A set of banded clouds can be seen in MODIS imagery during several days in January and February over the Gulf of Maine. They tend to occur with very cold winds from the NW following storms. MODIS 06FEB2015

There are several similar examples: 17 January and 6, 20 and 24 February

What causes these features? Why are they of that size?

  • 2
    $\begingroup$ Cloud structures like that are generally caused by pressure waves propagating through the atmosphere. What causes the pressure waves is dependent on local topography. $\endgroup$ Commented Feb 25, 2015 at 21:39
  • $\begingroup$ Amazing...if we were looking at an alien planet, would we think straight lines = artificial? I am not saying I don't believe you. Sure looks like an incredible phenomenon us non-scientists would have trouble understanding. Still trying to wade through the scientific stuff so that it makes sense... $\endgroup$
    – stormy
    Commented Feb 26, 2015 at 21:56

3 Answers 3


These are most certainly boundary layer rolls and not gravity waves. While there exists a visual similarity between the two phenomena, and both may exist in similar atmospheric conditions, they can be distinguished by two key characteristics:

  • Unlike gravity waves, which develop in the downwave direction (perpendicular to the crest), boundary layer rolls develop downwind (parallel to the apparent "crest");

  • Unlike gravity waves, whose crest propagates with intrinsic phase speed (related to interfacial density difference and gravitational acceleration) and whose energy propagates with intrinsic group velocity, the large-scale structure of boundary layer rolls is quasi-stationary.

Despite the visual resemblence between gravity waves and boundary layer rolls and possible interactions reported in the literature, it is my opinion that stating that boundary layer rolls are a form of gravity waves is incorrect and misleading.

It is clear based on the satellite picture that the wind is northwesterly because of the distance from land that the convection occurs over the ocean, which is approximately 50 km from the coast uniformly in the SW-NE direction. As the cold and dry air moves over (relatively) warm water, air rises and saturates with water vapor, forming clouds downwind from the coast. For these to be gravity waves, wind would have to be perpendicular to the wave crest, i.e. in the SW-NE direction, which is not the case.

Now that we agree that the wind is northwesterly, we see that the convective lines, the so-called "cloud streets", are aligned with the wind. They are long rolls of counter-rotating air that stretch the depth of the atmospheric boundary layer in height (~1 km) and are of width between 1 and 10 km. Water vapor condensation and cloud formation happens on the upward side of the roll, and evaporation happens on the downward side of the roll. Notice that in this particular example, they start off very thin, and become wider downwind. While these rolls are still not very well understood, the main mechanism for formation is likely to be a combination of rotation and friction (Ekman spirals) in presence of thermally unstable bottom boundary layer, and stable boundary layer cap, thus constraining the circulation into organized cells.

A review paper on boundary layer rolls by Etling and Brown (1993) can be found here.

  • $\begingroup$ I am pretty sure we are talking about the same phenomena, see J.G. Sang (1990) as cited by Etling and Brown (1993). That is why I mentioned that the interface has to be sharp, I am talking about interfacial waves gravity waves. If the interface is diffuse then internal waves can propagate in three dimensions and, as you say, the dispersion relation is $\omega = N \cos \theta$ and $\mathbf{K}\cdot\mathbf{u} = 0$ so that particle motion is perpendicular to the wavenumber vector.This is not the case here where for a sharp interface vertical motions are suppressed.I agree that the wind is from NW. $\endgroup$ Commented Feb 26, 2015 at 8:03
  • $\begingroup$ Take a look at the conclusions in J.G. Sang, particularly point (3). " ... These vortices or rolls, sometimes visualized as the cloud streets, compose the convective bands. Thus, in fact, the convective rolls are part of the convective waves... " $\endgroup$ Commented Feb 26, 2015 at 8:32
  • 1
    $\begingroup$ Some more literature on HCR: journals.ametsoc.org/doi/abs/10.1175/… journals.ametsoc.org/doi/abs/10.1175/… journals.ametsoc.org/doi/abs/10.1175/… $\endgroup$
    – casey
    Commented Feb 26, 2015 at 14:07
  • 1
    $\begingroup$ @IsopycnalOscillation I see your point; however I think this is where the nomenclature becomes weird/ambiguous. It is my understanding that the term "cloud streets" was originally coined by glider pilots; but later associated with both BL rolls and gravity waves in the atmospheric community. The cloud streets refer to a particular cloud formation, and are defined purely by appearance. The paper on convective gravity waves that you linked discusses waves spanning the whole height of the troposphere. On the other hand, BL rolls are constrained within the boundary layer (1-2 km). $\endgroup$ Commented Feb 26, 2015 at 22:42
  • 1
    $\begingroup$ Finally, to answer your comment, I do not think that the presence of cloud streets necessarily implies the presence of gravity waves or BL rolls. $\endgroup$ Commented Feb 26, 2015 at 22:44

These features are internal gravity waves (IGW) and are characterized by alternating areas of upwards and downwards movement of air parcels. They propagate on a sharp density interface between a relatively dry layer ambient air above and a layer of sea water laden air below. When the moist air rises the cloud is formed above the condensation level and eventually evaporates at the trailing edge of the wave. In other words, the cloud regions in the photograph represent the wave crest and the "empty space" in between is the wave trough.

An important requirement for IGWs is a stably stratified atmosphere, and specifically for these waves shown in the photograph, a relatively sharp interface. If the interface thickness is larger the continuous variation of density with height adds the possibility of vertical propagation.

IGWs can be generated by a variety of phenomena, ranging from orographic to non-orographic mechanisms. The former is mentioned in the comments and is relatively intuitive and easy to understand. In the latter case it is not so clear, there are several proposed mechanisms that may contribute to the generation of IGWs, for example, Bühler et al (1999) considered the generation of internal waves by an localized unstable shear layer at the top of the jet stream.

In general, three convective generation mechanisms are typically considered (Song et al, 2003):

  1. Thermal Forcing:

convective clouds are regarded as thermal forcings that generate gravity waves (nonstationary or stationary waves relative to the thermal forcing) in a stably stratified environment.

  1. Obstacle:

convective clouds are considered obstacles to the flow, and gravity waves are generated when the flow is subsequently blocked.

  1. Mechanical oscillator:

strong convective updrafts in convective storms stimulate the stable stratosphere and generate gravity waves above clouds. Convective cells acting as mechanical oscillators, as well as a nonsteady diabatic forcing, can generate gravity waves even under zero background wind relative to the convective cells

Oliver Bühler, Michael E. McIntyre, and John F. Scinocca, 1999: On Shear-Generated Gravity Waves that Reach the Mesosphere. Part I: Wave Generation. J. Atmos. Sci., 56, 3749–3763.

In-Sun Song, Hye-Yeong Chun, and Todd P. Lane, 2003: Generation Mechanisms of Convectively Forced Internal Gravity Waves and Their Propagation to the Stratosphere. J. Atmos. Sci., 60, 1960–1980.


Simple harmonic (SH) long crested cloud waves, commonly seen in windswept skies, occur between adjacent air strata flowing in different directions. One stratum is warmer, with higher humidity, causing cloud formation along the inter-stratum boundary. The simple harmonic wave formations conform to the same physics as the Reynolds U-tube experiment taught in fundamental fluid dynamics, where tube tilting to a critical level, causes SH waves to form along the compliant water/carbon bi-sulfide boundary. In the cloud waves and in the U-tube waves we see compliant boundary manifestations of the boundary layer oscillations of transition that Schubauer and Skramstad demonstrated in wind tunnel air flow along a flat plate in 1941.

Similar (invisible) shear waves develop when the sky is clear – when there is no condensation and cloud formation. These are recognized by pilots who, while flying through this variety of clear air turbulence (CAT), get a SH shuddering feeling “as if flying over a washboard”. These SH inter-stratum waves are occasionally revealed by SH waves developing in the lingering contrails of jet aircraft (Simple Harmonics, Aylmer Express, 2015, figures 16, 17, p. 14).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.