General circulation model results on migrating and nonmigrating tides in the mesosphere and lower thermosphere. Part I: comparison with observations

https://doi.org/10.1016/S1364-6826(02)00045-7Get rights and content

Abstract

The general circulation model of the Department of Numerical Mathematics of the Russian Academy of Science (Volodin and Schmitz, 2001, Tellus 53A (2001) 300) from the surface to mesospheric and lower thermospheric heights has been used to analyse the diurnal and semi-diurnal tides. The GCM includes tropospheric and stratospheric tidal forcings due to absorption of the radiation and latent heat release and uses the gravity wave breaking parameterization of Hines (J. Atmos. Sol. Terr. Phys. 59 (1997a) 371; J. Atmos. Sol. Terr. Phys. 59 (1997b) 387).

The model tides describe the observed tidal amplitudes and phases of eastward wind components at different northern hemispheric medium frequency radar sites (Andenes, Juliusruh, Saskatoon, Yamagawa and Hawaii) for January and July conditions. The separation of model tides into migrating and nonmigrating components shows that the nonmigrating part forms the total tide to a large extent, especially for the diurnal tide at low latitudes. The variability of diurnal and semi-diurnal tides is mostly determined by the variability of the nonmigrating part; the variability due to migrating tidal oscillations contributes only a small amount to the total variability. The nonmigrating diurnal model tide is strongly dependent on the longitude, with maxima in the western hemisphere at middle southern latitudes in January. In July, these tidal amplitudes are much weaker with maxima in the subtropics of the eastern hemisphere.

Introduction

Tides in the middle atmosphere have been studied comprehensively over the last three decades using model simulations and observations on the basis of different experimental methods including rocket soundings, radar and lidar measurements and space-based observations. Especially medium frequency radar (MFR) measurements have been used to study tides and their variability, e.g. by Manson and Meek (1985), Manson et al. (1999) Fritts and Isler (1994), Singer 1992, Singer 1997, and Jacobi et al. (1999). The measurements at several observational stations have been discussed in connection with migrating tides, which are waves travelling sun-synchronously westwards, and forced in the troposphere as well as in the stratosphere. These tides are able to propagate vertically up to the mesosphere and lower thermosphere (MLT).

The amplitude and phase dependences of the migrating diurnal and semi-diurnal components can be understood to a large extent on the basis of linear diurnal and semi-diurnal tidal wave propagation models as in Chapman and Lindzen (1970), Forbes (1982), Vial (1986), Hagan et al. (1995), and Wood and Andrews (1997). Linear propagation models take into account the zonal-mean zonal background wind and temperature as a function of latitude and height. The prescribed profiles of heating in the tropo- and stratosphere force the tidal oscillations. The damping of the tides in the MLT is due to eddy-diffusion processes.

It is known that the parameterization of gravity wave momentum deposition in linear models leads to difficulties in the determination of the seasonal variation of the migrating diurnal tide (Hagan et al., 1995). Furthermore, it is a common weakness of linear models that they are not able to describe the tidal variability resulting from internal interactions of different nonlinear wave processes appropriately. Fritts and Isler (1994) considered tidal variability on the basis of radar observations. Walterscheid and Vincent (1996) showed that the nonlinear interactions between the travelling two-day wave and the diurnal tide is an important process in the variability of the latter. McLandress (1997) discussed the importance of nonlinear effects for the simulation of the semi-annual amplitude variations of the diurnal tide. Besides those nonlinear processes as sources of the tidal variability, the forcing of the nonmigrating tide, due to tropospheric latent heat release for instance as discussed in Hamilton (1981) and Forbes et al. (1997), seems to be important as a source of the long-term variability, especially in the case of the diurnal tide. These nonmigrating components are not taken into account in classical 2d models at all, so the comparison between these model results and observations should only be understood as a first approach (Hagan et al., 1997).

General circulation models (GCMs) describe the full nonlinear interactions between mean fields, planetary waves and tides as well as their coupling with vertically propagating gravity waves. These small-scale waves cannot be resolved explicitly by the model, and so results in the MLT depend on the parameterization of gravity wave momemtum deposition. GCMs extending from the troposphere and stratosphere up to the MLT (e.g. Hamilton, 1995a; Hamilton et al., 1995b; Norton and Thurburn, 1997; McLandress, 1997; Miyahara and Miyoshi, 1997; Volodin and Schmitz, 2001, hereafter VS) offer the possibility to study the tidal oscillations and their dependence on nonlinear processes and nonmigrating components of tidal forcings. Miyahara et al. (1999) considered the contribution of the nonmigrating diurnal and semi-diurnal tidal components to the total tidal fields in the MLT on the basis of the middle atmosphere circulation model of the Kyushu University. Their results yield large diurnal tidal amplitudes at high latitudes where the tidal waves with a wavenumber larger than one are able to propagate vertically in a strong mean background wind field.

In this paper the tropo-strato-mesospheric GCM of the Department of Numerical Mathematics of the Russian Academy of Science (VS) is used, describing the atmosphere from tropospheric to MLT heights (upper boundary at ∼105km height) to study both diurnal and semi-diurnal tides. The GCM takes into account all the physical parameterizations which are usually included in tropo-stratosphere general circulation models (Alexeev et al., 1998) and therefore it should be able to describe the migrating and nonmigrating forcings in an appropriate manner. The main aim of this paper is a detailed comparison of total model tides with radar measurements in the MLT region at different geographical locations. On the basis of the model results, the migrating and nonmigrating tidal components are shown separately at different grid points, which should give a first indication of the contribution of nonmigrating tidal components to the total tide at single observational sites. A comparison of model tidal results on nonmigrating components with observational results can be performed only qualitatively because there are, on the one hand, some modelling uncertainties, while the determination of tides on the basis of measurements is, on the other hand, affected by observational biases. The model must describe the forcings of both migrating as well nonmigrating tides. They arise due to very different physical processes. Only the forcing of migrating tides are simulated appropriately in present state-of-the-art GCMs. Further errors result from uncertainties in the simulation of the mean fields of zonal wind and temperature up to the MLT which strongly influence the vertical tidal propagation. Also inaccurate representation of planetary waves on different time scales can lead to problems because these motions interact with the thermal forced tides. Finally, momentum deposition by dissipating gravity waves must also be included in a GCM, but here none of the parameterization schemes available for this appears fully satisfactory at this stage.

This paper has been organized as follows. In the next section, the data basis and the main GCM characteristics are given. Section 3 presents comparisons between model results and observations, especially with respect to the variability of both diurnal and semi-diurnal tides. The separation into migrating and nonmigrating tidal components is given in Section 4. A summarizing discussion follows in Section 5.

Section snippets

Data

In this paper, only tidal data derived from MFR measurements at northern hemispheric sites have been used for comparison with model results. The different observation sites and their data characteristics are given in Table 1. For the comparison between the observations and the daily model output, the following vectorized averaging procedure has been applied to determine the mean fields and standard deviations.

With the amplitude Ai and phase ϕi for distinct analyses i, (1⩽iN) the tidal

Diurnal and semi-diurnal tides in model and observations

Let us now compare the tides as derived from MFR measurements with the diurnal and semi-diurnal model tides. Results primarily from 1998 and 1999 are shown for the observed tides, with the exception of the two sites at HA and JU (abbreviations are given in Table 2) where data are available for nearly a decade.

Fig. 3, Fig. 4 show diurnal tidal amplitudes and phases of the eastward wind component for January and July, respectively. In January and high latitudes (AN) the model describes the

Migrating and nonmigrating tides

The total tidal variation at a distinct observational site consists of sun-synchronous travelling wave components and others which are not sun-synchronous. The procedure for separating the tides into migrating and nonmigrating parts was explained in Section 2. This separation was done only for the global set of model data because of the gaps in the global coverage of MFR sites. Assuming the model describes the tides at the observational sites in an appropriate manner, as discussed in Section 3,

Discussion and conclusion

Using the tropo-strato-mesosphere GCM of VS from the surface to about 105km altitude, the behaviour of diurnal and semi-diurnal tidal oscillations has been analysed and compared to MFR measurements at different northern hemispheric sites for January and July conditions. In the model, the tidal oscillations are generated as an internal process due to short- and longwave radiation absorption, and large-scale condensation and convective heating in the troposphere and stratosphere. The Hines 1997a,

Acknowledgements

Evgeny Volodin would like to acknowledge grants 99-65533 and 99-05-65442 of the Russian Fund of Basic Research. Alan Manson acknowledges grants from NSERC Canada and thanks the University of Saskatchewan and ISAS for support. David Fritts received support for this research under grant ATM 9813774 from the National Science Foundation. Werner Singer was partially supported by grant SI 501/3-1 of the Deutsche Forschungsgemeinschaft for the Andenes MF radar. We thank U. Achatz for reading the first

References (38)

  • Alexeev, V.A., Volodin, E.M., Galin, V.Ya., Dymnikov, V.P., Lykossov, V.N., 1998. Modelling of the present-day climate...
  • A. Arakawa et al.

    A potential enstrophy and energy conserving scheme for shallow water equations

    Monthly Weather Review

    (1981)
  • B.P. Briegleb

    Delta-Eddington approximation for solar radiation in the NCAR community climate model

    Journal of Geophysical Research

    (1992)
  • S. Chapman et al.

    Atmospheric Tides

    (1970)
  • M.D. Chou et al.

    Calculation of transmission functions in the infrared CO2 and O3 bands

    Journal of Geophysical Research

    (1991)
  • CIRA-86, 1988. In: Rees, D., Barnett, J.J., Labitzke, K. (Eds.), COSPAR International Reference Atmosphere: 1986 Part...
  • J.M. Forbes

    Atmospheric tides. Part 1. Model description and results for the solar diurnal component

    Journal of Geophysical Research

    (1982)
  • J.M. Forbes et al.

    Upper atmosphere tidal oscillations due to latent heat release in the tropical troposphere

    Annals of Geophysics

    (1997)
  • D.C. Fritts et al.

    Mean motions and tidal and two-day structure and variability in the mesosphere and lower thermosphere over Hawaii

    Journal of Atmospheric Science

    (1994)
  • Cited by (23)

    • Climatology of the main (24-h and 12-h) tides observed by meteor radars at Svalbard and Tromsø: Comparison with the models CMAM-DAS and WACCM-X

      2020, Journal of Atmospheric and Solar-Terrestrial Physics
      Citation Excerpt :

      The satellite observations from the HRDI and WINDII on board the UARS satellite, as well as the SABER and TIDI on board the TIMED satellite, have provided a global scale picture of the diurnal and semidiurnal tides seen in the neutral wind and temperature fields of the MLT region (Burrage et al., 1995; McLandress et al., 1994, 1996; Forbes and Wu, 2006; Oberheide et al., 2007; Wu et al., 2008a, 2008b; Mukhtarov et al., 2009; Pancheva et al., 2009, 2010). Various models, such as the Canadian Middle Atmosphere Model (CMAM), Whole Atmosphere Model (WAM), Whole Atmosphere Community Climate Model (WACCM), and different General Circulation Models (GCM) have been able to reproduce to some extent the global distribution and seasonal changes in the tidal fields as observed by satellites (McLandress, 2002a, 2002b; Miyahara and Forbes, 1991; Grieger et al., 2002; Akmaev et al., 2008; Chang et al., 2008). The Global Scale Wave Model (GSWM; Hagan et al., 1999; Hagan and Forbes, 2002, 2003) has been used mainly for comparison with ground-based radar measurements (Mitchell et al., 2002), however in this model the nonlinear tidal dynamics and interaction with planetary and gravity waves are neglected.

    • Diurnal tides at low latitudes: Radar, satellite, and model results

      2014, Journal of Atmospheric and Solar-Terrestrial Physics
      Citation Excerpt :

      Here we are not concentrating on the tidal variability issue, but we tried to explain how the short term variability of tides is present on the different scales of composite analyses. To demonstrate the differences between the various composite analyses, we derived the monthly mean tidal parameters for all composite analyses (4-d/1-d, 10-d/5-d, 60-d/1-d) using vector average (Grieger et al., 2002). As a typical example, the monthly mean diurnal tidal amplitudes for the above mentioned composite analyses over Cachoeira Paulista in 2005 are shown in Fig. 1.

    • Nocturnal temperature changes over tropics during CAWSES-III campaign: Comparison with numerical models and satellite data

      2010, Journal of Atmospheric and Solar-Terrestrial Physics
      Citation Excerpt :

      Also some small quantitative inaccuracies would easily lead to a substantial shift of a few degrees in the latitudinal structure of the migrating and nonmigrating tides in the nodal regions. It is now recognized, through observational analyses (e.g., Forbes, 1995; Manson et al., 2004) and modeling studies (e.g., Forbes et al., 2001; Grieger et al., 2002; Hagan and Forbes, 2003; Ward et al., 1999), that there is a whole spectrum of tides propagating from the troposphere to the MLT, that are related to longitudinal variations in the heating rates due to variations in absorbing gas concentrations, land–sea differences and their influence on latent heat release, topography and other factors. Nonlinear tide–tide interactions (Smith and Ortland, 2001) and tide–planetary wave interactions (Grieger et al., 2004) in the stratosphere and mesosphere are also known to play a key role.

    • A climatology of nonmigrating semidiurnal tides from TIMED Doppler Interferometer (TIDI) wind data

      2007, Journal of Atmospheric and Solar-Terrestrial Physics
      Citation Excerpt :

      For an overview of recent diurnal analyses, see Oberheide et al. (2006) and references therein. Tidal and general circulation models have been used to study semidiurnal nonmigrating tidal sources and morphology (Ward et al., 2005; Mayr et al., 2005; Hagan and Forbes, 2003; Grieger et al., 2002; Yamashita et al., 2002). These simulations gave an idea of the latent heat and planetary wave–tidal interaction forcing mechanisms and the spatio-temporal amplitude and phase distributions.

    • A numerical study of the impact of nonlinearity on the amplitude of the migrating diurnal tide

      2007, Journal of Atmospheric and Solar-Terrestrial Physics
      Citation Excerpt :

      In fact, many theoretical and observational works have revealed these interactions are ubiquitous in the real middle and upper atmosphere, e.g. tides and planetary waves by Hagan and Roble (2001) and Mayr et al. (2003) and tides and gravity waves by Miyahara and Forbes (1991) and Lu and Fritts (1993). These interactions may also lead to the disagreement of the observed diurnal tidal amplitudes with the GSWMs at altitudes above 90 km (Burrage et al., 1995; Palo et al., 1997; Bruinsma et al., 2002; Grieger et al., 2002; Pancheva et al., 2002; She et al., 2003; Batista et al., 2004; Zhang et al., 2004). 5.

    View all citing articles on Scopus
    View full text