Skip Navigation

Weatherwise -- May-June 2015

ResizeResize Text: Original Large XLarge

Weather Queries

What is the source of sound when a strong wind is blowing?

Barbara Schlatter

St. Louis, Missouri

This question has been on the back burner for several years because it is difficult to answer fully. A quick introduction to sound waves is helpful. Sound travels through air at a speed dependent only upon the temperature. At 59°F, sound travels at about 760 mph, faster than a commercial jet plane. Sound propagates through air as a compression wave, which is to say that molecules in the path of the wave move back and forth in the direction the wave is traveling. Molecules crowd more closely together at the peak of the wave (compression) and become more widely spaced at the bottom of the wave (rarefaction), as in Figure 1. The sound wave passes through the air molecules, and molecular collisions propel the wave front forward like a ripple.

The frequency of arrival of sound waves determines the pitch of the sound. The amplitude of a sound wave, as measured by the size of the pressure variations (very small) associated with its passage, determines how loud the sound is. The human ear is sensitive to frequencies from about 20–20,000 Hertz, but is most sensitive in the range 2,000–5,000 Hertz.

Everyone is familiar with musical notes, which have a distinct frequency (pitch), but the sounds generated by the wind are usually a combination of many frequencies, most of them low in the audio spectrum. Just as the mixture of wavelengths in visible light generates white light, so the mixture of many audio frequencies (equivalently, wavelengths) generates noise.

Figure 1.  Propagation of sound at a given pitch, represented by a sinusoidal pressure wave (top). Compression of molecules occurs at each wave crest; rarefaction occurs at each wave trough (bottom).

Wind makes familiar sounds. It rustles leaves in the trees, crackles dry autumn leaves swirling down the pavement, groans or roars through barren limbs in winter, sings around telephone wires, or rattles and creaks the house. Where does the noise come from? It comes partly from the objects set in motion by the wind. When objects deform or break in the wind, they generate noise internally. If objects vibrate, the vibration sets the adjacent air in motion at the same frequency, generating sound waves. A tuning fork does the same thing but without wind. Many objects deforming or vibrating in random ways create noise. Most often, we hear lower frequencies such as a jet plane or thunder would generate. The reason is that the atmosphere dampens higher frequencies more efficiently than lower frequencies. Thus, lower-frequency sound travels farther. How often do you hear the thumping bass from a passing car, even with the windows closed, but not the higher frequencies?

Wind noise is only partly the result of moving objects. If you've ever been downtown in a city with tall buildings lining the street on a windy day, you hear the wind, but often you discern nothing moving. The buildings, sidewalks, and streets aren't generating sound, but rather the air itself. Obstacles in the path of the wind create turbulent flow, making the moment-to-moment wind gusty and unpredictable. Small-scale swirls or eddies in the wind, called vortices, range in size from centimeters to many meters across, and these are capable of generating sound. How these small, rotating, and seemingly random entities generate organized compression waves that reach our ears as sound is the subject of aeroacoustics, a branch of physics that flowered in the 1950s. I am not qualified to explain this complex process, but the conversion of kinetic energy (the energy of motion in the air) to acoustic energy is at the heart of the matter.

Figure 2.  Mountaintop winds blowing over the crags can generate vortices that, in turn, produce low-frequency noise.

Here is another and perhaps better example of obstacles in the wind generating turbulence, which in turn produces noise. Hike above the timberline in the Rocky Mountains on a very windy day. The roar of the wind can stifle conversation and, at times, be almost deafening, but the rocks are not moving or vibrating. The sound comes from the air. As the wind funnels through passes, is deflected by cliffs, or swirls around rocky promontories, turbulent eddies of different sizes break off from the main flow and interact with each other (Figure 2), creating sound, mostly at low frequencies. In fact, sometimes these frequencies are below the audible range: infrasound. Infrasound can travel many miles. Though it cannot be heard, it has been detected with special instruments east of the mountains on very windy days.

What about sound generation in the free air, far from obstacles that might generate turbulence? It happens because the atmosphere generates turbulence naturally. A common source region is a strong inversion—a layer in which the temperature increases rapidly with height, perhaps 20°F in 700 feet). Inversions inhibit vertical mixing. It is common for strong winds to blow just above an inversion while the air is simultaneously calm in the colder air below.

Where I live, in Boulder, Colorado, it is common in winter to have a shallow cold air mass with light northeast winds pushed up against the east slope of the Front Range Rockies, while, at the same time, a strong mountain wave, aided by west winds crossing the Continental Divide, rides over the top of the inversion. At the inversion, this results in considerable shear: light winds in the cold air below and strong west winds in the warmer air above. Strong shear leads to turbulence. Sometimes, as the west wind tries to erode away the cold surface air, vortices in the turbulent layer generate enough noise to be heard at the ground: a dull roar is heard even though the surface air is nearly calm.

Severe thunderstorms generate extreme turbulence. In my younger days, I heard a roar aloft from a passing funnel cloud, but there was no wind at the ground. Tornadoes are associated with violent turbulence. They generate very loud noises and infrasound that can be detected hundreds of miles away. In fact, experimental infrasound networks have been used to detect tornadoes.

To summarize, it is quite likely that small vortices in turbulent flow are mainly responsible for sound waves generated in the air. If the wind encounters obstacles on the ground, noise from moving objects is added to that generated in the air.

I thank Al Bedard of NOAA's Earth System Research Laboratory for help in answering this question.

This question pertains to cloud cover near the Massachusetts coast on September 18, 2010, a day that thwarted local forecasters, who predicted sunshine. Can an inversion combined with solar heating cause a generalized cloud cover? If so, why do convective clouds in an inversion fuse to become a layer of overcast? I have always thought that convection represents discrete parcels of rising air, and the most that can come from this is a partly cloudy sky. See my photo of cloud cover, taken on the afternoon of September 17. These clouds were less extensive the following day, but still ruined the day for inland sun-lovers. On September 18, cloud growth seemed to increase inland during the morning and early afternoon, but the coast, with onshore flow, was largely in the clear. It was a glorious day there.

Tom Adams

Marblehead, Massachusetts

I have seen this scenario play out many times in the Midwest: A cold front passes. The next morning dawns clear, but there's a breeze from the west or northwest. By mid- to late morning, a few cumulus clouds appear, moving with the brisk flow aloft. These clouds grow deeper and begin to merge. Eventually by midday, the sky is overcast with stratocumulus clouds. The clouds dissipate slowly after sundown. The scenario often repeats for one or two more days. The conditions you describe are similar, except that cloud cover was much less extensive over water than land.

The weather maps for September 17, 2010, indicate that a surface cold front passed through eastern Massachusetts early in the day. A stable layer of air, in which the temperature changed little with altitude, from about 5,000 to 6,500 feet above the surface, marked the top of the cool air mass. Farther inland, an inversion (a layer with increasing temperature with altitude) marked the top of the cool air mass. By midday on September 18, the inversion became established over eastern Massachusetts. The air below the inversion was fairly moist. The air above the inversion was warmer and very dry.

So how does a stable layer, combined with solar heating, generate stratocumulus cloud cover? It happens in stages.

Stage 1: Within a fresh, cool air mass, the ground is usually warmer than the overlying air, and there's often a breeze, 10–25 mph. Heating of the air mass by the ground, especially when the sun heats the ground, and mechanical mixing of the air by the wind both foster the development of a steep lapse rate within the cold air, that is, a decrease in temperature with height of about 5.4°F per 1,000 feet of altitude. This lapse rate is called the dry adiabatic lapse rate, and, where it exists, up and down air motions are facilitated. The air becomes turbulent and well mixed.

Stage 2: The water vapor mixing ratio is the ratio of the mass of water vapor to the mass of dry air within the same volume. In a well-mixed layer, this mixing ratio is constant, as would be expected. The maximum possible mixing ratio decreases with decreasing temperature. Any volume of air subject to lifting in the mixed layer cools at a rate of 5.4°F per thousand feet of lift. If a volume of air is lifted sufficiently high that it cools to its dewpoint temperature, some of the water vapor will condense, and a cloud forms. When this happens, small, puffy cumulus clouds appear near the top of each little updraft.

Stage 3: If the cumulus cloud forms within the mixed layer, it can grow taller because condensation within the cloud raises its temperature, enough to make the cloud buoyant with respect to its surroundings. A saturated updraft (humidity 100%) cools less rapidly than 5.4°F per thousand feet of lift because of the energy released by condensation, and so the little cumulus cloud can grow as long as it stays warmer than its surroundings. The visual manifestation of this is cumulus clouds growing taller within the mixed layer and, with more heating of the ground by the sun shining between the clouds, probably more cumulus clouds forming.

Stage 4: The vertical growth is limited by the stable layer at the top of the cool air mass. As soon as a growing cumulus tries to penetrate this layer, it finds itself cooler than its surroundings. It loses its buoyancy and stops rising. The mass of rising air within the cloud has no place to go except to spread laterally, and this causes the tops of the cumulus clouds to merge, forming a continuous layer of clouds: stratocumulus. Figure 2 shows many cumulus clouds, thicker than those in Figure 1, merging into a layer of stratocumulus.

Figure 1. Stratocumulus clouds photographed looking south from the Marblehead, Massachusetts, Causeway, afternoon, September 17, 2010.

Stage 5: Once formed, stratocumulus clouds have an interesting way of preserving themselves. The top of this cloud layer emits infrared radiation, much like the surface of the earth. If the air above the stratocumulus is clear and dry, most of this radiation can escape to space, causing cooling at the cloud top. This cooling can help to maintain a large lapse rate within the cloud, even destabilizing the cloud layer, inducing more turbulent mixing within and below the cloud layer that can supplement the loss of plume-like turbulence that originally created the puffy cumulus. This cooling also can enhance the stable layer above the cloud, thereby limiting mixing of cloud air with the drier air above.

Figure 2. Cumulus clouds at late morning near Green Bay, Wisconsin, merging into stratocumulus.

Stage 6: The reasons for dissipation of stratocumulus clouds are unclear, but here are some possibilities: 60–70% of the sunlight, which originally heated the ground and hastened the formation of stratocumulus, is reflected by the cloud tops, once formed. Some of the remainder is absorbed within the cloud, causing heating, which can evaporate some of the condensate. The overcast greatly reduces the amount of sunlight reaching the ground, and so the low-level lapse rate can become less than dry adiabatic, suppressing the mixing and reducing convection below the cloud layer. Nighttime also has the same effect. Mixing of cloud-top air with dry air above causes local evaporation of cloud droplets and cloud thinning. Drying of the air below the cloud base will raise the condensation level, possibly high enough that clouds cannot form below the stable layer.

Why were there more stratocumulus clouds over land? Energy from the sun heats a very shallow layer of soil, whereas it penetrates deeply into sea water. Thus the ground experiences more heating during the day than the sea surface. In September, the ground is likely to be warmer than the ocean, especially during the day. This allows the mixed layer to develop more rapidly over land than sea and to become much deeper. That, in turn, makes the formation of cumulus and stratocumulus clouds much more likely over land than sea. You mentioned an onshore breeze. This would keep the immediate shoreline cooler than a few kilometers inland, putting you in the transition zone between mostly cloudy skies to the west and mostly clear skies to the east.

The public cares whether the day will be sunny or cloudy, but the formation and dissipation of post-frontal stratocumulus clouds are difficult to forecast. The detailed vertical profiles of temperature, moisture, and wind in the lowest 10,000 feet must be observed before sunrise. That is possible with the network of radiosonde stations across the country. Helium-filled balloons rise through the atmosphere, measuring temperature and moisture. Tracking the balloons at each altitude allows a determination of wind speed and direction. The problem is that conditions in the lowest 10,000 feet change during the day, and the next balloon sounding doesn't occur until late afternoon or evening. Even subtle changes during the day affect whether the cumulus clouds will remain scattered cotton balls in the sky or grow and merge into a solid cloud deck.

Thanks to Joseph B. Olson of NOAA's Earth System Research Lab for help in answering the second question.

Weatherwise Contributing Editor THOMAS W. SCHLATTER is a retired meteorologist and volunteer at NOAA's Earth System Research Laboratory in Boulder, Colorado. Submit queries to the author at, or by mail in care of Weatherwise, Taylor & Francis, 530 Walnut St., Suite 850, Philadelphia, PA 19106.

In this Issue

On this Topic

© 2017 Taylor & Francis Group · 530 Walnut Street, Suite 850, Philadelphia, PA · 19106