.7 0.5 0.7 60 60 0.3 0.3 0.3 0.3 0.1 0.1 0.1 0.1 −0.1 −0.1 −0.1 −0.1 −0.3 −0.3 −0.3 −0.3 −0.5 −0.5 −0.5 30 30 −0.3 −0.3 −0.3 −0.3 −0.3 −0.3 −0.3 −0.5 −0.5 −0.3 −0.5 −0.5 0 −0.5 0 −0.5 −0.5 −0.5 −0.1 −0.3 −0.3 −0.3 −0.3 −0.3 −30 −30 −0.7 −0.3 −0.3 −0.5 −0.5 −0.5 −0.5 −0.5 −0.5 −0.3 −0.3 −0.3 −0.1 −0.3 −0.1 −0.1 −0.1 0.1 0.1 0.1 0.1 0.3 0.3 0.3 0.7 0.5 0.3 0.7 0.5 0.5 0.9 −60 −60 0.5 0.5 0.7 0.5 0.5 0.3 0.3 0.7 0.3 0.3 01 01 01 J F M A M J J A S O N D J F M A M J J A S O N D Figure 14. Differences of nonorographic (left column) eastward and(right column) westward gravity wave stress (in mPa) between 4xCON- TROL and CONTROL at the launching altitude, 100 hPa and 1hPa, as indicated in the ﬁgure titles. Light red and blue shadings indicate posi- tive and negative statistically signiﬁcant differences, respectively (Student t-test, a50.01). subtropical anticyclones and deepens the subpolar lows in both hemispheres, in agreement with projec- tions from the Coupled Model Intercomparison Project Phase 5 (CMIP5) [e.g., Manzini et al., 2014]. The mag- nitude and locus of the SLP differences look insensitive to the use of parameterized NGW hooked to their sources. Figure 14 shows the difference between 4xCONTROL and CONTROL in eastward and westward stress at the launching altitude, 100 and 1 hPa. Interesting features emerge in this ﬁgure. At the launching level (Figures 14e and 14f), there is a poleward shift of the latitude bands with maximum stress in the extratropics of both hemispheres. This is consistent with the intensiﬁcation of the circulation described in Figure 13, and with the projected poleward shift in the storm tracks [Scaife et al., 2012]. It can also be seen that the annual cycle intensiﬁes. The poleward shift is also present at 100 hPa (Figures 14c and 14d), where the extratropical annual cycle is notably enhanced, particularly for the westward stress. At 1 hPa (Figures 14a and 14b), there is a weak reduction in eastward stress, and the enhanced annual cycle in the SH westward stress is collocat- ed with the maximum stress in CONTROL (Figure 1b). Figure 15 shows the corresponding plots for the difference between 4xGWLOG and GWLOG, where we can look into the effect of wind ﬁltering alone. At the launching level (Figures 15e and 15f), there is again a pole- ward shift. However, there is practically no signal of an annual cycle. This implies that the enhanced annual cycle in 4xCONTROL is due to changes in the strength of GW sources, while the poleward shift is due to a DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1521 latitude latitude latitude Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) East 4xGWLOG−GWLOG 1 hPa b) West 4xGWLOG−GWLOG 1 hPa 0.05 0.05 60 60 −0.05 30 30 0 0 −30 −0.05 −30 −0.05 0.05 0.05 −60 −60 0.15 0.25 0.15 J F M A M J J A S O N D J F M A M J J A S O N D c) East 4xGWLOG−GWLOG 100 hPa d) West 4xGWLOG−GWLOG 100 hPa 0.05 0.05 0.15 0.15 60 60 −0.05 0.15 0.05 30 30 0.05 −0.05 0 0 0.05 0.15 −0.05 −0.05 0.05 −30 −30 −0.05 0.05 −0.15 0.05 0.25 0.15 0.15 −0.05 0.35 −60 −60 0.15 0.15 0.05 0.05 005 005 J F M A M J J A S O N D J F M A M J J A S O N D e) East 4xGWLOG−GWLOG emitted f) West 4xGWLOG−GWLOG emitted 0.1 0.1 0.1 0.1 60 60 −0.1 −0.1 −0.1 −0.1 30 30 −0.3 −0.3 −0.3 −0.3 −0.5 0 0 −0.5 −0.3 −0.5 −0.3 −0.3 −0.3 −30 −30 −0.1 −0.1 −0.1 −0.1 0.1 0.1 0.1 0.1 −60 −60 0.1 0.1 0.1 J F M A M J J A S O N D J F M A M J J A S O N D Figure 15. As in Figure 14, but for the difference 4xGWLOG minus GWLOG. shift in the winds and storminess due to warmer conditions. The change in the strength of the annual cycle at 1 hPa, more pronounced for the westward component of the momentum ﬂux (Figure 15b), is mainly a result of changes in the wind ﬁltering, and not of changes in the sources. We next analyze the potential impact of triggering GWs from their sources on the seasonal cycle of the downwelling branches of the Brewer-Dobson circulation in a warmer climate. Figure 16 presents similar plots as Figure 12, but for the difference 4xCONTROL minus 4xGWLOG. The change in the NGW drag induced by linking the wave amplitude to their sources is very similar in warmer and in present climate con- ditions in both structure and magnitude (compare Figures 16a and 16b and Figures 12a and 12b). This simi- larity appears also in the w ; w , and w responses. Interestingly, some statistically signiﬁcant m m;NGWD m;DF changes in w show up in the NH (Figure 16g), but contrarily to what happens in the SH, they have the m;DF same sign as w (Figure 16c). We can then conclude that the self-adjustment of parameterized NGW m;NGWD amplitudes to climatological changes in the sources has a minor effect on the induced middle-atmospheric circulation changes in a warmer climate. We have just discussed the impact on the extra-tropical downwel- ling because we ﬁnd it to be the most sensitive aspect of the midlatitude circulation to respond to the GWs annual cycle. We nevertheless veriﬁed that this conclusion also applies to the zonal winds, and found that the differences between 4xCONTROL and 4XGWLOG in zonal mean zonal winds are almost identical to those betwen CONTROL and GWLOG (not shown but see supporting information). DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1522 latitude latitude latitude 0 0.01 −0.01 0.01 0.03 0.1 −0 03 0.01 0.1 0.1 −0.01 0.01 0.1 −0.3 −003 0.3 0.3 −3 −0.01 −1 −0.3 1 0.3 −0.3 3 0.01 0.03 −0.1 0.1 − −0.01 0.03 0.01 0.03 −3 0.1 0.3 0.1 −1 0 0..01 03 0.3 0.1 0.3 0.3 0.1 0.03 −0.1 0.1 −0.01 0.03 0.01 −0.03 −0.01 0.01 0.01 −0.3 −0.1 −1 0.3 0.1 0.03 0.01 −0.1 0.3 0.1 0.01 0.03 −0.1 0.1 0.03 0.01 0.01 0.03 −0.1 −0.01 0.01 −0.03 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) NGWD NH 4xCONTROL−4xGWLOG b) NGWD SH 4xCONTROL−4xGWLOG 0.01 0.01 −1 1 1 −0.03 −0.1 0.1 0.1 0.3 0.1 0.3 0.03 0.1 0.01 1 −0.01 1 −0.3 −0.1 10 −0.1 10 −0.03 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * c) w NH 4xCONTROL−4xGWLOG d) w SH 4xCONTROL−4xGWLOG m m 0.01 0.01 −1 0.01 −0.03 −0.1 0.1 0.1 0.3 0.03 −0.3 0.1 0.03 −0.01 0.01 0.3 0.1 0.03 1 −0.03 1 −0.01 0.01 10 10 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * e) w NH 4xCONTROL−4xGWLOG f) w SH 4xCONTROL−4xGWLOG m,NGWD m,NGWD 0.01 0.01 −0.01 −0.03 −1 0.1 0.1 0.3 −0.3 0.3 1 −0.03 1 0.1 0.1 −0.01 0.01 0.03 −0.1 10 10 −0.03 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * g) w NH 4xCONTROL−4xGWLOG h) w SH 4xCONTROL−4xGWLOG m,DF m,DF 0.01 0.01 −0.01 0.3 −0.01 0.1 0.1 −0.3 −0.1 −0.03 −0.1 0.1 1 1 −0.03 0.1 −0.1 10 10 0.3 0.01 100 100 J F M A M J J A S O N D J F M A M J J A S O N D Figure 16. As in Figure 12, but for 4xCONTROL and 4xGWLOG runs. In the tropics, the situation is not as clear, and it is more difﬁcult to deliver a clear message. We ﬁnd signiﬁ- cant changes in the amplitude and period of the QBO between 4xCONTROL and 4xGWLOG. In both runs the QBO period decreases drastically and the amplitude of the eastward phase is reduced. Also, in 4xGWLOG the oscillation of the winds is lost below 20 hPa, remaining in westward phase (see supporting information). Nonetheless, different settings and tuning of a given parameterization may have different - and somewhat inconsistent- QBO responses in simulations of a warmer climate [Schirber et al., 2015], so we do not consider that those changes be due to a crucial role of the GW sources. 5. Summary and Concluding Remarks In this work, we have presented the mean climate and variability of the middle atmosphere in the new ver- sion of the LMDz general circulation model. A novel characteristic of LMDz is that it includes a set of gravity wave parameterizations where the emitted stress is linked to the source characteristics, namely ﬂow over topography, convection, and fronts and jet imbalances. In general, LMDz with source-related GWD (i.e., CONTROL) shows good climatology and interannual variability as compared to ERA-Interim. Some well- known biases persist, as the lack of an equatorward tilt with height of the southern stratospheric polar night jet, and too strong summer easterly jets in both hemispheres. The model presents good statistics of sudden DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1523 −0.1 −0.03 −0.03 0.03 0.01 −0.01 −0.03 0.01 0.3 −0.03 0.03 0.01 −0.01 0.01 0.1 −0.03 −0.03 0.03 0.01 −0.1 −0.01 0.3 Pressure (hPa) Pressure (hPa) Pressure (hPa) Pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 stratospheric warmings, and internally generates a QBO in the tropical stratosphere with reasonable ampli- tude and mean period, as described in more detail by Lott and Guez [2013]. There are two major features that are reproduced in nonorographic GW parameterizations when the launched stress is tied to the intensity of the sources. The ﬁrst one is a realistic representation of momen- tum ﬂux intermittency; the second one is an annual cycle of the stress due to that in the GW sources. Regarding the reproduction of momentum ﬂux intermittency, de la Camara et al. [2016] have shown that it is a crucial factor in order to simulate the stratospheric ﬁnal warming in the SH with a realistic timing. In the present paper, we investigate the possible impact of the source-induced GW stress annual cycle on the mid- dle atmospheric circulation. For this, we have conducted additional experiments in which the intermittency is prescribed, but the launched GW stress is uncoupled from the sources (i.e., GWLOG). Our results show that including GW sources changes the seasonality of the middle atmospheric GW drag. The seasonality of the GW stress is ﬁltered out quite rapidly with altitude, and a quite reasonable midlati- tude climate can be obtained with a scheme without sources and prescribing the GW intermittency. Regarding the global Brewer-Dobson circulation, the GWD differences between CONTROL and GWLOG lead to changes in the seasonality of the Brewer-Dobson circulation that can be up to 10% in the NH, while in the SH the GWD variations are compensated by the resolved wave forcing. Our warmer climate simulations show that the GWD has a stronger seasonality when linked to the GW sources, but we do not ﬁnd any dra- matic ampliﬁcation of climate change in the troposphere or the stratosphere due to the changes in nonoro- graphic GWD speciﬁcation. This result is consistent with Sigmond and Scinocca [2010], who found that the inﬂuence of the basic state on the circulation response to a warmer climate is much larger than the inﬂu- ence of changes in the orographic GW drag. Our conclusions here are nevertheless based on a limited set of experiments, concerning zonal and time mean diagnostics. The results we ﬁnd regarding the midlatitude variability seem to indicate a stronger sensitivity to the GW annual cycle. Longer runs are needed to address this issue in present and future climate. Acknowledgments References The authors thank the comments from Abalos, M., W. J. Randel, and E. Serrano (2012), Variability in upwelling across the tropical tropopause and correlations with tracers in the two anonymous reviewers. This study lower stratosphere, Atmos. Chem. Phys., 12(23), 11,505–11,517, doi:10.5194/acp-12-11505-2012. was funded by the ANR/JPI-Climate/ Andrews, D. G., J. R. Holton, and C. B. Leovy (1987), Middle Atmosphere Dynamics, 489 pp., Academic Press, San Diego, Calif. Belmont Forum project GOTHAM Anstey, J. A., J. F. Scinocca, and M. Keller (2016), Simulating the QBO in an atmospheric general circulation model: Sensitivity to resolved (ANR-15-JCLI-0004-01), and the and parameterized forcing, J. Atmos. Sci., 73, 1649–1664, doi:10.1175/JAS-D-15-0099.1. European project ARISE2 (Horizon Beres, J. H. (2005), Implementation of a gravity wave source spectrum parameterization dependent on the properties of convection in the 2020, GAN653980). The data used in Whole Atmosphere Community Climate Model (WACCM), J. Geophys. Res., 110, D10108, doi:10.1029/2004JD005504. this study are available upon request Bushell, A. C., N. Butchart, S. H. Derbyshire, D. R. Jackson, G. J. Shutts, S. B. Vosper, and S. Webster (2015), Parameterized gravity wave to [email protected]. A. de la C amara momentum ﬂuxes from sources related to convection and large-scale precipitation processes in a global atmosphere model, J. Atmos. has been supported by the Advanced Sci., 72(11), 4349–4371, doi:10.1175/JAS-D-15-0022.1. Study Program at The National Center Butchart, N., et al. (2011), Multimodel climate and variability of the stratosphere, J. Geophys. Res., 116, D05102, doi:10.1029/2010JD014995. for Atmospheric Research (NCAR), Charlton, A. J., and L. M. Polvani (2007), A new look at stratospheric sudden warmings. Part I: Climatology and modeling benchmarks, which is sponsored by the U.S. J. Clim., 20(3), 449–469, doi:10.1175/JCLI3996.1. National Science Foundation. Charron, M., and E. Manzini (2002), Gravity waves from fronts: Parameterization and middle atmosphere response in a general circulation model, J. Atmos. Sci., 59(5), 923–941, doi:10.1175/1520-0469(2002)059¡0923:GWFFPA>2.0.CO;2. de la C amara, A., and F. Lott (2015), A parameterization of the gravity waves emitted by fronts and jets, Geophys. Res. Lett., 42, 2071–2078, doi:10.1002/GL063298. de la Camara, 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/JD022002. de la Camara, A., F. Lott, V. Jewtoukoff, R. Plougonven, and A. Hertzog (2016), On the gravity wave forcing during the southern stratospher- ic ﬁnal warming in LMDz, J. Atmos. Sci., 73, 3213–3226, doi:10.1175/JAS-D-15-0377.1. Eckermann, S. D. (2011), Explicitly stochastic parameterization of nonorographic gravity wave drag, J. Atmos. Sci., 68, 1749–1765, doi: 10.1175/JAS3684.1. Fritts, D. C. (1989), A review of gravity wave saturation processes, effects, and variability in the middle atmosphere, Pure Appl. Geophys., 130, 343–371, doi:10.1007/BF00874464. Fritts, D. C., and M. J. Alexander (2003), Gravity wave dynamics and effects in the middle atmosphere, Rev. Geophys., 41(1), 1003, doi: 10.1029/2001RG000106. Haynes, P. H., C. J. Marks, M. E. McIntyre, T. G. Shepherd, K. P. Shine (1991), On the ‘‘Downward Control’’ of extratropical diabatic circulations by eddy-induced mean zonal forces, J. Atmos. Sci., 48, 651–978. Hines, C. O. (1997), Doppler-spread parameterization of gravity-wave momentum deposition in the middle atmosphere. Part 2: Broad and quasi monochromatic spectra, and implementation, J. Atmos. Solar Terr. Phys., 59(4), 387–400, doi:10.1016/S1364-6826(96)00080-6. 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. 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. Weather Rev., 127(5), 788–801, doi:10.1175/1520-0493(1999)127¡0788:AOSBIA>2.0.CO;2. DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1524 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 Lott, F., and L. Guez (2013), A stochastic parameterization of the gravity waves due to convection and its impact on the equatorial strato- sphere, 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, Clim. Dyn., 25(7-8), 851–868, doi:10.1007/s00382-005-0064-x. Lott, F., R. Plougonven, and J. Vanneste (2010), Gravity waves generated by sheared potential vorticity anomalies, J. Atmos. Sci., 67(1), 157–170, doi:10.1175/2009JAS3134.1. Lott, F., R. Plougonven, and J. Vanneste (2012a), Gravity waves generated by sheared three-dimensional potential vorticity anomalies, J. Atmos. Sci., 69(7), 2134–2151, doi:10.1175/JAS-D-11-0296.1. Lott, F., L. Guez, and P. Maury (2012b), A stochastic parameterization of non-orographic gravity waves: Formalism and impact on the equa- torial stratosphere, Geophys. Res. Lett., 39, L06807, doi:10.1029/GL051001. Manzini, E., N. A. McFarlane, and C. McLandress (1997), Impact of the Doppler spread parameterization on the simulation of the middle atmosphere circulation using the MA/ECHAM4 general circulation model, J. Geophys. Res., 102, 25,751–25,762, doi:10.1029/97JD01096. Manzini, E., et al. (2014), Northern winter climate change: Assessment of uncertainty in CMIP5 projections related to stratosphere- troposphere coupling, J. Geophys. Res. Atmos., 119, 7979–7998, doi:10.1002/2013JD021403. Maury, P., F. Lott, L. Guez, and J.-P. Duvel (2013), Tropical variability and stratospheric equatorial waves in the IPSLCM5 model, Clim. Dyn., 40(9-10), 2331–2344, doi:10.1007/s00382-011-1273-0. McFarlane, N. A. (1987), The effect of orographically excited gravity wave drag on the general circulation of the lower stratosphere and tro- posphere, J. Atmos. Sci., 44, 1775–1800, doi:10.1175/1520-0469(1987)044¡1775:TEOOEG>2.0.CO;2. McLandress, C., T. G. Shepherd, S. Polavarapu, and S. R. Beagley (2012), Is missing orographic gravity wave drag near 60S the cause of the stratospheric zonal wind biases in chemistry-climate models?, J. Atmos. Sci., 69(3), 802–818, doi:10.1175/JAS-D-11-0159.1. 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, Q. J. R. Meteorol. Soc., 112(474), 1001–1039, doi:10.1002/ qj.49711247406. Randel, W. J., R. R. Garcia, and F. Wu (2002), Time-dependent upwelling in the tropical lower stratosphere estimated from the zonal-mean momentum budget, J. Atmos. Sci., 59, 2141–2152, doi:10.1175/1520-0469(2002)059¡2141:TDUITT>2.0.CO;2. Richter, J. H., F. Sassi, and R. R. Garcia (2010), Toward a physically based gravity wave source parameterization in a general circulation mod- el, J. Atmos. Sci., 67(1), 136–156, doi:10.1175/2009JAS3112.1. Rind, D., R. Suozzo, N. K. Balachandran, A. Lacis, and G. Russell (1988), The GISS global climate-middle atmosphere model. Part i: Model structure and climatology, J. Atmos. Sci., 45(3), 329–370, doi:10.1175/1520-0469(1988)045¡0329:TGGCMA>2.0.CO;2. Sassi, F., R. R. Garcia, B. A. Boville, and H. Liu (2002), On temperature inversions and the mesospheric surf zone, J. Geophys. Res., 107(D19), 4380, doi:10.1029/2001JD001525. Scaife, A. A., et al. (2012), Climate change projections and stratosphere-troposphere interaction, Clim. Dyn., 38(9-10), 2089–2097, doi: 10.1007/s00382-011-1080-7. Schirber, S., E. Manzini, and M. J. Alexander (2014), A convection-based gravity wave parameterization in a general circulation model: Implementation and improvements on the QBO, J. Adv. Model. Earth Syst., 6, 264–279, doi:10.1002/2013MS000286. Schirber, S., E. Manzini, T. Krismer, and M. Giorgetta (2015), The quasi-biennial oscillation in a warmer climate: Sensitivity to different gravity wave parameterizations, Clim. Dyn., 45, 825–836, doi:10.1007/s00382-014-2314-2. Sigmond, M., and J. F. Scinocca (2010), The inﬂuence of the basic state on the northern hemisphere circulation response to climate change, J. Clim., 23(6), 1434–1446, doi:10.1175/2009JCLI3167.1. Song, I.-S., and H.-Y. Chun (2005), Momentum ﬂux spectrum of convectively forced internal gravity waves and its application to gravity wave drag parameterization: Part i. theory, J. Atmos. Sci., 62, 107–124. Stephan, C., and M. J. Alexander (2015), Realistic simulations of atmospheric gravity waves over the continental U.S. using precipitation radar data, J. Adv. Model. Earth Syst., 7, 823–835, doi:10.1002/2014MS000396. Warner, C. D., and M. E. McIntyre (1996), On the propagation and dissipation of gravity wave spectra through a realistic middle atmo- sphere, J. Atmos. Sci., 53, 3213–3235. 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. Zulicke, € C., and D. Peters (2008), Parameterization of strong stratospheric inertia-gravity waves forced by poleward-breaking Rossby waves, Mon. Weather Rev., 136(1), 98–119, doi:10.1175/2007MWR2060.1. DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1525 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Advances in Modeling Earth Systems Wiley http://www.deepdyve.com/lp/wiley/climatology-of-the-middle-atmosphere-in-lmdz-impact-of-source-related-0nvaXPauLm

Loading next page...

References (47)

F. Lott, L. Guez (2013)
A stochastic parameterization of the gravity waves due to convection and its impact on the equatorial stratosphere
Journal of Geophysical Research: Atmospheres, 118
C. Zülicke, D. Peters (2008)
Parameterization of Strong Stratospheric Inertia Gravity Waves Forced by Poleward-Breaking Rossby Waves
Monthly Weather Review, 136
C. Hines (1991)
The Saturation of Gravity Waves in the Middle Atmosphere. Part II: Development Of Doppler-Spread Theory
Journal of the Atmospheric Sciences, 48
H. Chun, In‐Sun Song, T. Horinouchi (2005)
Momentum flux spectrum of convectively forced gravity waves : Can diabatic forcing be a proxy for convective forcing?
Journal of the Atmospheric Sciences, 62
J. Richter, F. Sassi, R. Garcia (2010)
Toward a Physically Based Gravity Wave Source Parameterization in a General Circulation Model
Journal of the Atmospheric Sciences, 67
A. Cámara, F. Lott, V. Jewtoukoff, R. Plougonven, A. Hertzog (2016)
On the Gravity Wave Forcing during the Southern Stratospheric Final Warming in LMDZ
Journal of the Atmospheric Sciences, 73
A. Charlton, L. Polvani, J. Perlwitz, F. Sassi, E. Manzini, K. Shibata, S. Pawson, J. Nielsen, D. Rind (2007)
A New Look at Stratospheric Sudden Warmings. Part II: Evaluation of Numerical Model Simulations
Journal of Climate, 20
T. Imamura (2006)
Meridional Propagation of Planetary-Scale Waves in Vertical Shear: Implication for the Venus Atmosphere
Journal of the Atmospheric Sciences, 63
S. Eckermann (2011)
Explicitly Stochastic Parameterization of Nonorographic Gravity Wave Drag
Journal of the Atmospheric Sciences, 68
A. Charlton, L. Polvani (2007)
A New Look at Stratospheric Sudden Warmings. Part I: Climatology and Modeling Benchmarks
Journal of Climate, 20
T. Palmer, G. Shutts, R. Swinbank (1986)
Alleviation of a systematic westerly bias in general circulation and numerical weather prediction models through an orographic gravity wave drag parametrization
Quarterly Journal of the Royal Meteorological Society, 112
C. Warner, M. McIntyre (1996)
On the Propagation and Dissipation of Gravity Wave Spectra through a Realistic Middle Atmosphere
Journal of the Atmospheric Sciences, 53
D. Fritts (1989)
A review of gravity wave saturation processes, effects, and variability in the middle atmosphere
pure and applied geophysics, 130
C. Hines (1997)
Doppler-spread parameterization of gravity-wave momentum deposition in the middle atmosphere. Part 2: Broad and quasi monochromatic spectra, and implementation
Journal of Atmospheric and Solar-Terrestrial Physics, 59
J. Anstey, J. Scinocca, Martin Keller (2016)
Simulating the QBO in an Atmospheric General Circulation Model: Sensitivity to Resolved and Parameterized Forcing
Journal of the Atmospheric Sciences, 73
F. Lott, L. Guez, P. Maury (2012)
A stochastic parameterization of non‐orographic gravity waves: Formalism and impact on the equatorial stratosphere
Geophysical Research Letters, 39
F. Sassi, R. Garcia, B. Boville, Huixin Liu (2001)
On temperature inversions and the mesospheric surf zone
Journal of Geophysical Research, 107
J. Beres, R. Garcia, B. Boville, F. Sassi (2005)
Implementation of a gravity wave source spectrum parameterization dependent on the properties of convection in the Whole Atmosphere Community Climate Model (WACCM)
Journal of Geophysical Research, 110
M. Charron, E. Manzini (2002)
Gravity Waves from Fronts: Parameterization and Middle Atmosphere Response in a General Circulation Model
Journal of the Atmospheric Sciences, 59
L. Jourdain, S. Bekki, F. Lott, F. Lefévre (2008)
The coupled chemistry-climate model LMDz-REPROBUS: description and evaluation of a transient simulation of the period 1980–1999
Annales Geophysicae, 26
C. Hines (1997)
Doppler-spread parameterization of gravity-wave momentum deposition in the middle atmosphere. Part 1: Basic formulation
Journal of Atmospheric and Solar-Terrestrial Physics, 59
E. Manzini, N. McFarlane, C. McLandress (1997)
Impact of the Doppler spread parameterization on the simulation of the middle atmosphere circulation using the MA/ECHAM4 general circulation model
Journal of Geophysical Research, 102
P. Maury, F. Lott, L. Guez, J. Duvel (2012)
Tropical variability and stratospheric equatorial waves in the IPSLCM5 model
Climate Dynamics, 40
C. Stephan, M. Alexander (2015)
Realistic simulations of atmospheric gravity waves over the continental U.S. using precipitation radar data
Journal of Advances in Modeling Earth Systems, 7
A. Cámara, F. Lott (2015)
A parameterization of gravity waves emitted by fronts and jets
Geophysical Research Letters, 42
A. Cámara, F. Lott, A. Hertzog (2014)
Intermittency in a stochastic parameterization of nonorographic gravity waves
Journal of Geophysical Research: Atmospheres, 119
In‐Sun Song, H. Chun (2005)
Momentum Flux Spectrum of Convectively Forced Internal Gravity Waves and Its Application to Gravity Wave Drag Parameterization. Part I: Theory
Journal of the Atmospheric Sciences, 62
M. Abalos, W. Randel, E. Serrano (2012)
Variability in upwelling across the tropical tropopause and correlations with tracers in the lower stratosphere
Atmospheric Chemistry and Physics, 12
F. Lott, L. Fairhead, F. Hourdin, P. Levan (2005)
The stratospheric version of LMDz: dynamical climatologies, arctic oscillation, and impact on the surface climate
Climate Dynamics, 25
E. Manzini, A. Karpechko, J. Anstey, M. Baldwin, R. Black, C. Cagnazzo, N. Calvo, A. Charlton-Perez, B. Christiansen, P. Davini, E. Gerber, M. Giorgetta, L. Gray, S. Hardiman, Y.‐Y. Lee, D. Marsh, B. McDaniel, A. Purich, Adam Scaife, D. Shindell, S. Son, S. Watanabe, G. Zappa (2014)
Northern winter climate change: Assessment of uncertainty in CMIP5 projections related to stratosphere‐troposphere coupling
Journal of Geophysical Research: Atmospheres, 119
S. Schirber, E. Manzini, Thomas Krismer, M. Giorgetta (2015)
The quasi-biennial oscillation in a warmer climate: sensitivity to different gravity wave parameterizations
Climate Dynamics, 45
P. Haynes, M. McIntyre, T. Shepherd, C. Marks, K. Shine (1991)
On the “Downward Control” of Extratropical Diabatic Circulations by Eddy-Induced Mean Zonal Forces
Journal of the Atmospheric Sciences, 48
D. G. Andrews, J. R. Holton, C. B. Leovy (1987)
Middle Atmosphere Dynamics
L. Wilcox, A. Charlton-Perez (2013)
Final warming of the Southern Hemisphere polar vortex in high‐ and low‐top CMIP5 models
Journal of Geophysical Research: Atmospheres, 118
Adam Scaife, T. Spangehl, D. Fereday, U. Cubasch, U. Langematz, H. Akiyoshi, S. Bekki, P. Braesicke, N. Butchart, M. Chipperfield, A. Gettelman, S. Hardiman, M. Michou, E. Rozanov, T. Shepherd (2012)
Climate change projections and stratosphere–troposphere interaction
Climate Dynamics, 38
C. McLandress, Theodore Shepherd (2012)
Is Missing Orographic Gravity Wave Drag near 60°S the Cause of the Stratospheric Zonal Wind Biases in Chemistry–Climate Models?
Journal of the Atmospheric Sciences, 69
F. Lott, R. Plougonven, J. Vanneste (2010)
Gravity Waves Generated by Sheared Potential Vorticity Anomalies
Journal of the Atmospheric Sciences, 67
S. Schirber, E. Manzini, M. Alexander (2014)
A convection‐based gravity wave parameterization in a general circulation model: Implementation and improvements on the QBO
Journal of Advances in Modeling Earth Systems, 6
D. Fritts, M. Alexander (2003)
Gravity wave dynamics and effects in the middle atmosphere
Reviews of Geophysics, 41
D. Rind, R. Suozzo, N. Balachandran, A. Lacis, G. Russell (1988)
The GISS Global Climate-Middle Atmosphere Model. Part I: Model Structure and Climatology
Journal of the Atmospheric Sciences, 45
M. Sigmond, J. Scinocca (2010)
The Influence of the Basic State on the Northern Hemisphere Circulation Response to Climate Change
Journal of Climate, 23
A. Bushell, N. Butchart, S. Derbyshire, D. Jackson, G. Shutts, S. Vosper, S. Webster (2015)
Parameterized Gravity Wave Momentum Fluxes from Sources Related to Convection and Large-Scale Precipitation Processes in a Global Atmosphere Model
Journal of the Atmospheric Sciences, 72
N. Butchart, A. Charlton-Perez, I. Cionni, S. Hardiman, P. Haynes, K. Krüger, P. Kushner, P. Newman, S. Osprey, J. Perlwitz, M. Sigmond, Lei Wang, H. Akiyoshi, J. Austin, S. Bekki, A. Baumgaertner, P. Braesicke, C. Brühl, M. Chipperfield, M. Dameris, S. Dhomse, V. Eyring, R. Garcia, H. Garny, P. Jöckel, J. Lamarque, M. Marchand, M. Michou, O. Morgenstern, Tetsu Nakamura, S. Pawson, D. Plummer, J. Pyle, E. Rozanov, J. Scinocca, T. Shepherd, K. Shibata, D. Smale, H. Teyssèdre, W. Tian, D. Waugh, Y. Yamashita (2011)
Multimodel climate and variability of the stratosphere
Journal of Geophysical Research, 116
N. McFarlane (1987)
The Effect of Orographically Excited Gravity Wave Drag on the General Circulation of the Lower Stratosphere and Troposphere
Journal of the Atmospheric Sciences, 44
F. Lott, R. Plougonven, J. Vanneste (2012)
Gravity waves generated by sheared three-dimensional potential vorticity anomalies
Journal of the Atmospheric Sciences, 69
W. Randel, R. Garcia, F. Wu (2002)
Time-dependent upwelling in the tropical lower stratosphere estimated from the zonal-mean momentum budget
Journal of the Atmospheric Sciences, 59
F. Lott (1999)
Alleviation of Stationary Biases in a GCM through a Mountain Drag Parameterization Scheme and a Simple Representation of Mountain Lift Forces
Monthly Weather Review, 127

Publisher: Wiley
Copyright: © 2016. American Geophysical Union. All Rights Reserved.
ISSN: 1942-2466
eISSN: 1942-2466
DOI: 10.1002/2016MS000753
Publisher site: See Article on Publisher Site

Abstract

PUBLICATIONS Journal of Advances in Modeling Earth Systems RESEARCH ARTICLE Climatology of the middle atmosphere in LMDz: Impact of 10.1002/2016MS000753 source-related parameterizations of gravity wave drag 1 2 1 Key Points: A. de la Camara , F. Lott , and M. Abalos Good stratospheric climatology using source-related GW parameterizations 1 2 National Center for Atmospheric Research, Boulder Colorado, USA, Laboratoire de Meteorologie Dynamique, IPSL and Gravity wave sources have an impact the annual cycle in the middle CNRS, Ecole Normale Superieure, Paris, France atmosphere Climate change is not signiﬁcantly affected by changes in GW sources Abstract Gravity wave (GW) parameterizations control the mean state and variability of the middle atmosphere in present-day climate models. The most recent parameterizations relate the GWs to their non- Supporting Information: orographic sources (fronts and convection), which impacts the annual cycle of the GW drag, and makes the Figure S1 GWs respond to the changing climate. These issues are addressed using the Laboratoire de Meteorologie Dynamique Zoom (LMDz) climate model, showing ﬁrst a climatology of the middle atmosphere in the pres- Correspondence to: ence of nonorographic GW sources. The model performance is comparable with that documented in earlier A. de la C amara, [email protected] model versions, illustrating that there are no major difﬁculties in including nonorographic GW sources in models. A twin experiment where the parameterization of GWs has no link with the nonorographic sources Citation: is also performed. Provided that in the twin experiment the launched GW stress is very intermittent, its cli- de la C amara, A., F. Lott, and M. Abalos matology compares reasonably well with the experiment with sources. This illustrates that GW intermittency (2016), Climatology of the middle is a key factor in GW dynamics, but also that the dynamical ﬁltering of the waves by the background ﬂow atmosphere in LMDz: Impact of source-related parameterizations of strongly modulates the signiﬁcance of the sources. Some impacts of having GW sources on the annual cycle gravity wave drag, J. Adv. Model. Earth of the zonal mean circulation of the middle atmosphere are nevertheless evident. In a changing climate, the Syst., 8, 1507–1525, doi:10.1002/ impact of introducing GW sources also seems to be substantially mitigated by the dynamical ﬁltering. The 2016MS000753. experiments and diagnostics are nevertheless limited in time and to the averaged climatology, respectively, calling for longer tests to measure the impacts on the atmospheric low frequency variability. Received 4 JUL 2016 Accepted 10 SEP 2016 Accepted article online 19 SEP 2016 Published online 6 OCT 2016 1. Introduction The representation of gravity waves (GW) is critical for the proper representation of the circulations of both the troposphere and the middle atmosphere in general circulation models (GCM). Orographic GWs were the ﬁrst to be parameterized, their effects helping to reduce biases in the upper tropospheric and lower stratospheric jets [e.g., Palmer et al., 1986; McFarlane, 1987; Lott, 1999]. Nonorographic GWs have been incor- porated thereafter, aiming at reducing very large biases in the stratosphere and mesosphere [e.g., Manzini et al., 1997; Sassi et al., 2002; Lott et al., 2005]. Contrarily to the orographic GWs, for which the source mechanisms are relatively well understood, the mechanisms exciting the nonorographic GWs are less evident [Fritts and Alexander, 2003]. For this reason, early parameterizations of the nonorographic GWs have no relation with their sources. Among these param- eterizations, the so-called ‘‘globally spectral’’ ones [Hines, 1997; Warner and McIntyre, 1996] assume that the GWs follow a saturated spectra, somehow in agreement with observations [Fritts, 1989, and references therein]. The good performance of the models that use these schemes [e.g., Lott et al., 2005, and references therein] witnesses that, for gravity waves, the dynamical ﬁltering due to the air density decrease with alti- tude and the vertical variations of the large scale winds play a central role determining the GW drag (GWD). V C 2016. The Authors. The globally spectral schemes are also used for the practical reason that they permit the treatment of a This is an open access article under the large ensemble of waves at a reasonable numerical cost. Nevertheless, the absence of sources in GW param- terms of the Creative Commons eterizations limit their potential calibration with the growing number of in situ and satellite observations, Attribution-NonCommercial-NoDerivs and is maybe a cause for systematic errors, at least in the Southern Hemisphere spring [McLandress et al., License, which permits use and 2012; de la Camara et al., 2016]. distribution in any medium, provided the original work is properly cited, the Consequently, many efforts have been made from theoretical, observational and modeling perspectives to use is non-commercial and no understand the mechanisms generating nonorographic GWs. As a result, many climate models now include modiﬁcations or adaptations are made. parameterizations of GWs generated by convection [Beres, 2005; Song and Chun, 2005; Lott and Guez, 2013; DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1507 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 Schirber et al., 2014; Bushell et al., 2015] and by fronts [Rind et al., 1988; Charron and Manzini, 2002; Richter et al., 2010] or planetary wave breaking [Zulicke and Peters, 2008]. Many of these parameterizations prefer to adopt a ‘‘multiwave’’ approach rather than a globally spectral one to treat the GWs. For the convective waves this is because it is quite easy to include a diabatic heating into a GW linear equation, and for the fronts an ageostrophic momentum forcing could play a similar role. This is the approach followed by the parameterizations that evaluate a frontogenesis function to identify where the GWs are excited [Rind et al., 1988; Charron and Manzini, 2002; Richter et al., 2010], a step that is quite demanding technically. For this rea- son, and also because there are no closed theories relating such an ‘‘ageostrophic forcing’’ to the GWs potentially produced, de la Camara and Lott [2015] use a simple relation between GWs and fronts that is based on theoretical results on GW emission from potential vorticity anomalies in sheared ﬂows [Lott et al., 2010, 2012a]. Interestingly, de la Camara et al. [2016] have recently demonstrated that the GW intermittency resulting from the introduction of sources (convection and fronts) is signiﬁcant to predict well the timing of the Southern Hemisphere stratospheric ﬁnal warming. It is also worthwhile to recall that by using stochastic techniques, the multiwave methods can be made much more computationally efﬁcient than initially thought (see discussions in Eckermann [2011] and Lott et al. [2012b]). A fundamental motivation to relate the GWs to their potential sources is that these sources can have an annual cycle and change when the climate changes. It is therefore important to test if this can affect the model annual cycle in the middle atmosphere and to analyze if the changes in the GW sources impact the prediction of the future climate. To have a more thorough understanding of their impact, a longer term objective is to include the GW sources in some of the models participating in the next climate model inter- comparison project (CMIP6). This is the approach followed by the Laboratoire de M et eorologie Dynamique Zoom (LMDz) GCM where all the parameterized GWs will be related to their sources, e.g., mountains, con- vection and fronts. This is in opposition with CMIP5 [Lott et al., 2005], where the LMDz model used the Hines [1997] globally spectral scheme to parameterize the nonorographic GWs. The ﬁrst purpose of this paper is therefore to carefully analyze the model middle atmospheric climate and variability, and to see if the frontal and convective GWs can do at least as well as the Hines [1997] uniform background of waves. The second is to start testing if having time and space varying sources inﬂuences the troposphere and middle atmosphere climate. The paper is organized as follows. Section 2 presents the LMDz model and summarizes the source-related GW parameterizations. Section 3 validates the model climatology and variability of the middle atmosphere against the European Centre for Middle-range Weather Forecast Interim Reanalysis (ERA-Interim), while sec- tion 4 addresses the impact of introducing nonorographic GW sources in the parameterizations. The main conclusions are given in section 5. 2. Model Description The version of LMDz we use has a 3.758 3 1.8758 longitude-latitude grid, and 71 levels in the vertical with the top at 0.01 hPa. Its vertical resolution is around 1 km in the lower stratosphere. The results shown are from a 20 year simulation (referred to as CONTROL), forced with climatological ﬁelds of sea surface tempera- ture, sea ice, soil temperature and composition over land. These climatological ﬁelds are averages over the period 1979–2005, as are the ozone ﬁelds, which are those predicted by the LMDz-Reprobus coupled climate-chemistry model [Jourdain et al., 2008]. 2.1. Gravity Wave Drag Parameterizations LMDz uses three distinct GWD parameterizations that account for GWs generated by topography [Lott, 1999], convection [Lott and Guez, 2013], and fronts [de la Camara and Lott, 2015]. The parameterizations of nonorographic GWs are based on the stochastic approach introduced by Eckermann [2011] and Lott et al. [2012b], and consists in sampling randomly the GW spectrum by launching 8 monochromatic waves at each grid point and ‘‘physical’’ time step (e.g., every 30 min). The waves chosen are purely zonal, and their zonal wavenumber k is chosen randomly between two extrema corresponding to wavelength between 1 km and 300 km and using a uniform statistics. The phase speed is chosen randomly according to a Gauss- –1 ian distribution with 40 ms standard deviation and centered on the wind velocity at the emission level (500 hPa for convective waves, 900 hPa for frontal waves). DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1508 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 As the sensitivity to the nonorographic sources is our ﬁrst objective, we next recall how this is done in LMDz (for the other aspects, like the treatment of the breaking, or the statistical superposition of the waves see Lott and Guez [2013] and de la Camara and Lott [2015]). For convective waves, the emitted GW stress at the launching altitude (z ) is: 2 2 2 2 2m Dz ~ ~ z RL jkj e k l W F 5q G P ; (1) c0 conv 3 q Hc NX r p jjkjj where q is the density at a reference level, G is a tunable, dimensionless parameter of order 1 (we take r c0 ; G 51:75), Dz a tunable characteristic depth of the heating source (we take Dz51 km), R is the ideal gas c0 constant, L is the latent heat of condensation, H5 7 km is the stratospheric scale height, c is the speciﬁc W p heat at constant pressure, k is the horizontal wavenumber vector, m is the vertical wavenumber 2 2 2 2 ~ ~ ~ (m 5N jkj =X ), N is the buoyancy frequency, X5x2k U is the intrinsic frequency (with x the ground- based frequency, and U the horizontal wind vector), and P is the grid-scale precipitation. Therefore, equa- tion (1) translates the gridscale precipitation into a subgrid scale GW stress. For frontal waves, the emitted GW stress is [see de la Camara and Lott, 2015]: z 0 ~ top Nðz Þ z dz k l 2p 0 0 2 0 0 0 ~ Uz ðz Þ F 5G q ðz ÞNðz Þf ðz Þe dz ; (2) f 0 fron 0 4f jjkjj where z is the top of the model, G is a tunable, dimensionless parameter of order 1 (we take G 52), dz top f 0 f 0 2z=H is a typical vertical depth of the vorticity anomaly (set to 1 km), f is the Coriolis parameter, q 5q e is the 0 r reference state density, f is the grid-scale relative vorticity, and U is the vertical shear. As we see equation (2) translates the resolved dynamics (gridscale vorticity and stability conditions) into a subgrid scale GW stress. 2.2. Characteristics of the Nonorographic gravity waves Figure 1 shows the annual cycle of eastward and westward momentum ﬂux (MF) at the launching altitude (i.e., convective GW stress at 500 hPa plus frontal GW stress at 900 hPa), 100 hPa and 1 hPa. At the level of emission (Figures 1e and 1f), the band of high MF in the tropics is due to the convectively generated GWs. The bands in the midlatitudes of both hemispheres are mainly due to frontal GWs, although convective GWs also contribute. The emitted MF is almost similar in amplitude for eastward and westward MF, and both exhibit a pronounced annual cycle. The tropical band migrates northward and gets stronger during the northern summer, consistent with the behavior of precipitation in the model (not shown). In the midlati- tudes, the emitted MF is weaker than in the tropics and presents higher values in winter of both hemi- spheres, consistent with stronger baroclinicity. The effect of wind ﬁltering on the GW propagation is evident, the MF getting smaller and smaller when entering the stratosphere at 100 hPa and the mesosphere at 1 hPa. At 100 hPa (Figures 1c and 1d), the MF still has some pattern similarities with the emitted MF, although the wind ﬁltering modulates the annual cycle. At 1 hPa (Figures 1a and 1b) the tropical band has been ﬁltered out to a large extent. In the extra- tropics, while the annual cycle pattern of westward MF somehow resembles that of the emitted ﬂux, the pattern of eastward ﬂuxes is almost in phase opposition with that emitted, presenting larger values during the summer months in both hemispheres. It is interesting to compare the GW stress at 100 hPa with the results of Richter et al. [2010], where the authors show similar plots for convective and frontal GW stresses separately at 100 hPa in WACCM3.5 (their Figures 2 and 3). The annual cycle of the GW stress in the tropics (Figures 1c and 1d) resembles the convec- tive GW stress in Richter et al. [2010], but the magnitude in our model is smaller by a factor of 2. At mid-to- high latitudes, the GW stress at 100 hPa is also qualitatively similar to the frontal GW stress in Richter et al. [2010], and this time the magnitude is smaller in our scheme by a factor of 1.5. Such differences in stress amplitude are not surprising given that the two models are very different, and most of all the WACCM mod- el top (0.0001 hPa) is much higher than the LMDz model top (0.01 hPa), meaning that a given stress can give much larger drag near the top in the ﬁrst model than in the second. This can yield modelers to tune the launched GW stress to control the drag amplitude near the model top. DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1509 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 b) Westward MF 1 hPa CONTROL a) Eastward MF 1 hPa CONTROL 1.5 1.35 1.2 1.05 30 30 0.9 0 0.75 0.6 −30 −30 0.45 0.3 −60 −60 0.15 J F M A M J J A S O N D J F M A M J J A S O N D c) Eastward MF 100 hPa CONTROL d) Westward MF 100 hPa CONTROL 2.7 60 60 2.4 2.1 1.8 0 0 1.5 1.2 −30 −30 0.9 0.6 −60 −60 0.3 J F M A M J J A S O N D J F M A M J J A S O N D f) Westward MF emitted CONTROL e) Eastward MF emitted CONTROL 5.4 60 60 4.8 4.2 3.6 0 0 3 2.4 −30 −30 1.8 1.2 −60 −60 0.6 J F M A M J J A S O N D J F M A M J J A S O N D Figure 1. Total eastward (left column) and westward (right column) momentum ﬂux at the level of emission, 100 hPa, and 1 hPa as indicat- ed, from nonorographic gravity waves as a function of latitude and time of the year (in mPa) in CONTROL. Figure 2 presents the drag imposed on the mean ﬂow by the frontal, convective, and orographic GW parameterizations for DJF and JJA. Frontal GWs are the main contributor to the total GWD in the southern 21 –1 extratropics in both seasons, with peak values larger than 621 ms d near 608S at mesospheric levels above 0.1 hPa. Convective GWD is weaker than the frontal drag in the extratropics, but it presents relative –1 maxima (about 63–6 ms d ) near 508 latitude in both hemispheres at the highest altitudes of the mod- el, presumably associated with the location of the storm tracks. The strong dissipation of MF in the tropics between 100 and 1 hPa described in Figure 1 is not evident here due to density effects, i.e., the drag is pro- portional to the the vertical divergence of the momentum ﬂux and inversely proportional to density. Oro- –1 graphic GWD is mainly active in northern winter extratropical stratosphere, reaching 29ms d at 0.1 hPa. 3. Mean Climate and Variability of the Middle Atmosphere 3.1. Zonal Mean Climate As this paper focuses on the impact of including sources in the nonorographic GWD schemes, we have tuned these parameterizations to ensure that LMDz has a climatology at least comparable to that docu- mented in its previous stratospheric version [Lott et al., 2005]. As we shall see, the improvements in some places are obvious, like in the QBO region, whereas in the midlatitudes the effects are more neutral. Note DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1510 latitude latitude latitude Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) DJF FGWD b) DJF CGWD c) DJF OGWD 0.01 0.1 10 18 100 12 e) JJA CGWD f) JJA OGWD d) JJA FGWD 0.01 −3 −6 0.1 −9 −12 1 −15 −18 −21 −24 −27 −30 −90 −45 0 45 90 −90 −45 0 45 90 −90 −45 0 45 90 latitude latitude latitude 21 –1 Figure 2. Longitudinally averaged drag (in ms d ) from the (left column) frontal (FGWD), (middle column) convective (CGWD) and (right column) orographic (OGWD) gravity waves for DJF and JJA in CONTROL. * * a) DJF U, Ψ CONTROL b) MAM U, Ψ CONTROL res res 0.01 0.01 0.1 0.1 95 1 1 10 10 100 100 1000 1000 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 15 * d) SON U, Ψ CONTROL −5 c) JJA U, Ψ CONTROL res res −15 0.01 0.01 −25 0.1 −35 0.1 −45 1 −55 −65 −75 −85 −95 1000 1000 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 latitude latitude –1 Figure 3. Zonally averaged zonal wind proﬁles (in ms , shaded), and stream function of the residual mean meridional circulation (con- tours) in CONTROL. Magenta contours represent positive values (i.e., clockwise circulation), and cyan contours represent negative values (i.e., counter-clockwise circulation). DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1511 pressure (hPa) pressure (hPa) pressure (hPa) pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 * * a) DJF U, Ψ ERAI b) MAM U, Ψ ERAI res res 0.01 0.01 0.1 0.1 95 1 1 10 10 100 100 1000 1000 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 15 d) SON U, Ψ ERAI −5 c) JJA U, Ψ ERAI res res −15 0.01 0.01 −25 0.1 −35 0.1 −45 1 −55 −65 −75 −85 100 100 −95 1000 1000 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 latitude latitude Figure 4. As in Figure 3 but for ERA-Interim. that having a model version with GW sources and a QBO but without degrading the model in other places was an implicit objective of the present study. To illustrate this, Figure 3 shows the seasonal averages of zonal-mean zonal wind proﬁles. It shows well- –1 deﬁned polar night jets in the solstices with values up to 40 and 85 ms in the boreal and austral jet cores, –1 respectively. The summer easterly jets present maximum values of 270 ms in the subtropics at around 1 hPa, and the winds show transition conditions in the equinoxes. These zonal mean winds compare well with those corresponding to an earlier model version [Lott et al., 2005, Figure 3], but some biases that were present in the previous version of the model remain. When compared to ERAI in Figure 4 we see that the –1 largest biases in the model are in the summer easterly jets, with winds 20 ms stronger in LMDz than in ERAI. Also, the SH easterly jet in DJF splits into two parts (Figure 3a). The strength of the polar night jet is comparable in the two data sets, although the boreal jet in LMDz is weaker than in ERAI in the upper strato- sphere and lower mesosphere. This is more clearly seen in Figure 5, which speciﬁcally shows the wind speed in the jet core and its latitudinal position as a function of height during the northern and southern winters (note that Figures 5 and 6 include results for GWLOG, which will be discussed in section 4). The lati- tudinal tilt of the jets with altitude is well captured, with the exception of the southern jet in JJA. This bias is common to most climate models [Butchart et al., 2011]. The model performance in MAM and SON shows good agreement with ERAI (Figures 3 and 4). To complete the description of the zonal mean circulation, the contours in Figure 3 display the mass streamfunction in CONTROL, representing the residual mean meridional circulation W : res @W res 52q cos / v ; (3) @z where q 5q ðzÞ is the background density, / is latitude, and v is the latitudinal component of the residual 0 0 circulation in the Transformed Eulerian Mean (TEM) formalism [Andrews et al., 1987]: DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1512 pressure (hPa) pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) Strength NH Jet (DJF) b) Latitude NH Jet (DJF) 0.01 0.01 0.03 CONTROL 0.03 0.1 0.1 ERAI 0.3 0.3 GWLOG 1 1 3 3 10 10 30 30 100 100 300 300 0 25 50 75 100 20 40 60 80 c) Strength SH Jet (JJA) d) Latitude SH Jet (JJA) 0.01 0.01 0.03 0.03 0.1 0.1 0.3 0.3 1 1 3 3 10 10 30 30 100 100 300 300 0 25 50 75 100 −80 −60 −40 −20 zonal wind (m/s) latitude Figure 5. Zonal wind speed and latitude and latitude of the jet maxima (top) of the NH DJF climatology and (bottom) of the SH JJA clima- tology, for CONTROL (blue line), ERAI (green line), GWLOG (red line, see section 4). 1 @ q v h v v2 (4) q @z @h=@z In DJF and JJA (Figure 3c), the main circulation cell presents upward motions in the tropics, extending to the summer hemisphere, that reach mesospheric altitudes, and downward motions in the winter high lati- tudes. The meridional motion in the mesosphere above 1 hPa is responsible for the dynamical maintenance of winter pole temperatures much warmer than summer pole temperatures in the mesosphere (not shown). A secondary, shallow circulation cell can also be seen in the summer hemisphere lower stratosphere. All these features compare well with ERAI (Figure 4). During MAM and SON the circulation cells grow deeper in the autumn and shallower in the spring hemispheres, in good agreement with the reanalysis. a) NH Max variability (DJF) b) NH Latitude max variability (DJF) 0.01 0.01 0.03 0.03 CONTROL 0.1 0.1 ERAI 0.3 0.3 GWLOG 1 1 3 3 10 10 30 30 100 100 300 300 0 5 10 15 20 25 20 40 60 80 c) SH Max variability (JJA) d) SH Latitude max variability (JJA) 0.01 0.01 0.03 0.03 0.1 0.1 0.3 0.3 1 1 3 3 10 10 30 30 100 100 300 300 0 5 10 15 20 25 −80 −60 −40 −20 zonal wind (m/s) latitude Figure 6. Location and amplitude of the maximum interannual standard deviation of the zonal mean zonal wind (top) in the NH in DJF poleward of 458N and (bottom) in the SH in JJA poleward of 308S, for CONTROL (blue line), ERAI (green line), GWLOG (red line, see section 4). DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1513 pressure (hPa) pressure (hPa) pressure (hPa) pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) NP T 10hPa CONTROL b) NP T 10hPa, ERAI c) NP T 10hPa GWLOG J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J d) SP T 10hPa CONTROL e) SP T 10hPa, ERAI f) SP T 10hPa GWLOG J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D Figure 7. Polar temperatures (in K) at 10 hPa and 858 latitude in CONTROL, ERA-Interim, and GWLOG (see section 4) for (a, b, c) the North- ern Hemisphere, and (d, e, f) the Southern Hemisphere as a function of time of the year. 3.2. Interannual Variability To analyze the simulated variability, Figure 6 shows the amplitude and latitudinal location of maximum interannual variability of the polar night jets as a function of height, for CONTROL and ERAI. For the North- ern Hemisphere (NH) winter, the amplitude of the maximum variability is well represented below 1 hPa as compared to ERAI, but the model underestimates it in the lower mesosphere (Figure 6a). The latitudinal location of this variability is not so well represented (Figure 6b). While in ERAI the position tilts equatorward with height between 30 and 1 hPa, in LMDz it tilts poleward. This bias is common to many climate models, and needs further investigation [Butchart et al., 2011]. For the Southern Hemisphere (SH) winter, the com- parison to ERAI provides similar conclusions. A slight equatorward tilt with height of the maximum variabili- ty does appear in CONTROL below 1 hPa, but with a much steeper slope than in ERAI. The maximum variability is 10–208 poleward in CONTROL as compared to ERAI, possibly due to differences in the locations of the jet core (Figure 5d). A complementary view of the interannual variability is given by the time series of polar temperature in both hemispheres (at 858 latitude) at 10 hPa in Figure 7. The seasonal evolution and variability, as well as the inter- hemispheric contrasts, are generally well captured. However the model presents too much variability, as evi- denced by sporadic warmings in 0.5 October and November in the North CONTROL (0.76/year) Pole (Figure 7a), or the spread in ERAI (0.67/year) 0.4 temperatures in the South Pole dur- GWLOG (0.58/year) ing the austral winter/spring that are not present in the ERAI data. Figure 0.3 8 further shows an histogram of fre- quencies of major sudden warmings 0.2 (MSW) in the NH for both CONTROL and ERAI, sorted by winter month. 0.1 We have followed the method by Charlton and Polvani [2007] to iden- tify these events. The frequency of Nov Dec Jan Feb Mar events is higher in CONTROL than in ERAI (0.76/year versus 0.67/year), Figure 8. Frequency of major stratospheric sudden warmings (number of MSW per conﬁrming larger simulated vari- year) in the NH for CONTROL (blue bars), ERAI (green bars), and GWLOG (red bars, see section 4), sorted by month. The total frequency is also indicated in the ﬁgure legend. ability. Importantly, the intraseasonal DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1514 frequency of major sudden warmings Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) QBO CONTROL 1 2 3 4 5 6 7 8 9 10 b) QBO ERAI −5 −15 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 c) QBO GWLOG −25 −35 −45 1 2 3 4 5 6 7 8 9 10 year Figure 9. Zonal mean zonal wind averaged over the equatorial band 58S–58N for (a) CONTROL, (b) ERAI, and (c) GWLOG (see section 4). –1 Contour interval: 10 m s , bold grey contour indicates the zero-wind line. distribution of major warmings consistently presents higher frequencies as the winter season progresses, peaking up in February. The differences against ERAI include too high frequencies in November and February, and too low in December. Given the multiple factors inﬂuencing the occurrence of MSWs, we consider that the performance of LMDz compares well with that of the previous model version (not shown here, but see Lott et al. [2005, Figure 13]). Additionally, the model does not present a signiﬁcant delay in the simulation of the stratospheric ﬁnal warming in the SH [de la Cam ara et al., 2016], a bias that most climate models still have [e.g., Butchart et al., 2011; McLandress et al., 2012; Wilcox and Charlton-Perez, 2013]. In the tropical lower stratosphere, the QBO dominates the interannual variability of the zonal winds. Figure 9 shows the zonal winds at the Equator as a function of time and height. The model internally generates a QBO with an average period of 28 months that closely matches that in ERAI (27 months). Yet there are some discrepancies between the model and the reanalysis in Figure 9, such as wind velocities that are up –1 to 10 ms weaker in the model, especially during the westward phase. Also the QBO in CONTROL does not descend as low as it does in ERAI, and it lacks the westerlies stalling that often occurs below 30 hPa (see e.g., the years 2009–2010 in Figure 9b). The causes are multiple, but we suspect that the underestimation of the slow Kelvin waves in LMDz might play a signiﬁcant role [Maury et al., 2013]. A ﬁner vertical resolution in the lower stratosphere (1 km in this model version) might also contribute to improve the QBO simulation especially at lower levels [see Anstey et al., 2016]. The reader is referred to Lott and Guez [2013] for further details on the simulation of the QBO in LMDz and the comparison with observations. For completeness, we recall here that the QBO was absent in Lott et al. [2005]. 4. Impact of Source-Varying GWD Parameterizations In this section we evaluate the impact of including sources of nonorographic gravity waves (NGW). First, we describe the twin experiments performed, and then we will present the results, with the focus on DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1515 Pressure (hPa) Pressure (hPa) Pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 the simulated annual cycle in the middle atmosphere, and on possible impacts under future climate conditions. 4.1. Model Experiments Recent studies have shown that linking the parameterized GW amplitudes to their nonorographic sources naturally produces intermittent MFs, the probability density function of absolute momentum ﬂuxes ﬁtting a lognormal distribution [de la Camara et al., 2014; de la Camara and Lott, 2015; Stephan and Alexander, 2015]. These studies also suggest that the NGW intermittency can help reduce model biases, simply because for a given averaged launched momentum ﬂux, few large amplitude waves break at lower altitude than a large number of small amplitude waves. Therefore, to evaluate the role of the NGW sources speciﬁcally, we next replace the source terms in the convective and frontal schemes (i.e., the P and f terms in equations (1) and (2)) by random numbers produced by a lognormal distribution. The characteristics of the distribution are tuned to obtain a reasonable zonal mean climatology (see next section). This run is referred to as GWLOG. We also apply a latitudinal weighting in the modiﬁed convective GW scheme to launch larger stress in the tropics and help generate a QBO. Speciﬁcally, the latitudinal weighting function chosen is 2 30 f ð/Þ5ð0:15sin 2/11:1cos /Þ. This function has a narrow maximum at the equator and two secondary peaks at 458 latitude, qualitatively mimicking the averaged latitudinal distribution of precipitation [e.g., Lott and Guez, 2013]. The magnitudes of the maxima have been chosen ad hoc to obtain a reasonable climatolo- gy (see next section). A different potential impact of having source-related NGW schemes is that parameterized wave amplitudes will change if climate changes. To investigate this point, we perform two additional experiments. First we make a 20 year experiment, named 4xCONTROL, where the GW speciﬁcations are as in CONTROL, but increasing the CO concentrations by a factor of four, and by adding everywhere 4 K to the prescribed SST. Second we make another 20 year experiment, named 4xGWLOG, similar to 4xCONTROL but using the GW speciﬁcations of GWLOG. 4.2. Climatology of the Simulation Without GW Sources To make a fair comparison between the simulations with and without NGW sources we have tried to make them as close as possible in terms of the GW drag in the midlatitudes, the middle atmosphere jets in the midlatitudes and subtropics, and the QBO in the tropical lower stratosphere. Figure 10 displays the nonoro- graphic GWD for DJF and JJA in the CONTROL and GWLOG runs. The lognormal distributions of emitted GW stress used in GWLOG provide GWD proﬁles qualitatively similar to those in CONTROL. Gravity wave drag –1 values larger than 63ms d are found above 1 hPa in the mesosphere of the two runs. Quantitatively, the drag in CONTROL is slightly weaker than in GWLOG, particularly in the summer hemisphere. Concerning the impacts on the mean climate, we return to Figure 5 that shows the strength and location of the wintertime polar jets in GWLOG. The zonal mean climate of GWLOG is comparable to that of CON- TROL during the solstices. Although the panels in Figure 5 focus on the winter westerly jets, essentially because the GW parameterizations are ﬁrst intended to improve them, it is important to say that similari- ties are found in the midlatitudes during other seasons. Beyond the zonal means, it is much more difﬁcult to control the variability, as illustrated in Figure 6 where the variability of the jet in GWLOG is also shown. The most notorious differences appear in the SH, such as larger variability of the jet in GWLOG (Figure 6c), and the absence of the equatorward tilt with height in the jet variability below 1 hPa (Figure 6d). Still concerning the variability but coming back to the NH, Figure 8 also shows the MSW statistics for GWLOG. The mean winter frequency is reasonable (0.58 per year), but GWLOG fails in capturing the intra- seasonal distribution of the major warmings. This result may be due to chance, but it is interesting that removing the relation with the NGW sources degrades the SSW seasonality. We nevertheless need to test it with longer model simulations. On the other hand, there is some improvement in MSW frequency in December, where most climate models do not get nearly enough warmings just like the control run. Finally, and concerning the tropical region, Figure 9 shows that GWLOG also has an internally generated QBO, its period is slightly longer than in CONTROL. Above the QBO region, the semi-annual signal seems more pronounced in GWLOG. DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1516 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) DJF CONTROL b) DJF GWLOG 0.01 0.1 10 24 100 16 d) JJA GWLOG 0 c) JJA CONTROL 0.01 −4 −8 −12 0.1 −16 −20 −24 −28 −32 −36 −40 −90 −45 0 45 90 −90 −45 0 45 90 latitude latitude 21 –1 Figure 10. Zonally averaged drag from the NGW parameterizations (in ms d ), for (a, b) DJF and (c, d) JJA, from the CONTROL and 21 –1 GWLOG runs as indicated. The magenta and cyan lines indicate 10.1 and 20.1 ms d contours, respectively. 4.3. Impact on the Annual Cycle We showed in Figure 1 that the GW stress emitted in CONTROL presents a strong annual cycle, presumably due to the annual cycle of the GW sources activity. Figure 11 presents the eastward and westward NGW stress as a function of latitude and time of the year for GWLOG. As expected, a very weak seasonality appears at the altitude of emission since NGW sources are not considered in this run. At 100 hPa, seasonal differences start to show up, and at 1 hPa a strong annual cycle is present due to momentum ﬂux dissipa- tion. We can now compare this performance in GWLOG with that in CONTROL (Figure 1). At 1 hPa both east- ward and westward momentum ﬂuxes are very similar in magnitude and seasonal evolution in both runs. This contrasts with the stress at 100 hPa, where the annual cycle is much stronger and peak values are much larger for CONTROL than for GWLOG, specially for the westward direction (e.g., 2.1 versus 1.2 mPa at 508S in August). It can be interpreted then that the GW stress entering the mesosphere in our simulations is only weakly dependent on the seasonal cycle of the stress at lower altitudes, and in particular on the sea- sonal cycle introduced by the GW sources. On the other hand, this implies a distinct momentum ﬂux dissi- pation in the stratosphere between these two runs, which may result in differences of GW drag in the stratosphere. In terms of nonorographic GW drag, the difference between CONTROL and GWLOG is also signiﬁcant as illustrate the Figures 12a and 12b where the annual cycle of the drag averaged for the northern (508N– 808N) and southern (508S–808S) high latitudes, are shown. In the NH (Figure 12a), there is a band of negative differences in the lowermost stratosphere during the whole year, perhaps pointing to larger westward net stress emitted (i.e., producing a negative drag) in CONTROL than in GWLOG. Above 50 hPa, a marked sea- sonal cycle appears, with positive differences during summer and negative during winter, changing sign in the mesosphere above 0.5 hPa. In the SH (Figure 12b), there is a noticeable annual cycle in the drag up to 0.1 hPa, with positive differences in summer and negative in winter. The magnitude is also small, reaching 21 –1 up to 20.3 ms d in the upper stratosphere in JJA. The negative differences during the summer months descend throughout the season and reach the lower stratosphere by September. The impact of the GW seasonality on the annual cycle of the zonal winds at midlatitudes is not very signiﬁ- cant, consistently with the fact that we tuned GWLOG with this objective (see supporting information). The situation is somehow different if we look at the Brewer-Dobson circulation, as we show below. To evaluate DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1517 pressure (hPa) pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 b) Westward MF 1 hPa GWLOG a) Eastward MF 1 hPa GWLOG 1.5 1.35 1.2 1.05 30 30 0.9 0 0 0.75 0.6 −30 −30 0.45 0.3 −60 −60 0.15 J F M A M J J A S O N D J F M A M J J A S O N D c) Eastward MF 100 hPa GWLOG d) Westward MF 100 hPa GWLOG 2.7 60 60 2.4 2.1 1.8 0 0 1.5 1.2 −30 −30 0.9 0.6 −60 −60 0.3 J F M A M J J A S O N D J F M A M J J A S O N D f) Westward MF emitted GWLOG e) Eastward MF emitted GWLOG 5.4 4.8 4.2 3.6 0 0 3 2.4 −30 −30 1.8 1.2 −60 −60 0.6 J F M A M J J A S O N D J F M A M J J A S O N D Figure 11. As in Figure 1 but for LMDz-GWLOG. the impact of the GWs on the Brewer-Dobson circulation we use the TEM formalism, where the zonal momentum equation is given by [Andrews et al., 1987]: @u @u 2 v f1w 5DF1X (5) @t @z ^ 1 and where ðv ; w Þ are the meridional and vertical components of the TEM residual circulation, f5f2 acos / @ðu cos /Þ ~~ rF with f the Coriolis parameter, DF5 is the force applied by the resolved waves with F the @/ q a cos / Eliassen-Palm (EP) ﬂux, and X is the force applied by unresolved processes (in our case the parameterized gravity wave drag). In equation (5), the left-hand terms represent the circulation response to the forcing applied by the right-hand terms. We next evaluate the vertical motion over the high latitudes (i.e., downwel- ling), focusing on the possible response of the mean meridional circulation to the NGW drag differences between CONTROL and GWLOG. Following Randel et al. [2002] and Abalos et al. [2012], we combine equa- tion (5) and the TEM continuity equation [Andrews et al., 1987] to derive the vertical component of the resid- ual circulation: () 0 2 z=H 1 2z =H 0 2e e cos / @u ð/; z Þ 0 0 0 w ð/; zÞ5 DFð/; z Þ1Xð/; z Þ2 dz : (6) 2 0 @t z fð/; z Þ a cos /; d/ / / 1 1 We take / 560 N, / 580 N for the NH, and / 580 S, / 560 S for the SH. Note that the forcing from the 1 2 1 2 total GW drag (orographic plus nonorographic, i.e., X ) is explicitly taken into account in equation (6). DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1518 latitude latitude latitude −0.01 0.03 0.03 −0.1 −0.01 −0.03 −0.03 0.03 −0.1 0.1 −1 −3 −1 1 1 −0.3 0.3 −0 0.03 0 0.3 0.01 .01 .1 3 −0.1 − 0.1 0.0 0.01 0.01 3 −0.03 −0.01 0.01 0.03 0.1 −0.01 −0.3 0.01 0.01 − −0 0.03 .01 0.01 0.03 −0.1 −0.3 0.3 0.3 0.03 − −0.0 0.01 3 −0.1 0.1 0.01 0.01 0.1 0.03 0.03 0.01 −0.01 0.03 − −0 0.0 .01 3 −0.3 −0.1 −0.3 0.3 0.1 0.03 0.01 0.03 0.01 0.03 −1 0.1 0.3 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) NGWD NH CONTROL−GWLOG b) NGWD SH CONTROL−GWLOG 0.01 0.01 0.1 0.1 1 −3 0.3 0.3 −0.03 −0.3 0.10.3 0.3 0.01 −0.1 0.1 −1 −0.01 −0.01 1 1 0.03 0.01 0.01 −0.03 −0.01 −0.1 0.03 10 10 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * c) w NH CONTROL−GWLOG d) w SH CONTROL−GWLOG m m 0.01 0.01 0.1 0.03 −0.3 −1 0.1 0.1 0.3 −0.01 0.01 −0.3 0.3 −0.03 −0.1 0.3 1 −1 0.1 0.1 0.1 −0.1 1 1 −0.03 −0.01 0.3 10 10 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * e) w NH CONTROL−GWLOG f) w SH CONTROL−GWLOG m,NGWD m,NGWD 0.01 0.01 0.01 −0.03 −0.3 0.01 −0.1 −1 0.1 0.1 0.1 0.3 −0.01 −1 −0.1 0.03 −0.1 1 0.3 0.3 1 −0.03 −0.03 0.1 0.01 0.1 −0.01 0.03 0.03 10 10 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * g) w NH CONTROL−GWLOG h) w SH CONTROL−GWLOG m,DF m,DF 0.01 0.01 0.3 −0.01 0.1 0.1 −0.3 −0.1 −0.01 0.01 −0.1 1 −0.03 −0.01 0.03 −0.03 −0.1 0.1 −0.01 0.01 0.3 0.1 1 1 0.01 0.3 10 10 100 100 J F M A M J J A S O N D J F M A M J J A S O N D Figure 12. Differences in the (a, b) NGWD and (c–h) vertical component of the residual mean meridional circulation derived from the TEM momentum balance equation, between the CONTROL and GWLOG runs as a function of height and time of the year. The data are longitu- –1 dinally averaged over the 508–808 latitude band in both hemispheres. Contours are at 60.01, 60.03, 60.1, 60.3, 61, 63ms d for –1 NGWD, and mms for w . Light red and blue shading indicate positive and negative statistically signiﬁcant differences, respectively (Student t-test, a50.01). The panels in the second row of Figure 12 show the annual evolution of the differences of w in the north- ern and southern high-latitudes. In the NH (Figure 12c), the pattern is similar to that of the NGWD (Figure 12a). The fact that the patterns in w are found at lower altitudes than those in the NGWD is consistent with equation (6), which links the vertical motion to the drag at that level and above. We see positive differ- ences in the winter mesosphere and negative differences in the summer mesosphere (note that the regions of statistical signiﬁcant differences are somewhat limited). This means that the amplitude of the annual cycle of w is around 10% weaker in CONTROL than in GWLOG in the mesosphere, and around 10% stron- ger in the lower stratosphere (note that in the lower stratosphere the value is not statistically signiﬁcant). In the SH, the differences in w do not present a clear pattern and are barely signiﬁcant. To address whether the w differences in the NH between CONTROL and GWLOG emerge from the NGWD differences, we evaluate separately the contributions from the NGWD and from the resolved forcing (i.e., DF in equation (5)) to the vertical component of the residual circulation [Haynes et al., 1991]. We do so by com- puting w using NGWD alone (i.e., w ): m m;NGWD DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1519 0.01 −0.03 0.01 −0.01 0.01 0.01 0.1 −0.01 0.3 0.1 −0.03 −0.01 0.03 0.03 0.1 0.1 −0 01 −0.03 0.01 0.01 −0.01 −0.01 −0.1 0.01 −0.03 0 01 −0.01 −0.03 Pressure (hPa) Pressure (hPa) Pressure (hPa) Pressure (hPa) 7 −1 −9 −7 −7 −9 60 E 120 W −1 120 W 120 E 60 W −1 −1 −1 −9 −7 −1 −7 −5 −5 −9 −3 −1 −1 −1 −1 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) DJF SLP 4xCONTROL−CONTROL b) SON SLP 4xCONTROL−CONTROL o o 180 W 0 o o 0 180 W c) DJF SLP 4xGWLOG−GWLOG d) SON SLP 4xGWLOG−GWLOG o o 180 W 0 o o 0 180 W Figure 13. Differences of sea level pressure between 4xCONTROL and CONTROL, and between 4xGWLOG and GWLOG, for (a, c) DJF (in the NH), and (b, d) SON (in the SH). Contours start at 61 hPa, with an interval of 2 hPa. Light red and blue shadings indicate positive and negative statistically signiﬁcant differences, respectively (Student t-test, a50.01). () ð 2 z=H 2z =H 2e e cos / 0 0 w ð/; zÞ5 ; X ð/; z Þ; dz ; (7) NGWD m;NGWD 2 ^ 0 a cos /; d/ z fð/; z Þ / / 1 1 and using the divergence of the EP ﬂux alone (i.e., w ): m;DF () ð 0 2 z=H 2z =H 2e e cos / 0 0 w ð/; zÞ5 ; DFð/; z Þ; dz ; (8) m;DF / 2 ^ 0 a cos /; d/ z fð/; z Þ 1 1 The corresponding plots for w and w are shown in the third and bottom rows, respectively, of m;NGWD m;DF Figure 12. In the NH, the main contribution to the change in vertical motion is due to the changes in NGWD (Figure 12e). The differences in w strongly resemble in both magnitude and evolution those in w , m;NGWD m while no clear pattern is observed for w . m;DF In the SH, the w pattern agrees with the pattern in the forcing (Figures 12b and 12f). Interestingly, m;NGWD the residual circulation induced by the resolved forcing opposes almost exactly (Figure 12h) that induced by the NGWD, resulting in the insigniﬁcant w differences in the SH. We interpret that in the NH the ampli- tude and variability of the resolved waves are sufﬁciently large not to be sensitive to the rather small differ- ences in the annual cycle of the NGWD. In contrast, in the SH the amplitude and variability of the resolved waves are not as large, and they respond compensating the forcing from the parameterized NGWs. 4.4. Impact on a Warmer Climate In this section we analyze the potential impact of NGW with source-depending amplitudes on a warmer cli- mate. Figure 13 displays sea level pressure (SLP) differences 4xCONTROL-CONTROL and 4xGWLOG-GWLOG, in DJF (NH) and SON (SH). The tropospheric circulation response to warmer conditions reinforces the DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1520 −1 −1 −9 −1 −3 −1 −5 −7 −1 −3 −1 −7 −7 −5 −3 −1 −5 −3 −13 −3 −1 −9 −3 −11 −3 −11 −5 −9 120 W 120 W 60 E 60 W 60 E 60 W 120 E 120 E −1 −5 −7 −9 −7 −3 −11 −5 −13 −3 −5 −7 −13 −11 −5 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) East 4xCONTROL−CONTROL 1 hPa b) West 4xCONTROL−CONTROL 1 hPa 0.05 0.05 0.15 60 60 30 30 −0.05 0 0 −30 −0.05 −30 −0.05 0.25 0.05 0.15 0.05 −60 −60 0.45 0.35 0.15 0.25 J F M A M J J A S O N D J F M A M J J A S O N D c) East 4xCONTROL−CONTROL 100 hPa d) West 4xCONTROL−CONTROL 100 hPa 0.05 0.15 0.15 0.25 0.35 0.25 0.45 0.55 0.35 0.65 0.45 0.05 60 60 0.55 −0.05 −0.15 0.15 0.15 0.05 −0.15 0.05 −0.05 30 30 −0.15 0.05 −0.05 −0.05 −0.05 −0.05 0.15 0.05 0 0 0.05 −0.05 −0.05 −0.15 0.05 −0.05 0.15 −0.05 −0.15 −30 −30 0.05 −0.05 −0.15 0.05 0.05 0.15 0.45 0.15 0.55 0.25 −0.15 0.25 0.65 0.35 0.75 −0.05 0.350.35 −60 −60 0.75 0.65 0.25 0.55 0.35 0.45 0.15 0.25 0.15 J F M A M J J A S O N D J F M A M J J A S O N D e) East 4xCONTROL−CONTROL emitted f) West 4xCONTROL−CONTROL emitted 0.3 0.3 0.3 0.3 0.5 0.5 0.5 0.7 0.5 0.7 60 60 0.3 0.3 0.3 0.3 0.1 0.1 0.1 0.1 −0.1 −0.1 −0.1 −0.1 −0.3 −0.3 −0.3 −0.3 −0.5 −0.5 −0.5 30 30 −0.3 −0.3 −0.3 −0.3 −0.3 −0.3 −0.3 −0.5 −0.5 −0.3 −0.5 −0.5 0 −0.5 0 −0.5 −0.5 −0.5 −0.1 −0.3 −0.3 −0.3 −0.3 −0.3 −30 −30 −0.7 −0.3 −0.3 −0.5 −0.5 −0.5 −0.5 −0.5 −0.5 −0.3 −0.3 −0.3 −0.1 −0.3 −0.1 −0.1 −0.1 0.1 0.1 0.1 0.1 0.3 0.3 0.3 0.7 0.5 0.3 0.7 0.5 0.5 0.9 −60 −60 0.5 0.5 0.7 0.5 0.5 0.3 0.3 0.7 0.3 0.3 01 01 01 J F M A M J J A S O N D J F M A M J J A S O N D Figure 14. Differences of nonorographic (left column) eastward and(right column) westward gravity wave stress (in mPa) between 4xCON- TROL and CONTROL at the launching altitude, 100 hPa and 1hPa, as indicated in the ﬁgure titles. Light red and blue shadings indicate posi- tive and negative statistically signiﬁcant differences, respectively (Student t-test, a50.01). subtropical anticyclones and deepens the subpolar lows in both hemispheres, in agreement with projec- tions from the Coupled Model Intercomparison Project Phase 5 (CMIP5) [e.g., Manzini et al., 2014]. The mag- nitude and locus of the SLP differences look insensitive to the use of parameterized NGW hooked to their sources. Figure 14 shows the difference between 4xCONTROL and CONTROL in eastward and westward stress at the launching altitude, 100 and 1 hPa. Interesting features emerge in this ﬁgure. At the launching level (Figures 14e and 14f), there is a poleward shift of the latitude bands with maximum stress in the extratropics of both hemispheres. This is consistent with the intensiﬁcation of the circulation described in Figure 13, and with the projected poleward shift in the storm tracks [Scaife et al., 2012]. It can also be seen that the annual cycle intensiﬁes. The poleward shift is also present at 100 hPa (Figures 14c and 14d), where the extratropical annual cycle is notably enhanced, particularly for the westward stress. At 1 hPa (Figures 14a and 14b), there is a weak reduction in eastward stress, and the enhanced annual cycle in the SH westward stress is collocat- ed with the maximum stress in CONTROL (Figure 1b). Figure 15 shows the corresponding plots for the difference between 4xGWLOG and GWLOG, where we can look into the effect of wind ﬁltering alone. At the launching level (Figures 15e and 15f), there is again a pole- ward shift. However, there is practically no signal of an annual cycle. This implies that the enhanced annual cycle in 4xCONTROL is due to changes in the strength of GW sources, while the poleward shift is due to a DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1521 latitude latitude latitude Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) East 4xGWLOG−GWLOG 1 hPa b) West 4xGWLOG−GWLOG 1 hPa 0.05 0.05 60 60 −0.05 30 30 0 0 −30 −0.05 −30 −0.05 0.05 0.05 −60 −60 0.15 0.25 0.15 J F M A M J J A S O N D J F M A M J J A S O N D c) East 4xGWLOG−GWLOG 100 hPa d) West 4xGWLOG−GWLOG 100 hPa 0.05 0.05 0.15 0.15 60 60 −0.05 0.15 0.05 30 30 0.05 −0.05 0 0 0.05 0.15 −0.05 −0.05 0.05 −30 −30 −0.05 0.05 −0.15 0.05 0.25 0.15 0.15 −0.05 0.35 −60 −60 0.15 0.15 0.05 0.05 005 005 J F M A M J J A S O N D J F M A M J J A S O N D e) East 4xGWLOG−GWLOG emitted f) West 4xGWLOG−GWLOG emitted 0.1 0.1 0.1 0.1 60 60 −0.1 −0.1 −0.1 −0.1 30 30 −0.3 −0.3 −0.3 −0.3 −0.5 0 0 −0.5 −0.3 −0.5 −0.3 −0.3 −0.3 −30 −30 −0.1 −0.1 −0.1 −0.1 0.1 0.1 0.1 0.1 −60 −60 0.1 0.1 0.1 J F M A M J J A S O N D J F M A M J J A S O N D Figure 15. As in Figure 14, but for the difference 4xGWLOG minus GWLOG. shift in the winds and storminess due to warmer conditions. The change in the strength of the annual cycle at 1 hPa, more pronounced for the westward component of the momentum ﬂux (Figure 15b), is mainly a result of changes in the wind ﬁltering, and not of changes in the sources. We next analyze the potential impact of triggering GWs from their sources on the seasonal cycle of the downwelling branches of the Brewer-Dobson circulation in a warmer climate. Figure 16 presents similar plots as Figure 12, but for the difference 4xCONTROL minus 4xGWLOG. The change in the NGW drag induced by linking the wave amplitude to their sources is very similar in warmer and in present climate con- ditions in both structure and magnitude (compare Figures 16a and 16b and Figures 12a and 12b). This simi- larity appears also in the w ; w , and w responses. Interestingly, some statistically signiﬁcant m m;NGWD m;DF changes in w show up in the NH (Figure 16g), but contrarily to what happens in the SH, they have the m;DF same sign as w (Figure 16c). We can then conclude that the self-adjustment of parameterized NGW m;NGWD amplitudes to climatological changes in the sources has a minor effect on the induced middle-atmospheric circulation changes in a warmer climate. We have just discussed the impact on the extra-tropical downwel- ling because we ﬁnd it to be the most sensitive aspect of the midlatitude circulation to respond to the GWs annual cycle. We nevertheless veriﬁed that this conclusion also applies to the zonal winds, and found that the differences between 4xCONTROL and 4XGWLOG in zonal mean zonal winds are almost identical to those betwen CONTROL and GWLOG (not shown but see supporting information). DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1522 latitude latitude latitude 0 0.01 −0.01 0.01 0.03 0.1 −0 03 0.01 0.1 0.1 −0.01 0.01 0.1 −0.3 −003 0.3 0.3 −3 −0.01 −1 −0.3 1 0.3 −0.3 3 0.01 0.03 −0.1 0.1 − −0.01 0.03 0.01 0.03 −3 0.1 0.3 0.1 −1 0 0..01 03 0.3 0.1 0.3 0.3 0.1 0.03 −0.1 0.1 −0.01 0.03 0.01 −0.03 −0.01 0.01 0.01 −0.3 −0.1 −1 0.3 0.1 0.03 0.01 −0.1 0.3 0.1 0.01 0.03 −0.1 0.1 0.03 0.01 0.01 0.03 −0.1 −0.01 0.01 −0.03 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 a) NGWD NH 4xCONTROL−4xGWLOG b) NGWD SH 4xCONTROL−4xGWLOG 0.01 0.01 −1 1 1 −0.03 −0.1 0.1 0.1 0.3 0.1 0.3 0.03 0.1 0.01 1 −0.01 1 −0.3 −0.1 10 −0.1 10 −0.03 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * c) w NH 4xCONTROL−4xGWLOG d) w SH 4xCONTROL−4xGWLOG m m 0.01 0.01 −1 0.01 −0.03 −0.1 0.1 0.1 0.3 0.03 −0.3 0.1 0.03 −0.01 0.01 0.3 0.1 0.03 1 −0.03 1 −0.01 0.01 10 10 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * e) w NH 4xCONTROL−4xGWLOG f) w SH 4xCONTROL−4xGWLOG m,NGWD m,NGWD 0.01 0.01 −0.01 −0.03 −1 0.1 0.1 0.3 −0.3 0.3 1 −0.03 1 0.1 0.1 −0.01 0.01 0.03 −0.1 10 10 −0.03 100 100 J F M A M J J A S O N D J F M A M J J A S O N D * * g) w NH 4xCONTROL−4xGWLOG h) w SH 4xCONTROL−4xGWLOG m,DF m,DF 0.01 0.01 −0.01 0.3 −0.01 0.1 0.1 −0.3 −0.1 −0.03 −0.1 0.1 1 1 −0.03 0.1 −0.1 10 10 0.3 0.01 100 100 J F M A M J J A S O N D J F M A M J J A S O N D Figure 16. As in Figure 12, but for 4xCONTROL and 4xGWLOG runs. In the tropics, the situation is not as clear, and it is more difﬁcult to deliver a clear message. We ﬁnd signiﬁ- cant changes in the amplitude and period of the QBO between 4xCONTROL and 4xGWLOG. In both runs the QBO period decreases drastically and the amplitude of the eastward phase is reduced. Also, in 4xGWLOG the oscillation of the winds is lost below 20 hPa, remaining in westward phase (see supporting information). Nonetheless, different settings and tuning of a given parameterization may have different - and somewhat inconsistent- QBO responses in simulations of a warmer climate [Schirber et al., 2015], so we do not consider that those changes be due to a crucial role of the GW sources. 5. Summary and Concluding Remarks In this work, we have presented the mean climate and variability of the middle atmosphere in the new ver- sion of the LMDz general circulation model. A novel characteristic of LMDz is that it includes a set of gravity wave parameterizations where the emitted stress is linked to the source characteristics, namely ﬂow over topography, convection, and fronts and jet imbalances. In general, LMDz with source-related GWD (i.e., CONTROL) shows good climatology and interannual variability as compared to ERA-Interim. Some well- known biases persist, as the lack of an equatorward tilt with height of the southern stratospheric polar night jet, and too strong summer easterly jets in both hemispheres. The model presents good statistics of sudden DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1523 −0.1 −0.03 −0.03 0.03 0.01 −0.01 −0.03 0.01 0.3 −0.03 0.03 0.01 −0.01 0.01 0.1 −0.03 −0.03 0.03 0.01 −0.1 −0.01 0.3 Pressure (hPa) Pressure (hPa) Pressure (hPa) Pressure (hPa) Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 stratospheric warmings, and internally generates a QBO in the tropical stratosphere with reasonable ampli- tude and mean period, as described in more detail by Lott and Guez [2013]. There are two major features that are reproduced in nonorographic GW parameterizations when the launched stress is tied to the intensity of the sources. The ﬁrst one is a realistic representation of momen- tum ﬂux intermittency; the second one is an annual cycle of the stress due to that in the GW sources. Regarding the reproduction of momentum ﬂux intermittency, de la Camara et al. [2016] have shown that it is a crucial factor in order to simulate the stratospheric ﬁnal warming in the SH with a realistic timing. In the present paper, we investigate the possible impact of the source-induced GW stress annual cycle on the mid- dle atmospheric circulation. For this, we have conducted additional experiments in which the intermittency is prescribed, but the launched GW stress is uncoupled from the sources (i.e., GWLOG). Our results show that including GW sources changes the seasonality of the middle atmospheric GW drag. The seasonality of the GW stress is ﬁltered out quite rapidly with altitude, and a quite reasonable midlati- tude climate can be obtained with a scheme without sources and prescribing the GW intermittency. Regarding the global Brewer-Dobson circulation, the GWD differences between CONTROL and GWLOG lead to changes in the seasonality of the Brewer-Dobson circulation that can be up to 10% in the NH, while in the SH the GWD variations are compensated by the resolved wave forcing. Our warmer climate simulations show that the GWD has a stronger seasonality when linked to the GW sources, but we do not ﬁnd any dra- matic ampliﬁcation of climate change in the troposphere or the stratosphere due to the changes in nonoro- graphic GWD speciﬁcation. This result is consistent with Sigmond and Scinocca [2010], who found that the inﬂuence of the basic state on the circulation response to a warmer climate is much larger than the inﬂu- ence of changes in the orographic GW drag. Our conclusions here are nevertheless based on a limited set of experiments, concerning zonal and time mean diagnostics. The results we ﬁnd regarding the midlatitude variability seem to indicate a stronger sensitivity to the GW annual cycle. Longer runs are needed to address this issue in present and future climate. Acknowledgments References The authors thank the comments from Abalos, M., W. J. Randel, and E. Serrano (2012), Variability in upwelling across the tropical tropopause and correlations with tracers in the two anonymous reviewers. This study lower stratosphere, Atmos. Chem. Phys., 12(23), 11,505–11,517, doi:10.5194/acp-12-11505-2012. was funded by the ANR/JPI-Climate/ Andrews, D. G., J. R. Holton, and C. B. Leovy (1987), Middle Atmosphere Dynamics, 489 pp., Academic Press, San Diego, Calif. Belmont Forum project GOTHAM Anstey, J. A., J. F. Scinocca, and M. Keller (2016), Simulating the QBO in an atmospheric general circulation model: Sensitivity to resolved (ANR-15-JCLI-0004-01), and the and parameterized forcing, J. Atmos. Sci., 73, 1649–1664, doi:10.1175/JAS-D-15-0099.1. European project ARISE2 (Horizon Beres, J. H. (2005), Implementation of a gravity wave source spectrum parameterization dependent on the properties of convection in the 2020, GAN653980). The data used in Whole Atmosphere Community Climate Model (WACCM), J. Geophys. Res., 110, D10108, doi:10.1029/2004JD005504. this study are available upon request Bushell, A. C., N. Butchart, S. H. Derbyshire, D. R. Jackson, G. J. Shutts, S. B. Vosper, and S. Webster (2015), Parameterized gravity wave to [email protected]. A. de la C amara momentum ﬂuxes from sources related to convection and large-scale precipitation processes in a global atmosphere model, J. Atmos. has been supported by the Advanced Sci., 72(11), 4349–4371, doi:10.1175/JAS-D-15-0022.1. Study Program at The National Center Butchart, N., et al. (2011), Multimodel climate and variability of the stratosphere, J. Geophys. Res., 116, D05102, doi:10.1029/2010JD014995. for Atmospheric Research (NCAR), Charlton, A. J., and L. M. Polvani (2007), A new look at stratospheric sudden warmings. Part I: Climatology and modeling benchmarks, which is sponsored by the U.S. J. Clim., 20(3), 449–469, doi:10.1175/JCLI3996.1. National Science Foundation. Charron, M., and E. Manzini (2002), Gravity waves from fronts: Parameterization and middle atmosphere response in a general circulation model, J. Atmos. Sci., 59(5), 923–941, doi:10.1175/1520-0469(2002)059¡0923:GWFFPA>2.0.CO;2. de la C amara, A., and F. Lott (2015), A parameterization of the gravity waves emitted by fronts and jets, Geophys. Res. Lett., 42, 2071–2078, doi:10.1002/GL063298. de la Camara, 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/JD022002. de la Camara, A., F. Lott, V. Jewtoukoff, R. Plougonven, and A. Hertzog (2016), On the gravity wave forcing during the southern stratospher- ic ﬁnal warming in LMDz, J. Atmos. Sci., 73, 3213–3226, doi:10.1175/JAS-D-15-0377.1. Eckermann, S. D. (2011), Explicitly stochastic parameterization of nonorographic gravity wave drag, J. Atmos. Sci., 68, 1749–1765, doi: 10.1175/JAS3684.1. Fritts, D. C. (1989), A review of gravity wave saturation processes, effects, and variability in the middle atmosphere, Pure Appl. Geophys., 130, 343–371, doi:10.1007/BF00874464. Fritts, D. C., and M. J. Alexander (2003), Gravity wave dynamics and effects in the middle atmosphere, Rev. Geophys., 41(1), 1003, doi: 10.1029/2001RG000106. Haynes, P. H., C. J. Marks, M. E. McIntyre, T. G. Shepherd, K. P. Shine (1991), On the ‘‘Downward Control’’ of extratropical diabatic circulations by eddy-induced mean zonal forces, J. Atmos. Sci., 48, 651–978. Hines, C. O. (1997), Doppler-spread parameterization of gravity-wave momentum deposition in the middle atmosphere. Part 2: Broad and quasi monochromatic spectra, and implementation, J. Atmos. Solar Terr. Phys., 59(4), 387–400, doi:10.1016/S1364-6826(96)00080-6. 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. 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. Weather Rev., 127(5), 788–801, doi:10.1175/1520-0493(1999)127¡0788:AOSBIA>2.0.CO;2. DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1524 Journal of Advances in Modeling Earth Systems 10.1002/2016MS000753 Lott, F., and L. Guez (2013), A stochastic parameterization of the gravity waves due to convection and its impact on the equatorial strato- sphere, 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, Clim. Dyn., 25(7-8), 851–868, doi:10.1007/s00382-005-0064-x. Lott, F., R. Plougonven, and J. Vanneste (2010), Gravity waves generated by sheared potential vorticity anomalies, J. Atmos. Sci., 67(1), 157–170, doi:10.1175/2009JAS3134.1. Lott, F., R. Plougonven, and J. Vanneste (2012a), Gravity waves generated by sheared three-dimensional potential vorticity anomalies, J. Atmos. Sci., 69(7), 2134–2151, doi:10.1175/JAS-D-11-0296.1. Lott, F., L. Guez, and P. Maury (2012b), A stochastic parameterization of non-orographic gravity waves: Formalism and impact on the equa- torial stratosphere, Geophys. Res. Lett., 39, L06807, doi:10.1029/GL051001. Manzini, E., N. A. McFarlane, and C. McLandress (1997), Impact of the Doppler spread parameterization on the simulation of the middle atmosphere circulation using the MA/ECHAM4 general circulation model, J. Geophys. Res., 102, 25,751–25,762, doi:10.1029/97JD01096. Manzini, E., et al. (2014), Northern winter climate change: Assessment of uncertainty in CMIP5 projections related to stratosphere- troposphere coupling, J. Geophys. Res. Atmos., 119, 7979–7998, doi:10.1002/2013JD021403. Maury, P., F. Lott, L. Guez, and J.-P. Duvel (2013), Tropical variability and stratospheric equatorial waves in the IPSLCM5 model, Clim. Dyn., 40(9-10), 2331–2344, doi:10.1007/s00382-011-1273-0. McFarlane, N. A. (1987), The effect of orographically excited gravity wave drag on the general circulation of the lower stratosphere and tro- posphere, J. Atmos. Sci., 44, 1775–1800, doi:10.1175/1520-0469(1987)044¡1775:TEOOEG>2.0.CO;2. McLandress, C., T. G. Shepherd, S. Polavarapu, and S. R. Beagley (2012), Is missing orographic gravity wave drag near 60S the cause of the stratospheric zonal wind biases in chemistry-climate models?, J. Atmos. Sci., 69(3), 802–818, doi:10.1175/JAS-D-11-0159.1. 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, Q. J. R. Meteorol. Soc., 112(474), 1001–1039, doi:10.1002/ qj.49711247406. Randel, W. J., R. R. Garcia, and F. Wu (2002), Time-dependent upwelling in the tropical lower stratosphere estimated from the zonal-mean momentum budget, J. Atmos. Sci., 59, 2141–2152, doi:10.1175/1520-0469(2002)059¡2141:TDUITT>2.0.CO;2. Richter, J. H., F. Sassi, and R. R. Garcia (2010), Toward a physically based gravity wave source parameterization in a general circulation mod- el, J. Atmos. Sci., 67(1), 136–156, doi:10.1175/2009JAS3112.1. Rind, D., R. Suozzo, N. K. Balachandran, A. Lacis, and G. Russell (1988), The GISS global climate-middle atmosphere model. Part i: Model structure and climatology, J. Atmos. Sci., 45(3), 329–370, doi:10.1175/1520-0469(1988)045¡0329:TGGCMA>2.0.CO;2. Sassi, F., R. R. Garcia, B. A. Boville, and H. Liu (2002), On temperature inversions and the mesospheric surf zone, J. Geophys. Res., 107(D19), 4380, doi:10.1029/2001JD001525. Scaife, A. A., et al. (2012), Climate change projections and stratosphere-troposphere interaction, Clim. Dyn., 38(9-10), 2089–2097, doi: 10.1007/s00382-011-1080-7. Schirber, S., E. Manzini, and M. J. Alexander (2014), A convection-based gravity wave parameterization in a general circulation model: Implementation and improvements on the QBO, J. Adv. Model. Earth Syst., 6, 264–279, doi:10.1002/2013MS000286. Schirber, S., E. Manzini, T. Krismer, and M. Giorgetta (2015), The quasi-biennial oscillation in a warmer climate: Sensitivity to different gravity wave parameterizations, Clim. Dyn., 45, 825–836, doi:10.1007/s00382-014-2314-2. Sigmond, M., and J. F. Scinocca (2010), The inﬂuence of the basic state on the northern hemisphere circulation response to climate change, J. Clim., 23(6), 1434–1446, doi:10.1175/2009JCLI3167.1. Song, I.-S., and H.-Y. Chun (2005), Momentum ﬂux spectrum of convectively forced internal gravity waves and its application to gravity wave drag parameterization: Part i. theory, J. Atmos. Sci., 62, 107–124. Stephan, C., and M. J. Alexander (2015), Realistic simulations of atmospheric gravity waves over the continental U.S. using precipitation radar data, J. Adv. Model. Earth Syst., 7, 823–835, doi:10.1002/2014MS000396. Warner, C. D., and M. E. McIntyre (1996), On the propagation and dissipation of gravity wave spectra through a realistic middle atmo- sphere, J. Atmos. Sci., 53, 3213–3235. 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. Zulicke, € C., and D. Peters (2008), Parameterization of strong stratospheric inertia-gravity waves forced by poleward-breaking Rossby waves, Mon. Weather Rev., 136(1), 98–119, doi:10.1175/2007MWR2060.1. DE LA CAMARA ET AL. GW SOURCES IMPACTS ON MIDDLE ATMOSPHERE 1525

Journal

Journal of Advances in Modeling Earth Systems – Wiley

Published: Dec 1, 2016

Keywords: ; ; ;

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Climatology of the middle atmosphere in LMDz: Impact of source‐related parameterizations of gravity wave drag

Climatology of the middle atmosphere in LMDz: Impact of source‐related parameterizations of gravity wave drag

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Climatology of the middle atmosphere in LMDz: Impact of source‐related parameterizations of gravity wave drag

Climatology of the middle atmosphere in LMDz: Impact of source‐related parameterizations of gravity wave drag

References (47)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies