+:+
Where possible I have tried to write things for a general audience whilst staying true to terms that may arise in the geosciences.
Listed here are a few definitions for clarity that it may be useful clarify the definitions for:
Fluid | Anything that flows1. This includes water and air. |
A fictitious force that arises due to existing in a rotating reference frame.
N2L: \[\begin{equation} \vec{F} = m\vec{a} = m \left(\frac{\mathrm{d}^2\vec{r}}{\mathrm{d}t^2}\right) \end{equation}\]
In the rotating frame: \[\begin{equation} \vec{r} = \vec{r_r} + \left(\vec{\Omega}\times\vec{r_r}\right) \end{equation}\]
Such that (for constant \(|\vec{\Omega}|\)): \[\begin{equation} \vec{F} = m \left(\frac{\mathrm{d}^2\vec{r_r}}{\mathrm{d}t^2}\right) - 2m \vec{\Omega}\times\frac{\mathrm{d}\vec{r_r}}{\mathrm{d}t} - m \vec{\Omega}\times(\vec{\Omega}\times\vec{r_r}) \end{equation}\]
\[\begin{align} \vec{F} & = m \left(\frac{\mathrm{d}^2\vec{r_r}}{\mathrm{d}t^2}\right) - 2m \vec{\Omega}\times\frac{\mathrm{d}\vec{r_r}}{\mathrm{d}t} - m \vec{\Omega}\times(\vec{\Omega}\times\vec{r_r}) \\ \vec{F} & = m\vec{a_r} - 2m \vec{\Omega}\times\vec{u_r} - m \vec{\Omega}\times(\vec{\Omega}\times\vec{r_r}) \\ \end{align}\]
MIT demonstration provides the easiest way to view this.
Relevance for geophysical fluid dynamics?
The Earth is a rotating reference frame!
\(\vec{\Omega} = 7.292\times 10^{-5}\) rad/s
As fluids move in the atmosphere and ocean they will experience the Coriolis force.
We often think in pressure and velocity. These are tangible variables, but carry a danger of thinking in terms of causality…
Pressure makes fluids move (velocity) but velocity causes pressure changes (Bernoulli and flight)!
Further, velocity conveys information at its speed, whilst pressure is significantly faster (instantaneous or speed of sound).
Vorticity solves this problem for us.
Hunga-Tonga Lamb Wave by Matthew Barlow using NASA GOES images
Vorticity1 is the local rotation or angular momentum of the fluid.
\[\begin{equation} \vec{\omega} = \nabla \times \vec{u} \,\, , \end{equation}\]
\[\begin{equation} \vec{\omega} = \left[\frac{\partial u_z}{\partial u_y}-\frac{\partial u_y}{\partial u_z} \,\, , \,\, \frac{\partial u_x}{\partial u_z}-\frac{\partial u_z}{\partial u_x} \,\, , \,\, \frac{\partial u_y}{\partial u_x}-\frac{\partial u_x}{\partial u_y}\right] \end{equation}\]
\[\begin{equation} \frac{\mathrm{D}\vec{u}}{\mathrm{D}t} = -\frac{1}{\rho}\nabla{p} + \nu\nabla^2{\vec{u}} \,\, , \end{equation}\]
\[\begin{equation} \nabla{\bf\cdot}{\vec{u}} = 0 \,\, . \end{equation}\]
Taking the curl of these equations:
\[\begin{equation} \frac{\mathrm{D}\vec{\omega}}{\mathrm{D}t} = \left( \vec{\omega} {\bf\cdot} \nabla \right) \vec{u} + \nu\nabla^2{\vec{\omega}} \end{equation}\]
\[\begin{equation} \frac{\mathrm{D}\vec{\omega}}{\mathrm{D}t} = \left( \vec{\omega} {\bf\cdot} \nabla \right) \vec{u} + \nu\nabla^2{\vec{\omega}} \,\, . \end{equation}\]
All terms are local to the point under consideration.
Note in-particular, in the absence of viscous forces:
Large-scale meteorology is primarily concerned with the vertical component of vorticity.
Vortex shedding visualisation CC-BY-SA-4.0 by Thierry Dugnolle
Flow visulisation by R. Gontijo and W. Cerqueira (Fair use)
Vortex street over the Juan Fernández Islands from Landsat 7 - NASA (public domain)
vorticity
Simple Harmonic Motion:
\[\begin{equation} \frac{\partial^2 u}{\partial t^2} = - \omega^2 u \end{equation}\]
permits solutions:
\[\begin{equation} u = u_0 e^{i \omega t} \end{equation}\]
i.e. oscillations at frequency \(\omega\).
Wave equation:
\[\begin{equation} \frac{\partial^2 u}{\partial t^2} = c^2 \nabla^2 u \end{equation}\]
permits solutions:
\[\begin{equation} u = u_0 e^{i (\vec{k} \cdot \vec{x} - \omega t)} \end{equation}\]
i.e. oscillations of frequency \(\omega\) travelling in direction \(\vec{k}\).
Pendulum by Wikinana38 used under CC BY-SA 4.0
Mass-spring system by Oleg Alexandrov in public domain
Travelling plane wave by Pajs in public domain
Atmospheric profiles by cmglee used under CC BY-SA 3.0
Gravity is the restoring force.
Consider a small parcel of air, density \(\rho_0\) displaced by \(\Delta z\) 1.
N2L (\(\, f = \rho_0 \, a\)) gives:
\[\begin{equation} -g \Delta \rho = -g[\rho(z)-\rho(z+\Delta z)] = \rho_0 \frac{\partial^2 \Delta z}{\partial t^2} \end{equation}\]
Linearising in a slowly varying atmosphere s.t. \([\rho(z)-\rho(z+\Delta z)] = -\frac{\partial \rho(z)}{\partial z} \Delta z\):
\[\begin{equation} \frac{\partial^2 \Delta z}{\partial t^2} = \frac{g}{\rho_0} \frac{\partial \rho(z)}{\partial z} \Delta z \,\,\,\,\,\,\,\,\,\, , \,\,\,\,\,\,\,\,\,\, \omega^2 = N^2 = - \frac{g}{\rho_0} \frac{\partial \rho(z)}{\partial z} \end{equation}\]
This is an equation for SHM with the buoyancy force (gravity) causing the parcel to oscillate about the point of neutral density with Brunt–Väisälä frequency \(N\).
Skipping over a bunch of maths (see Vallis (2017)) in the atmosphere we get1:
\[\begin{equation} \frac{\partial^2}{\partial t^2}(\nabla^2 + \frac{\partial^2}{\partial z^2}) \vec{u} = -N^2\nabla^2 \vec{u} \end{equation}\]
which is a ‘wave-like’ equation with solutions:
\[\begin{equation} \vec{u} = \vec{u}_0 e^{i (\vec{k} \cdot \vec{x} - \omega t)} \end{equation}\]
where2:
\[\begin{equation} \omega^2 = \frac{\vec{k}^2_\perp N^2}{\vec{k}^2} \end{equation}\]
From the maths we can deduce1:
\(N^2 > 0\) required for stability
(i.e. stratification of \(\mathrm{d}\rho/\mathrm{d}z < 0\))
\(\vec{u} \cdot \vec{c}_p \sim \vec{u} \cdot \vec{k} = 0\)
⇒ Fluid (air, water) motion is parallel to wavefronts
\(|{\vec{u}|} \sim N^{-1}\)
⇒ amplitude grows as waves propogate higher
\[\begin{equation} \frac{\partial^2}{\partial t^2}(\nabla^2 + \frac{\partial^2}{\partial z^2}) \vec{u} = -N^2\nabla^2 \vec{u} \end{equation}\]
\[\begin{equation} \vec{u} = \vec{u}_0 e^{i (\vec{k} \cdot \vec{x} - \omega t)} \end{equation}\]
\[\begin{equation} \omega^2 = \frac{\vec{k}^2_\perp N^2}{\vec{k}^2} \end{equation}\]
From the maths we can deduce1:
\[\begin{equation} \frac{\partial^2}{\partial t^2}(\nabla^2 + \frac{\partial^2}{\partial z^2}) \vec{u} = -N^2\nabla^2 \vec{u} \end{equation}\]
\[\begin{equation} \vec{u} = \vec{u}_0 e^{i (\vec{k} \cdot \vec{x} - \omega t)} \end{equation}\]
\[\begin{equation} \omega^2 = \frac{\vec{k}^2_\perp N^2}{\vec{k}^2} \end{equation}\]
Generation:
Results:
Gravity Waves generated by topography by Bruce Sutherland.
Gravity Waves over Amsterdam Island by NASA (public domain).
Heavily based on Vallis (2017)
See also Davidson (2013)
Wayne Hocking has some excellent notes with nice diagrams illustrating some of the more complex features.
A nearly periodic reversal of the zonal wind in the equatorial stratosphere.
A nearly periodic reversal of the zonal wind in the equatorial stratosphere.
\(z\) (m) is a natural choice for a vertical coordinate.
However, we could use anything that has a 1:1 mapping to z. Pressure is one such value1:
\[\begin{equation} \frac{\mathrm{d}p}{\mathrm{d}z} = -\rho(z) g \end{equation}\]
usually represented in hPa (hectoPascals)2
The advantage of using pressure is that it simplifies some equations, and is directly linked to observations.
Atmospheric pressure profile from North Carolina State University used under fair use
A nearly periodic reversal of the zonal wind in the equatorial stratosphere.
A nearly periodic reversal of the zonal wind in the equatorial stratosphere.
A nearly periodic reversal of the zonal wind in the equatorial stratosphere.
A nearly periodic reversal of the zonal wind in the equatorial stratosphere.
A nearly periodic reversal of the zonal wind in the equatorial stratosphere.
QBO plot by Pierre cb in the public domain.
Recall that when waves exist in a mean flow if u = c_p then vertical group velocity is zero
mean flow acts to slow upwards propogation of GW and they deposit energy.
Model based on upward propogation on GW and the effect on zonal flow. Holton and Lindzden (1968)
Waves generated in upper equatorial troposphere. deep convection etc.
As they propogate upwards eventually cp = u_bar => wave slows giving more time for dissipation to act.
Dissipation deposits energy, increasing mean flow u_bar. This causes the point at which waves slow to be lower in atmosphere, causing dissipation in a cycle etc etc.
This does not explain reversal, however -> 2 wave model
wave that matches mean flow will be dissipated, whilst the other will propogate higher. dissipation low accelerates the flow, causing wind to ‘descend’ other wave dissipating higher will ‘accellerate’ the ‘slower’ flow in the opposite direction Leading to a higher reversal and dissipation.
Heavily based on Vallis (2017)
See also Davidson (2013)
Wayne Hocking has some excellent notes with nice diagrams illustrating some of the more complex features.
Image by NASA (public domain)
Image by NOAA (public domain)
Image by NOAA (public domain)
The Stommel model and Sverdrup flows tell us about the vertically integrated flow.
In reality we also need to understand the vertical variations.
SST from NOAA (public domain)
Animation by BOM Aus. (public domain)
A question of different timescales:
These slides at:
https://jack.atkinson.net/slides/queens_pdra.html
I can be reached at: