2:15 PM - 2:30 PM
[AAS02-03] Toward New Vertical Wind Science from Global Circulation to Turbulence
Keywords:vertical winds, Brewer-Dobson circulation, stratosphere-troposphere exchange, gravity waves, convection, turbulence
Vertical winds, which are associated with basic atmospheric components such as convection, gravity waves, and general circulation, are an important physical quantity related to latent heat release and transport of momentum, mass and minor constituents. Compared to the horizontal winds, the vertical winds are weaker and dominate on smaller horizontal scales, making it difficult to capture, and hence their studies have been limited. However, with the accumulation of observational data at even limited locations and the introduction of GCMs that can resolve a major spectral range of gravity waves, it has become possible to reconsider the dynamics of vertical winds by a variety of observational, modelling, and theoretical studies.
Atmospheric radars such as the MU radar (36°N, 135°E) and the PANSY radar (69°S, 40°E) are the only instruments that can directly observe vertical winds with fine resolution and good accuracy. The observations show that there are active and quite periods in the vertical wind fluctuations, and that the frequency spectra are much different between the two periods, namely, proportional to -5/3 power of omega for the active period and constant up to the Brunt-Vaisala frequency for the quite one (e.g., Ecklund et al., 1982). The frequency spectra for active periods are explained by the phase modulation of topographically-forced gravity waves (Sato 1990). This characteristic observed frequency spectrum of vertical wind fluctuations has not been well simulated even by a high-resolution numerical model (Tomikawa et al., 2015) in the troposphere, while for the mesosphere, the spectra close to the observations have been obtained by a similar high-resolution model (Shibuya and Sato, 2019). This suggests that vertical wind characteristics largely depend on the altitude region.
The momentum redistribution by gravity waves, which are characterized by strong vertical wind components, is an essential component of the momentum budget in the whole atmosphere, and must be adequately represented in models for accurate climate prediction. However, it is difficult to describe characteristics of non-orographic gravity waves such as amplitudes, intermittency, and wavenumbers which are necessary for parameterizations, because of various and complex wave sources. This insufficient expression of gravity waves are considered to be one of the main causes of the cold bias in the winter stratosphere in climate models (Geller et al., 2013). In order to solve this problem, it is necessary to accumulate observational and numerical studies to obtain guidelines for better parameterizations (e.g., Minamihara et al., 2020). Another possible new method is statistical downscaling using deep learning (Matsuoka et al., 2020).
It is also important to investigate small-scale processes, including tropopause folding and turbulence, to quantify the stratosphere and troposphere exchange. Dynamics of such small-scale processes depends on the vertical flow. Atmospheric radars provide accurate turbulence intensity estimates (Kohma et al. 2019).
Needless to say, global-scale material circulation also involves vertical flow. The zonal mean Lagrangian flow, which is expressed as the sum of mean ageostrophic wind and Stokes drift associated with waves, is not very easy to estimate. However, with the development of data assimilation techniques, new atmospheric reanalyses have become more accurate in estimating ageostrophic winds, allowing to indirectly estimate the contribution of gravity wave forcing and to identify the deficiency in parameterization (Sato & Hirano, 2019). In the future, it is necessary to confirm the realism of the vertical wind estimates at smaller spatio-temporal scales and to clarify the possibilities and limitations of using atmospheric reanalysis data.
In this paper, we have reviewed the major issues in vertical wind science related to turbulence, gravity waves, tropopause, and atmospheric general circulation. All of these issues would be interesting if they could be enhanced to quantitative studies that combine observations, models, and theories including ideal model experiments. In addition, since the target phenomena are small and the research involves handling big data, the use of AI is also effective. Since this is an unexplored research area, there is a great possibility of finding new perspectives and new phenomena. The study of vertical wind studies may open up new avenues for significant improvement of models and future use of satellite Doppler observations.
Atmospheric radars such as the MU radar (36°N, 135°E) and the PANSY radar (69°S, 40°E) are the only instruments that can directly observe vertical winds with fine resolution and good accuracy. The observations show that there are active and quite periods in the vertical wind fluctuations, and that the frequency spectra are much different between the two periods, namely, proportional to -5/3 power of omega for the active period and constant up to the Brunt-Vaisala frequency for the quite one (e.g., Ecklund et al., 1982). The frequency spectra for active periods are explained by the phase modulation of topographically-forced gravity waves (Sato 1990). This characteristic observed frequency spectrum of vertical wind fluctuations has not been well simulated even by a high-resolution numerical model (Tomikawa et al., 2015) in the troposphere, while for the mesosphere, the spectra close to the observations have been obtained by a similar high-resolution model (Shibuya and Sato, 2019). This suggests that vertical wind characteristics largely depend on the altitude region.
The momentum redistribution by gravity waves, which are characterized by strong vertical wind components, is an essential component of the momentum budget in the whole atmosphere, and must be adequately represented in models for accurate climate prediction. However, it is difficult to describe characteristics of non-orographic gravity waves such as amplitudes, intermittency, and wavenumbers which are necessary for parameterizations, because of various and complex wave sources. This insufficient expression of gravity waves are considered to be one of the main causes of the cold bias in the winter stratosphere in climate models (Geller et al., 2013). In order to solve this problem, it is necessary to accumulate observational and numerical studies to obtain guidelines for better parameterizations (e.g., Minamihara et al., 2020). Another possible new method is statistical downscaling using deep learning (Matsuoka et al., 2020).
It is also important to investigate small-scale processes, including tropopause folding and turbulence, to quantify the stratosphere and troposphere exchange. Dynamics of such small-scale processes depends on the vertical flow. Atmospheric radars provide accurate turbulence intensity estimates (Kohma et al. 2019).
Needless to say, global-scale material circulation also involves vertical flow. The zonal mean Lagrangian flow, which is expressed as the sum of mean ageostrophic wind and Stokes drift associated with waves, is not very easy to estimate. However, with the development of data assimilation techniques, new atmospheric reanalyses have become more accurate in estimating ageostrophic winds, allowing to indirectly estimate the contribution of gravity wave forcing and to identify the deficiency in parameterization (Sato & Hirano, 2019). In the future, it is necessary to confirm the realism of the vertical wind estimates at smaller spatio-temporal scales and to clarify the possibilities and limitations of using atmospheric reanalysis data.
In this paper, we have reviewed the major issues in vertical wind science related to turbulence, gravity waves, tropopause, and atmospheric general circulation. All of these issues would be interesting if they could be enhanced to quantitative studies that combine observations, models, and theories including ideal model experiments. In addition, since the target phenomena are small and the research involves handling big data, the use of AI is also effective. Since this is an unexplored research area, there is a great possibility of finding new perspectives and new phenomena. The study of vertical wind studies may open up new avenues for significant improvement of models and future use of satellite Doppler observations.