Forests versus hurricanes

How trees may be a first defense in the battle against disaster

Photo by kakela/Flickr

By Douglas Sheil, originally posted on Forests News

The 12-year absence of major hurricane landfalls in the continental USA ended earlier this year: Hurricane Harvey, Hurricane Irma… perhaps the list and the destruction will continue. But what does this have to do with forests?

As an ecologist concerned with how forests respond to damage, I know how destructive cyclonic storms—hurricanes, typhoons and cyclones (henceforth simply ‘cyclones’)—can be. But aside from the ecological impacts, recent research suggests that forests and cyclones share a more fundamental link: their relationship with atmospheric moisture.

Both forests and cyclones are characterized by immense volumes of rain. While the Amazon forest, even deep within the South American continent, can maintain rainfall of over two meters a year, a single cyclonic storm can generate rain at a rate of two cubic kilometers per day. All this rain derives from the atmosphere.

“More forests may mean fewer and less destructive cyclones.” –Douglas Sheil, @CIFOR senior associate

Tweet Quote

If my colleagues Anastassia Makarieva and Victor Gorshkov are correct, the processes that sustain cyclones also sustain much of the world’s forests. At the same time, more forests may mean fewer and less destructive cyclones.

Recently, we have proposed and illustrated new ideas involving the physical principles that maintain rainfall in forested continental interiors, which are the same that power cyclones. Whether these ideas stand up to scientific scrutiny or not (and so far all the evidence suggests they will), there are good reasons to believe that extensive areas of forest can reduce both the likelihood and the severity of cyclones.


While the prevailing view is that winds are determined by temperature gradients, Anastassia, Victor and others, including myself, have advanced a theory describing how evaporation and condensation influence atmospheric dynamics.

While this theory was once viewed as controversial, accumulating understanding and evidence mean that it is gaining acceptance.

One important insight from this theory is that areas that sustain high rates of rainfall, such as forests and cyclones, also generate low surface-level atmospheric pressure that draws in surrounding low-level air and any water vapor it contains, which converges and rises, leading to condensation of any moisture it contains, feeding further pressure drops and resulting in a positive feedback that can be sustained as long as moist air is available. This mechanism explains how both forests and cyclones capture and concentrate such high levels of rain.

A growing body of evidence
Conventional atmospheric science assumes that the change in particle densities that occurs as water evaporates and condenses is largely irrelevant to understanding winds, as the key large-scale pressure differences are dominated by temperature. Our work indicates that this neglect is unjustified and that the change in molecule numbers during evaporation and condensation is a major mechanism shaping wind patterns and atmospheric moisture transport.The physical details of the theory and its implications have been published in peer-reviewed physics journals (1,2,3) and atmospheric science journals (4,5,6,7) and a compilation of relevant publications can be found here. We have developed introductions and summaries for non-specialists too (8,9,10). In advancing these ideas we have described failings in the temperature-driven theory of winds (11)and revised the fundamental equations governing atmospheric dynamics (12).

The proposed physical mechanisms are consistent with energetic and thermodynamic laws. Water vapor provides a source of energy, a substantial fraction of which can accelerate air when the vapor condenses, thus causing winds (this process can be further encouraged by aerosols—that is, by particles and compounds released from the forest into the atmosphere). This theory explains how high rainfall can be maintained within continental interiors and how cyclones, by concentrating so much available energy, can be so powerful.

Is there evidence for these ideas? Yes. For example, the theory predicts that rainfall over actively transpiring forests will be more likely when more moisture has locally accumulated, suggesting that there will be a positive difference in local atmospheric pressure prior to rain (when the difference switches to negative). This has been observed in long-term data from many sites in the Amazon, and is absent from neighboring, less forested regions.

Furthermore, if the maintenance of high rainfall in continental interiors is dependent on actively transpiring forests, it should disappear each winter over Siberia. Again, this prediction matches observations.

We can claim various other achievements, too. For example, if we estimate the global rate at which the kinetic energy of winds is generated (atmospheric power) we find that this also matches our theoretical predictions.

El Yunque Rainforest in Puerto Rico a few days before Hurricane Sandy. Photo by Timothy McAuliffe/Flickr


Predicting the intensity of tropical cyclones is a recognized challenge. While past efforts focused on energy derived from oceanic heat, the new theory indicates that overall storm power is determined by condensation of pre-existing atmospheric moisture. Furthermore, we predict that this power is linearly related to the total rate of precipitation as has been confirmed both by our own research and that of others.

Our work suggests that forests may protect continental regions from extreme storms. Both the formation and maintenance of cyclones appear to depend on sufficient supplies of water vapor. Extensive forests draw away moisture over the land. When cyclones do form near the coast of continental West Africa, they do so at a latitude where sufficient water vapor is available, that is, between that of wet forests and dry deserts (deserts yield only dry air), where the desiccating effect of neither dominates. This helps explain why these oceanic storms form in two belts 10°–20° north and south of the equator.

“Forests may protect continental regions from extreme storms.” –Douglas Sheil, @CIFOR senior associate

Tweet Quote

Evidence suggests that by importing atmospheric moisture from the ocean, forests deplete the vapor available to generate and support cyclones. Given their high vapor emissions, it may seem logical that forests themselves should generate and support cyclonic storms. But this does not occur because the severe drag-friction over forests absorbs energy and slows down winds.

In addition, forests have a strong diurnal cycle, and transpiration (unlike ocean evaporation) ceases overnight. Furthermore, while cyclonic storms concentrate air flows and resulting condensation into their eye, forests release moisture and aerosols in a more diffuse form, ensuring that condensation remains more broadly spread in both space and time. Thus, we believe that cyclonic storms cannot maintain themselves over, or near to, extensive forests. Though when storms do arise elsewhere, they may still ultimately be drawn toward the forest.


Such considerations and speculations suggest why cyclonic storms are near absent in the South Atlantic where any storm must arise while closely flanked by the major rain forests of South America and Africa (Hurricane Catarina in 2004 was the only exception). Any loss of these forests would increase available moisture over the neighboring ocean, and lower friction, and thus raise the likelihood of storms. Indeed all cyclones may become more frequent and more powerful as neighboring forest cover declines.

All these proposed influences are scale-dependent, and while the influence of large forests such as those in the Amazon are likely to be extensive, a narrow band of trees will have less impact. This suggests that, unfortunately, Australia’s remaining wet forests, and Madagascar’s remaining East coast forests will have only minor abilities to reduce cyclonic storms – though these could be boosted by large-scale rehabilitation.

We already know that forest loss has led to a marked reduction in atmospheric moisture over many regions with likely negative consequences for rainfall-dependent regions downwind. Once forests become more fragmented and reduced in extent they will not only provide less moisture to those downwind but will also have less impact on the local climate, likely leading to droughts. Droughts and subsequent fires in turn may start a cycle of further forest loss and depletion leading to further desiccation.

Forests play multiple roles in sustaining the global water cycle and stabilizing the climate, and removing forest will influence these processes negatively. We can now also add a possible increase in cyclonic storms to the list of potential effects. Ignoring the plight of forests is asking for trouble.

For more information on this topic, please contact Douglas Sheil at or Anastassia Makarieva at