Rossby wave

Meanders of the northern hemisphere's jet stream developing (a, b) and finally detaching a "drop" of cold air (c). Orange: warmer masses of air; pink: jet stream; blue: colder masses of air.

Atmospheric Rossby waves result from the conservation of potential vorticity and are influenced by the Coriolis force and pressure gradient. The rotation causes fluids to turn to the right as they move in the northern hemisphere and to the left in the southern hemisphere. For example, a fluid that moves from the equator toward the north pole will deviate toward the east; a fluid moving toward the equator from the north will deviate toward the west. These deviations are caused by the Coriolis force and conservation of potential vorticity which leads to changes of relative vorticity. This is analogous to conservation of angular momentum in mechanics. In planetary atmospheres, including Earth, Rossby waves are due to the variation in the Coriolis effect with latitude. Carl-Gustaf Arvid Rossby first identified such waves in the Earth's atmosphere in 1939 and went on to explain their motion.

One can identify a terrestrial Rossby wave as its phase velocity, marked by its wave crest, always has a westward component. However, the collected set of Rossby waves may appear to move in either direction with what is known as its group velocity. In general, shorter waves have an eastward group velocity and long waves a westward group velocity.

The terms "barotropic" and "baroclinic" are used to distinguish the vertical structure of Rossby waves. Barotropic Rossby waves do not vary in the vertical, and have the fastest propagation speeds. The baroclinic wave modes, on the other hand, do vary in the vertical. They are also slower, with speeds of only a few centimeters per second or less.[3]

Most investigations of Rossby waves have been done on those in Earth's atmosphere. Rossby waves in the Earth's atmosphere are easy to observe as (usually 4-6) large-scale meanders of the jet stream. When these deviations become very pronounced, masses of cold or warm air detach, and become low-strength cyclones and anticyclones, respectively, and are responsible for day-to-day weather patterns at mid-latitudes. The action of Rossby waves partially explains why eastern continental edges in the Northern Hemisphere, such as the Northeast United States and Eastern Canada, are colder than Western Europe at the same latitudes.[4]

Deep convection (heat transfer) to the troposphere is enhanced over very warm sea surfaces in the tropics, such as during El Niño events. This tropical forcing generates atmospheric Rossby waves that have a poleward and eastward migration.

Poleward-propagating Rossby waves explain many of the observed statistical connections between low- and high-latitude climates.[5] One such phenomenon is sudden stratospheric warming. Poleward-propagating Rossby waves are an important and unambiguous part of the variability in the Northern Hemisphere, as expressed in the Pacific North America pattern. Similar mechanisms apply in the Southern Hemisphere and partly explain the strong variability in the Amundsen Sea region of Antarctica.[6] In 2011, a Nature Geoscience study using general circulation models linked Pacific Rossby waves generated by increasing central tropical Pacific temperatures to warming of the Amundsen Sea region, leading to winter and spring continental warming of Ellsworth Land and Marie Byrd Land in West Antarctica via an increase in advection.[7]

Atmospheric Rossby waves, like Kelvin waves, can occur on any rotating planet with an atmosphere. The Y-shaped cloud feature on Venus is attributed to Kelvin and Rossby waves.[8]

Oceanic Rossby waves are large-scale waves within an ocean basin. They have a low amplitude, in the order of centimetres (at the surface) to metres (at the thermocline), compared with atmospheric Rossby waves which are in the order of hundreds of kilometres. They may take months to cross an ocean basin. They gain momentum from wind stress at the ocean surface layer and are thought to communicate climatic changes due to variability in forcing, due to both the wind and buoyancy. Both barotropic and baroclinic waves cause variations of the sea surface height, although the length of the waves made them difficult to detect until the advent of satellite altimetry. Satellite observations have confirmed the existence of oceanic Rossby waves.[9]

Baroclinic waves also generate significant displacements of the oceanic thermocline, often of tens of meters. Satellite observations have revealed the stately progression of Rossby waves across all the ocean basins, particularly at low- and mid-latitudes. These waves can take months or even years to cross a basin like the Pacific.

Rossby waves have been suggested as an important mechanism to account for the heating of the ocean on Europa, a moon of Jupiter.[10]

Rossby wave instabilities are also thought to be found in astrophysical discs, for example, around newly forming stars.[11][12]

It has been proposed that a number of regional weather extremes in the Northern Hemisphere associated with blocked atmospheric circulation patterns may have been caused by quasiresonant amplification of Rossby waves. Examples include the 2013 European floods, the 2012 China floods, the 2010 Russian heat wave, the 2010 Pakistan floods and the 2003 European heat wave. Even taking global warming into account, the 2003 heat wave would have been highly unlikely without such a mechanism.

Normally freely travelling synoptic-scale Rossby waves and quasistationary planetary-scale Rossby waves exist in the mid-latitudes with only weak interactions. The hypothesis, proposed by Vladimir Petoukhov, Stefan Rahmstorf, Stefan Petri, and Hans Joachim Schellnhuber, is that under some circumstances these waves interact to produce the static pattern. For this to happen, they suggest, the zonal (east-west) wave number of both types of wave should be in the range 6–8, the synoptic waves should be arrested within the troposphere (so that energy does not escape to the stratosphere) and mid-latitude waveguides should trap the quasistationary components of the synoptic waves. In this case the planetary-scale waves may respond unusually strongly to orography and thermal sources and sinks because of "quasiresonance".[13]

A 2017 study by Mann, Rahmstorf, et al. connected the phenomenon of manmade Arctic amplification to planetary wave resonance and extreme weather events.[14]

The idea that interannual oceanic Rossby waves are resonantly forced stems from the seminal work of Warren B. White.[15] Recent work proposes that the climate system preferentially responds to solar and orbital forcing with the mediation of long-period Rossby waves.[16] Supported by both observational and theoretical considerations, this new approach complements the Milankovitch theory [17] because changes in the forcing are too small to explain the observed climate variations as simple linear responses.[18]

Propagating cyclonically around the subtropical gyres, the so-called Gyral Rossby waves (GRWs) owe their origin to the gradient ${\displaystyle \beta }$ of the Coriolis parameter relative to the mean radius of the gyres.[19] The resulting modulated western boundary current, whose velocity is added to that of the steady anticyclonic wind-driven current, accelerates/decelerates according to the phase of GRWs. This amplifies the oscillation of the thermocline because of a positive feedback loop ensuing from the temperature gradient between the high and low latitudes of the gyres.

Multi-frequency GRWs overlap, behaving as coupled oscillators with inertia resonantly forced by solar and orbital cycles in subharmonic modes. So, the efficiency of forcing increases considerably as the forcing period approaches a natural period of the GRWs.

The climate response to orbital forcing with the mediation of GRWs allows explaining the glacial-interglacial cycles because of the resulting positive feedback. Endowing the climate response with a resonant feature, this mediation help explain the Mid-Pleistocene Transition (MPT) by involving orbital variations as the only external forcing as well as the little ice ages that occurred in both hemispheres. In the same way as during the MPT, but with periods 10 times longer, a transition occurred at the hinge of Pliocene-Pleistocene.[16] Both transitions as well as the observed adjustment of the South Pacific gyre to the resonance conditions during the MPT are new clues in favor of the mediation of GRWs.

To start with, a zonal mean flow, U, can be considered to be perturbed where U is constant in time and space. Let ${\displaystyle {\vec {u}}=\langle u,v\rangle }$ be the total horizontal wind field, where u and v are the components of the wind in the x- and y- directions, respectively. The total wind field can be written as a mean flow, U, with a small superimposed perturbation, u′ and v′.

${\displaystyle u=U+u'(t,x,y)\!}$

${\displaystyle v=v'(t,x,y)\!}$

The perturbation is assumed to be much smaller than the mean zonal flow.

${\displaystyle U\gg u',v'\!}$

Relative vorticity η, u and v can be written in terms of the stream function ${\displaystyle \psi }$ (assuming non-divergent flow, for which the stream function completely describes the flow):

${\displaystyle {\begin{aligned}u'&={\frac {\partial \psi }{\partial y}}\\[5pt]v'&=-{\frac {\partial \psi }{\partial x}}\\[5pt]\eta &=\nabla \times (u'\mathbf {\hat {\boldsymbol {\imath }}} +v'\mathbf {\hat {\boldsymbol {\jmath }}} )=-\nabla ^{2}\psi \end{aligned}}}$

Considering a parcel of air that has no relative vorticity before perturbation (uniform U has no vorticity) but with planetary vorticity f as a function of the latitude, perturbation will lead to a slight change of latitude, so the perturbed relative vorticity must change in order to conserve potential vorticity. Also the above approximation U >> u' ensures that the perturbation flow does not advect relative vorticity.

${\displaystyle {\frac {d(\eta +f)}{dt}}=0={\frac {\partial \eta }{\partial t}}+U{\frac {\partial \eta }{\partial x}}+\beta v'}$

with ${\displaystyle \beta ={\frac {\partial f}{\partial y}}}$. Plug in the definition of stream function to obtain:

${\displaystyle 0={\frac {\partial \nabla ^{2}\psi }{\partial t}}+U{\frac {\partial \nabla ^{2}\psi }{\partial x}}+\beta {\frac {\partial \psi }{\partial x}}}$

Using the method of undetermined coefficients one can consider a traveling wave solution with zonal and meridional wavenumbers k and ℓ, respectively, and frequency ${\displaystyle \omega }$:

${\displaystyle \psi =\psi _{0}e^{i(kx+\ell y-\omega t)}\!}$

This yields the dispersion relation:

${\displaystyle \omega =Uk-\beta {\frac {k}{k^{2}+\ell ^{2}}}}$

The zonal (x-direction) phase speed and group velocity of the Rossby wave are then given by

${\displaystyle {\begin{aligned}c&\equiv {\frac {\omega }{k}}=U-{\frac {\beta }{k^{2}+\ell ^{2}}},\\[5pt]c_{g}&\equiv {\frac {\partial \omega }{\partial k}}\ =U-{\frac {\beta (\ell ^{2}-k^{2})}{(k^{2}+\ell ^{2})^{2}}},\end{aligned}}}$

where c is the phase speed, c_g is the group speed, U is the mean westerly flow, ${\displaystyle \beta }$ is the Rossby parameter, k is the zonal wavenumber, and ℓ is the meridional wavenumber. It is noted that the zonal phase speed of Rossby waves is always westward (traveling east to west) relative to mean flow U, but the zonal group speed of Rossby waves can be eastward or westward depending on wavenumber.

The Rossby parameter is defined:

${\displaystyle \beta ={\frac {\partial f}{\partial y}}={\frac {1}{a}}{\frac {d}{d\varphi }}(2\omega \sin \varphi )={\frac {2\omega \cos \varphi }{a}}}$

${\displaystyle \varphi }$ is the latitude, ω is the angular speed of the Earth's rotation, and a is the mean radius of the Earth.

If ${\displaystyle \beta =0}$, there will be no Rossby waves; Rossby waves owe their origin to the gradient of the tangential speed of the planetary rotation (planetary vorticity). A "cylinder" planet has no Rossby waves. It also means that at the equator of any rotating, sphere-like planet, including Earth, one will still have Rossby waves, despite the fact that ${\displaystyle f=0}$, because ${\displaystyle \beta >0}$. (Equatorial Rossby wave).