The Intersection

i-526939160be7072173f64cb1b8f608bb-BarbaraTrackMay30.gif

Well, we’ve got a second named storm in the Northeast Pacific. As NHC forecaster James Franklin notes, this is an unusual (although not un-heard of) occurrence:

THE DEPRESSION IS UPGRADED TO A 35 KT TROPICAL STORM. ONLY TWICE BEFORE…IN 1956 AND 1984…HAVE THERE BEEN TWO EASTERN NORTH PACIFIC NAMED STORMS IN MAY.

Barbara is apparently steadily intensifying–the Wisconsin folks estimate the intensity at 45 knots by now. Because this storm could become a hurricane before striking Mexico, it is going to be watched very closely.

Meanwhile, what is the point of a storm like Barbara, or any other hurricane? Are these storms just random weather? Or are they a critical part of the climate system, one mechanism that the planet “uses” to keep cool around its tropical waistline?

That’s the question investigated in a new paper just published in the journal Nature by Ryan Sriver and Matthew Huber of Purdue University (PDF; subscription only). The paper is entitled “Observational evidence for an ocean heat pump induced by tropical cyclones.” Here’s the freely available editor’s summary of the work:

Tropical cyclones are known to mix the upper layers of the ocean, a process that ‘pumps’ heat downwards and thereby cools the ocean surface on local scales. It has been suggested that they may play an important role in ocean mixing at the global scale, and new calculations suggest that tropical cyclones are indeed responsible for significant cooling and vertical mixing of the surface ocean in tropical regions. Heat pumped downwards must be balanced by heat transport towards the poles, so some 15% of peak poleward oceanic heat transport may be associated with tropical cyclone-induced mixing. The amount of mixing induced by tropical cyclones is also related to sea surface temperature, suggesting that future changes in tropical sea surface temperatures may have significant effects on ocean circulation and ocean heat transport, as both processes are affected by ocean mixing. Climate change models could benefit from taking these effects into account.

This actually doesn’t do full justice to the potential implications of this work. Allow me to quote the authors themselves, who have kindly provided a copy of the paper. From their concluding paragraph:

Our analysis suggests that changes in global cyclone frequency, duration and/or intensity are closely related to the amount of heat pumped into–and available to be subsequently transported by–the oceans. This relationship may have implications for changes in heat transport associated with past and future climate change.
Extrapolation of our results suggests that future increases in tropical temperatures may result in increased dissipation, mixing, heat storage, and eventually heat transport. Moreover, this positive response in transport might feed back on climate by redistributing heat poleward, diminishing the Equator-to-pole temperature gradient, and
raising global mean temperature. We have provided some evidence that cyclone-induced mixing is a fundamental physical mechanism that may act to stabilize tropical temperatures, mix the upper ocean, and cause polar amplification of climate change.

In other words: Tropical cyclones play an essential role in getting heat out of the tropics, which they do by churning up the oceans and “pumping” warm water downward, in turn leading to a net poleward transport of ocean heat. So if the globe warms and hurricanes intensify, they will also send more heat poleward through the oceans–thus further warming the polar regions. The Arctic is already experiencing greatly amplified global warming; this could greatly increase that phenomenon.

The result, as the authors note, could be a considerable lessening of the equator-to-pole temperature gradient. And it could mean that one way the planet responds to global warming is to create dramatically more intense hurricanes.

Food for thought, to say the least. There is much more on this emerging line of scientific thinking in Storm World.

Comments

  1. #1 llewelly
    May 30, 2007

    The result, as the authors note, could be a considerable lessening of the equator-to-pole temperature gradient.

    To me, this implies the polar regions will warm significantly more than the global average, while the tropical regions warm significantly less (and that has to my knowledge been the trend, at least in the Northern Hemisphere).

    And it could mean that one way the planet responds to global warming is to create dramatically more intense hurricanes.

    I’m not clear where you are going with this. The prior sentence implies that global warming will have reduced impact on tropical ocean heat, because hurricanes will act to transport that heat away from the tropics. That is a negative feedback. It may mean that for a 2 C global rise, we see only 1 C rise in tropical ocean temps (with, say, 4 C in polar regions), with commensurate increases in hurricane intensities, rather than the 2 C rise in tropical ocean temps, and significantly greater increases in hurricane intensities, that would occur in the case of more or less uniform warming.

    I don’t see how hurricane induced-mixing strengthens (or weakens) the theory that global warming results in more intense hurricanes. (Unless it turns out that lessening equator-to-pole gradients has other effects which favor hurricanes, such as reducing shear (not unlikely), but that is another matter not covered by the paper.) Rather, it’s an impact of more intense hurricanes, and a negative feedback, at least to the extent that it moves ocean heat out of tropical regions.

  2. #2 Chris Mooney
    May 31, 2007

    Llewelly,
    I’m not the authors of this paper, but would guess they would respond that just because the equator-to-pole temp gradient would be reduced doesn’t mean the tropics won’t still be hotter than they are now.

    Also, yes, it’s a negative feedback–but in order for the feedback to be significant, the hurricanes have to be intense, because then more ocean heat transport occurs because more mixing occurs.

    That’s the hypothesis anyway. It goes back to the work of MIT’s Kerry Emanuel. All of this is in quite an early stage of research….it’s intriguing, though.

  3. #3 llewelly
    May 31, 2007

    Chris:

    I’m not the authors of this paper, but would guess they would respond that just because the equator-to-pole temp gradient would be reduced doesn’t mean the tropics won’t still be hotter than they are now.

    Agreed. In my hypothetical example I wrote ‘only 1 C rise in tropical ocean temps (with, say, 4 C in polar regions), with commensurate increases in hurricane intensities,’ . When I write ‘negative feedback’, I’m thinking something akin to a brake (whose resistance increases as a function of velocity), not an instantaneous reverse.

  4. #4 huber
    June 6, 2007

    I usually try to avoid getting involved in blogging conversations, but “llewelly” makes a reasonable point and it might be helpful for me to expand upon it.

    In a previous paper (Sriver and Huber, 2006, GRL) we showed a strong positive relationship between the low-pass filtered time series of integrated tropical storm intensity (PD) and SST. Thus if ‘global warming’ leads to an increase in tropical temperatures we expect to see some increase in PD and hence an increase in ocean mixing. As llewelly points out, perhaps we might see a degree of warming whereas we might have seen two degrees of warming without the feedback. Thus we would also expect tropical cyclones to not increase in integrated intensity as much as if this (local) negative feedback did not exist. I think this is the point that llewelly was trying to make and it is probably completely valid. Nevertheless, greenhouse gas induced warming->increased PD->somewhat reduced tropical warming->somewhat reduced increase in PD still is still potentially a recipe for increased warming at the poles (via the advection of the downward pumped heat to higher latitudes) which can feedback onto sea ice formation. This is only speculation at this point. If, however, this feedback does occur then global mean temperature may increase (see the relevant citations in our paper–it’s a nonlinear feedback due to albedo changes) thus potentially wrecking that nice negative feedback in the tropics we started out with.

    Also note that the process really changes quantitatively once the water that is upwelled in cold wakes is not so cold anymore (see Emanuel’s original papers (2002) on this on his web site).

    So you’re both right. :)

  5. #5 Timothy Chase
    June 18, 2007

    One point which occured to me…

    You had pointed out a while back that although 2006 was relatively pacific in the Atlantic, it was a record year for cyclones in the Pacific ocean. Given this and the principle that hurricanes increase poleward flow, this would seem to suggest that we should expect there to have been a higher amount of warm water currents entering the arctic waters by way of the Bering Strait. I know that they monitor the water currents – although I am not sure where to get the information to verify this.

    Have we seen just this sort of correlation?

  6. #6 Alastair B. McDonald
    June 19, 2007

    Tim,

    There is evidence on my web page that the sea ice in the region of the Arctic fed by the Bering Strait is showing record low concentrations. This could be the result of warm water entering from the Pacific. See: http://www.abmcdonald.freeserve.co.uk/north.htm

  7. #7 Per
    June 19, 2007

    Or to cite the honourable Henry Louis Le Chatelier:

    If a tropical system at equilibrium experiences a change in concentration(of water vapor = local sea temperature), temperature(Cloud tops), volume(of hot water below = depth of thermocline), or total pressure(at sea level); the equilibrium will shift in order to partially counter-act the imposed change(tropical sst’s).. and make global warming worse.

  8. #8 llewelly
    June 19, 2007

    huber, thank you (belatedly) for your informative response.

  9. #9 Timothy Chase
    June 20, 2007

    Thank you, Alastair.

    Incidently, I remembered that I actually brought up loss/gain of energy in collision between absorbtion and re-emission first – but then you came in with the numbers a couple of days later. I figured the latter was more important.

    *

    Per – I will take you word for it – at the moment: my wife undoubtedly expects me to go to the grocery store and it is almost time to call it a night.

  10. #10 Timothy Chase
    July 29, 2007

    You probably know about it already, but there is a paper coming out tomorrow:

    Heightened Tropical Cyclone Activity in the North Atlantic: Natural Variability or Climate Trend?
    Greg J. Holland and Peter J. Webster
    Online July 30,2007
    Philosophical Transactions of the Royal Society of London

    It reads like a (critical) review of the literature for the first six pages (and has three or so pages of references) then moves on to their analysis.

    They identify and filter out short-term cyclic phenomena using the sort of techniques Tamino is highly skilled at (e.g., in order to remove the variability associated with El Nino-Southern Oscillation), and are able to identify three distinct climate regimes from 1905 to present.

    Coming from a strict amateur after having done little more than a quick scan, I would have to say that I am impressed.

    I will be looking forward to hearing your views…

The site is currently under maintenance and will be back shortly. New comments have been disabled during this time, please check back soon.