The genesis of two category-five hurricanes (Katrina and Rita) in a row over the Gulf of Mexico is an unprecedented and troubling occurrence. But for most tropical meteorologists the truly astonishing “storm of the decade” took place in March 2004. Hurricane Catarina–so named because it made landfall in the southern Brazilian state of Santa Catarina–was the first recorded South Atlantic hurricane in history.

Textbook orthodoxy had long excluded the possibility of such an event; sea temperatures, experts claimed, were too low and wind shear too powerful to allow tropical depressions to evolve into cyclones south of the Atlantic equator. Indeed, forecasters rubbed their eyes in disbelief as weather satellites downlinked the first images of a classical whirling disc with a well-formed eye in these forbidden latitudes.

In a series of recent meetings and publications, researchers have debated the origin and significance of Catarina. A crucial question is this: Was Catarina simply a rare event at the outlying edge of the normal bell curve of South Atlantic weather, just as, for example, Joe DiMaggio’s incredible fifty-six-game hitting streak in 1941 represented an extreme probability in baseball (an analogy made famous by Stephen Jay Gould)? Or was Catarina a “threshold” event, signaling some fundamental and abrupt change of state in the planet’s climate system?

Scientific discussions of environmental change and global warming have long been haunted by the specter of nonlinearity. Climate models, like econometric models, are easiest to build and understand when they are simple linear extrapolations of well-quantified past behavior–that is, when causes maintain a consistent proportionality to their effects.

But all the major components of global climate–air, water, ice and vegetation–are actually nonlinear: At certain thresholds they can switch from one state of organization to another, with catastrophic consequences for species too finely tuned to the old norms. Until the early 1990s, however, it was generally believed that these major climate transitions took centuries, if not millennia, to accomplish. Now, thanks to the decoding of subtle signatures in ice cores and sea-bottom sediments, we know that global temperatures and ocean circulation can, under the right circumstances, change abruptly–in a decade or even less.

The paradigmatic example is the so-called “Younger Dryas” event, 12,800 years ago, when an ice dam collapsed, releasing an immense volume of meltwater from the shrinking Laurentian ice sheet into the Atlantic Ocean via the instantly created St. Lawrence River. This “freshening” of the North Atlantic suppressed the northward conveyance of warm water by the Gulf Stream and plunged Europe back into a thousand-year ice age. Abrupt switching mechanisms in the climate system–such as relatively small changes in ocean salinity–are augmented by causal loops that act as amplifiers. Perhaps the most famous example is sea-ice albedo: The vast expanses of white, frozen Arctic Ocean ice reflect heat back into space, thus providing positive feedback for cooling trends. Alternatively, shrinking sea-ice levels increase heat absorption, accelerating both further melting and planetary warming.

Thresholds, switches, amplifiers, chaos–contemporary geophysics assumes that earth history is inherently revolutionary. This is why many prominent researchers–especially those who study topics like ice-sheet stability and North Atlantic circulation–have always had qualms about the consensus projections of the Intergovernmental Panel on Climate Change (IPCC), the world authority on global warming.

In contrast to Bushite flat-earthers and shills for the oil industry, these researchers base their skepticism on fears that the IPCC models fail to adequately allow for catastrophic nonlinearities like the Younger Dryas. Where other researchers model the late-twenty-first-century climate that our children will live with upon the precedents of the Altithermal (the hottest phase of the current Holocene period, 8,000 years ago) or the Eemian (the previous, even warmer interglacial episode, 120,000 years ago), growing numbers of geophysicists toy with the possibilities of runaway warming returning the earth to the torrid chaos of the Paleocene-Eocene Thermal Maximum (PETM: 55 million years ago), when the extreme and rapid heating of the oceans led to massive extinctions.

Dramatic new evidence has emerged recently that we may be headed, if not back to the dread, almost inconceivable PETM, then to a much harder landing than envisioned by the IPCC. As I flew toward Louisiana and the carnage of Katrina three weeks ago, I found myself reading the August 23 issue of EOS, the newsletter of the American Geophysical Union. I was pole-axed by an article titled “Arctic System on Trajectory to New, Seasonally Ice-Free State,” co-authored by twenty-one scientists from almost as many universities and research institutes. Even two days later, walking among the ruins of the Lower Ninth Ward, I found myself worrying more about the EOS article than the disaster surrounding me.

The article begins with a recounting of trends familiar to any reader of the Tuesday Science section of the New York Times: For almost thirty years, Arctic sea ice has been thinning and shrinking so dramatically that “a summer ice-free Arctic Ocean within a century is a real possibility.” The scientists, however, add a new observation–that this process is probably irreversible. “Surprisingly, it is difficult to identify a single feedback mechanism within the Arctic that has the potency or speed to alter the system’s present course.”

An ice-free Arctic Ocean has not existed for at least 1 million years; the authors warn that the earth is inexorably headed toward a “super-interglacial” state “outside the envelope of glacial-interglacial fluctuations that prevailed during recent Earth history.” They emphasize that within a century, global warming will probably exceed the maximum Eemian temperature and thus obviate all the models that have made this their essential scenario. They also suggest that the total or partial collapse of the Greenland Ice Sheet is a real possibility–an event that would definitely throw a Younger Dryas wrench into the Gulf Stream.

If they are right, then we are living on the climate equivalent of a runaway train that is picking up speed as it passes the stations marked “Altithermal” and “Eemian.” “Outside the envelope,” moreover, means that we are not only leaving behind the serendipitous climatic parameters of the Holocene–the last 10,000 years of mild, warm weather that have favored the explosive growth of agriculture and urban civilization–but also those of the late Pleistocene that fostered the evolution of Homo sapiens in eastern Africa.

Other researchers undoubtedly will contest the extraordinary conclusions of the EOS article and–we must hope–suggest the existence of countervailing forces to this scenario of an Arctic albedo catastrophe. But for the time being, at least, research on global change is pointing toward worst-case scenarios.

All of this, of course, is a perverse tribute to industrial capitalism and extractive imperialism as geological forces so formidable that they have succeeded in scarcely more than two centuries–indeed, mainly in the last fifty years–in knocking the earth off its climatic pedestal and propelling it toward the nonlinear unknown.

The demon in me wants to say: Party and make merry. No need now to worry about Kyoto, recycling your aluminum cans or using too much toilet paper, when, soon enough, we’ll be debating how many hunter-gatherers can survive in the scorching deserts of New England or the tropical forests of the Yukon.

The good parent in me, however, screams: How is it possible that we can now contemplate with scientific seriousness whether our children’s children will themselves have children? Let ExxonMobil answer that in one of its sanctimonious ads.