To understand how we can end the threat of tropical cyclones (hurricanes, typhoons, cyclones), it is useful to review the basics on how these systems develop.
Targeting the problem at the source
Most Atlantic hurricanes that make US landfall begin from a limited portion of the tropical Atlantic: the Caribbean Sea, a small region just South of the Cape Verde Islands (where over 75% of Atlantic tropical storms originate), the Gulf of Mexico, and off the coast of Venezuela and Guyana. Hurricanes seldom form elsewhere because conditions for hurricane genesis are rarely met elsewhere in the Atlantic.
The Pacific Ocean and the Indian Ocean gyres are more complex, and weather patterns can emerge that would allow genesis of cyclones in more numerous small regions. But there are still a limited number of relatively small regions that simultaneously meet all of the conditions for cyclone genesis. These regions can be targeted and treated by Tropical Cloud Generation (TCG).
Hurricanes begin as partial circular flows over warm seas that fit the following qualifications:
1. The Sea Surface Temperature (SST) – usually defined at a depth of 10 m – is above 26.5°C,
2. The relative humidity in the atmosphere over a large area up to the mid troposphere (~5 km) is very high (above 80%),
3. There is low atmospheric shear to altitudes of at least 7 km, and
4. There is a sufficient level of cyclonic vorticity in the lower troposphere in the genesis region (they need a little kick to get started because the warm SST extends over such a long distance).
5. The region must be sufficiently spaced away from the Equator (usually at least ~5 degrees latitude).
If all of those conditions aren’t met, then a hurricane won’t begin, which means that by treating a region in a way that removes one of those conditions, you prevent hurricanes from starting.
Hurricane genesis overview
Water vapor is less dense than atmospheric air (which may be counter-intuitive because it is common to think of humid air in reference to air laden with small liquid water particles – like fog – and liquid water is much heavier than air, but water vapor only has ~62% of the density of atmospheric air). In regions over warm ocean surfaces, the warm low-density humid air rises due to its buoyancy, and the low-density column of air creates a central low pressure region. The low pressure then causes inflow at sea-level from the surrounding area, which – drawn along the surface of the surrounding warm sea – takes on greater humidity and continues to drive the buoyancy-driven rising column of air, driving an ascending vortex that creates increasing atmospheric boundary-layer inflow and tangential cyclonic flow (counter clockwise in the Northern Hemisphere) from the Coriolis effect. Coriolis effect.
Right: Schematic representation of inertial circles of air masses in the absence of other forces, calculated for a wind speed of approximately 50 to 70 m/s (110 to 160 mph). Source – “Coriolis force” Wikipedia
If there is little high-altitude wind shear, this ascending vortex of moist air near the center rises, expands, and cools, causing its moisture to condense out in heavy rain. When it reaches the upper troposphere, it mostly flows outwards in an anti-cyclonic flow (clockwise in the Northern Hemisphere), and continues to cool, causing more moisture to condense out. The air rushing inward towards the low pressure rises in storm bands of high convection interspersed by regions where the cool dry heavy air falls back towards the sea – where it will once again become warm and humid from the surface-wind-driven fluxes of moist enthalpy at the sea surface as it is pulled inward towards the low pressure center, thus completing the positive-feedback heat-engine loop. (Earlier we said the central rising air “mostly flows outwards” because in fact once the storm becomes well developed, the strong upward flow just outside the eyewall drives eddies near the troposphere that produce substantial downward flow of dry air in the center of the eye – which isn’t too clear in the graphic shown here.)
If these circular flows are not disrupted they can slowly build into massive closed heat engines operating between the warm sea surface and the cool upper troposphere. Their intensification and maintenance against normal dissipative mechanisms depends on self-induced heat transfer from the ocean. If the seas are cool when the cyclone travels over it, its strength is greatly reduced. If high altitude wind shear is present, the circular system cannot close and organize, and existing cyclones will weaken and see their organized circular systems disrupted.
Tropical cyclones cannot build into monsters unless all four of the prerequisites listed above are present. Once they do develop, they cannot maintain their strength over regions of cooler seas, low humidity, or high wind shear. These facts then offer us a path to disrupt their formation: eliminate at least one of these conditions in regions that have all of these conditions, and there are no hurricanes.
Anticyclones
The same mechanisms could be reversed by cooling a small patch of ocean surface surrounded by warm ocean: The air over the cool patch would become heavier and sink, pushing outwards along the warm surrounding sea surface (anticyclonic because it’s outward flow), absorbing heat and moisture as it flows outward, rising, and flowing back inward at higher altitude (here cyclonic because it’s inward flow) over the cool patch as the cooled column of air over the cool patch descends. That completes the closed loop of the anti-cyclone. A big difference is that conservation of angular momentum causes the velocity of the wind in the anticyclone to decrease as it flows away from the eye – the same principle that causes the velocity of the wind to increase in the cyclone as it flows toward the eye. Another big difference is that the cyclone maintains its organization by continually absorbing heat from the warm waters that surround it for hundreds of kilometers – so it can easily travel with the winds over the warm ocean after it is organized and maintain its structure (as it generates its own self-sustaining pressure patterns that are stronger than the influence of mild wind shear). The anti-cyclone, on the other hand, requires a centralized cooled zone that is much cooler than the surrounding waters in order to maintain its organization, which means that it must remain fixed to the cooled patch (which will drift with the speed of the surface currents) and will gradually wind down as the waters surrounding the cooled patch are also cooled from the cloud cover and increased evaporation rate.
The center of the cool patch might be expected to slowly warm because its longwave radiation is reduced and the skies above it are cloudless (as in the eye of a cyclone/hurricane). However, preliminary calculations suggest the evaporative cooling into the dry air above it will greatly exceed the above effects and will actually further cool and expand the cool patch until it grows to the point that dissipative boundary-layer mechanisms exceed convective forces. Because the anticyclone depends on staying “tethered” to the cool patch (which is relatively small), it is more easily disrupted by surface and mid-tropospheric winds unless it is quite large. The separate small anticyclones that would develop over each of the separate cool patches (where winds are very low) could also drive small separate cyclonic storms over the warm waters between the cool patches, but these storms would be nothing like hurricanes and would not combine to form such because of the anticyclonic vorticity of the neighboring anticyclones and the reduced mean sea surface temperature. Simulations are needed to better understand the interplay of these effects, the optimum starter size for the cool patches, and the extent to which they could grow. However, current global circulation models would probably not be the best place to begin, as they have transverse resolution of only ~100 km, and transverse resolution of no more than a 200 m is needed to adequately capture what will happen here, as will become apparent in the next section, on Sea Breezes.
Sea-Breeze-induced Clouds and Storms
When the winds are not very low, anticyclones will not develop around the cool patches, but there will still be downward convection over the cool patches and upward convection over the warm water (mostly downwind) where clouds and rain can be expected, as often seen on-shore in the afternoon on tropical islands and sometimes offshore from a land breeze in the very early morning.
For an introduction to the meteorology of sea breezes, the interested reader is referred to this page , which summarizes a more in-depth study and simulations that are validated from weather data off the east coast of central Florida in August, which we briefly summarize below.
When the synoptic winds (the prevailing mid-tropospheric winds over a horizontal scale of more than 1000 km) are onshore (from offshore), the sea breeze front (SBF, the leading edge of the cool marine front, in the lower 1-2 km) begins in mid morning and penetrates inland at speeds of 10-30 km/hr, to distances as far as 300 km by early evening, though a distance of 20-100 km is more typical. Transverse wind shear in the lower 1 km, transverse ground-level thermal gradients, and strong mid-day heating over land stimulate and drive Horizontal Convective Rolls (HCRs) aligned with the wind direction that alternately add and subtract to the convective updraft at the SBF, and clouds then often form in “streets” aligned with the updrafts between the HCRs. These HCRs are typically ~5 km wide and 1-2 km high by late afternoon. Peak vertical velocities up to 4 m/s aren’t uncommon, and the HCRs can be 100 km long, though 30 km is more typical. Storms are more likely to develop if the onshore synoptic winds are below 10 km/hr, but they are possible even with onshore winds more than twice that.
When the synoptic winds are offshore (from onshore), the sea breeze penetration is much more limited (sometimes only a few km) but the frontal contrast is much greater and thunderstorms are more likely, even with synoptic winds above 35 km/hr. The figure below illustrates a portion of the progression from mid morning to mid afternoon for the case of mild offshore synoptic winds.
Monthly Weather Review, May, 1999 (volume 127m p. 858
Not shown above are the weaker perpendicular updraft and downdraft rolls that also form ahead of the SBF, along the front, before initiation of the HCRs, especially for the case where the synoptic wind is parallel to the coast line. These perpendicular rolls might have peak updraft velocities up to 3 m/s at about 1.5 km altitude directly above the (ground-level) front line, where they form a roll cloud, and rapidly decrease to small downdraft (under 0.2 m/s) about 10 km ahead of the front line.
It’s worth repeating: These atmospheric effects – the sea breeze, the HCRs, the cloud arrays, the storms – are all driven by the large thermal gradient between the land and the sea. The above illustrated case (afternoon sea breeze, where the land is warm and the sea is cool) is somewhat similar to what is expected on the upwind side of a cool patch, where the synoptic wind is coming from over warm waters and colliding with a cool breeze in the lower few hundred meters coming toward (and under) it from the cool patch. However, there are several differences that we suspect will lead to greater cloud and storm generation relative to normal sea breeze cases.
1. For the coastal case, the mid-afternoon thermal gradient seen in the first 30 km inland from the coast is typically ~40 mK/km (at 300 m altitude). Over the open seas, the only time that a thermal gradient of more than ~3-4 mK/km can be found in nature is in an active cyclone or at the edge of a storm front. However, in the case of Tropical Cloud Generation (TCG), a single sea mixer would be able to mix a 60-km diameter patch of cooled sea surface about a monthwith an initial thermal gradient of 1000 mK/km for several km around it (though this will gradually decrease as the HCRs and cloud cover gradually cool the surrounding seas, and the cool patch gradually warms).
2. The surface-level thermal gradients are continually strong (day and night), not just for 6-8 hours at a time. (So, the sea breeze effects can build continuously for weeks as the region surrounding the mixed patch cools, rather than gradually building and then quickly ebbing every day).
3. The warm air for the patchy mixing case will be much more humid (it’s over water, not land).
4. The parallel HCRs, which dramatically enhance convective transfer from the warm surface to the atmosphere above it, should be stronger for the patchy mixing case than for normal sea breeze case. In the patchy mixing case, there will be places where the synoptic winds are parallel to large surface-level thermal gradients (on the upwind and downwind sides of the cool patches) and there will be as many places where they are perpendicular to them (on the sides of the cool patches). The high transverse thermal gradients in the synoptic flow along the sides of the cool patches are expected to be extremely effective at driving strong HCRs, but now not just over warm surfaces because HCRs couple strongly transversely. Those initiating over the sides of the cool patches (with downdraft over the cool patch and updraft over the warm water) will drive additional HCRs in both directions, over the warm waters and the cool patches, dramatically increasing the convective cooling rate of the lower 2 km of the atmosphere over the cool patches and further increasing the front intensity and cloud generation.
5. Finally, in addition to convective and turbulent mechanisms (and evaporation and condensation) there is also radiative coupling, which is very small, but increases more rapidly than linearly with moisture content. The typical relatively dry mid-latitude atmosphere is very transparent to longwave radiation in the 8-13 microns range, and radiation from the tropical sea is mostly in the 5-16 microns range. However, absorption and emission in moist tropical air (at wavelengths below 8 microns, above 13 microns, and at some narrow bands in between – ozone at 9.6 microns, and H2O-H2O dimer modes at 8.2, 10.1, 11.2, and 13.7 microns) provides some instant thermal coupling even without clouds or turbulence. That coupling initiates slow cooling of the lower 2-3 km of the troposphere when clear warm air blows from warm waters to the first cool patch in its path even if there were zero turbulence (which will never be the case). This cooling alone would be great enough in only ~half an hour (after the first 3-8 km over a cool patch) to start the convection and increased turbulence which would then dominate as the flow continues over the first cool patch and subsequent warm spaces and cool patches. Hence, the first cool patch might not create clouds on its upwind side (though it certainly would downwind), but once the turbulence and convection in the atmospheric boundary layer is initiated, subsequent cool patches would then create lots of clouds and storms both upwind and downwind.
Again, simulations are needed to better understand what can be expected for different sizes and spacing of cool patches with various wind speeds. The most important optimization criterion will be maximizing cloud generation for typical conditions, as clouds can have a negative climate forcing of more than 120 W/m2.