Several decades ago, a pair of physicists proposed a model called the thermostat hypothesis, describing how the formation of clouds over Pacific waters sets an upper limit on how hot the tropics can get.

Recent findings have rebutted the need for any such model, showing that temperatures probably exceeded the hypothetical limit during a warm spell 56 million years ago.

First proposed in 1991, the thermostat hypothesis has had a warmer reception among skeptics of human-caused global warming than with most climatologists, because it suggests that there's a limit to how bad things can get.

Using evidence of temperatures collected in tropical waters, proponents of the hypothesis have suggested that the tropics are unlikely to ever reach average sea surface temperatures much greater than 31 degrees Celsius (88 degrees Fahrenheit), thanks to a feedback loop between heat and cloud formation that stabilises temperatures over those areas of the ocean.

While there have been responses criticising the hypothetical mechanisms, there has been limited evidence demonstrating the temperature of tropical waters consistently rising above 31 degrees at any time in history.

Now a team from Purdue University has published evidence that shows not only were tropical ocean waters warmer during a period called the Paleocene-Eocene Thermal Maximum (PETM), they were the warmest the ocean has been in the past 100 million years.

This period of history – also called the Eocene Thermal Maximum 1 – falls roughly 200,000 years when the average global temperature was about 8 degrees Celsius (14 degrees Fahrenheit) warmer than today, most likely the result of carbon dioxide being released into the atmosphere over a period of a few thousand years or so.

Calculating the temperature of the ocean millions of years in the past requires digging up sediments which were deposited at that time and then analysing the chemistry of the remains of dead marine organisms such as foraminifera.

In this case, the researchers had their job cut out for them, given the difficulty in accessing tropical marine sediment from the time of the PETM.

They found a thin section of material in Nigeria, and studied the types of organisms within it, as well as measuring the ratio of isotopes in their shells and analysing an organic residue called TEX86 to estimate temperatures during that period.

Combined, their results indicated sea surface temperatures of around 33 degrees Celsius (91 degrees Fahrenheit), rising to a lofty 36 degrees Celsius (97 degrees Fahrenheit) during the PETM.

"This is the first time that we've found really good information, in a very detailed way, where we saw major changes in the tropics directly associated with warming past a key threshold in the past 60 million years," said one of the team, Matthew Huber.

The PETM was accompanied by a mass extinction event, with up to half of all foraminifera disappearing over the course of about 1,000 years, eroding the foundations of a far-reaching food web.

"The records produced in this study indicate that when the tropics warmed that last little bit, a threshold was passed and parts of the tropical biosphere seem to have died," said Huber.

Huber and his team attribute this mass die-off to the swift warming of the tropical oceans, posing a stark warning of similar events occurring in our not too distant future.

Of course, some might take comfort in the fact that the world has seen such temperatures in the distant past, with ecosystems inevitably bouncing back in the long run.

But while the temperature rise might have a precedent, such a rapid rise in ocean temperature will undoubtedly spell catastrophe for communities in poorer parts of the world, whose territories overlap with much of the tropics.

"If you say there's no tropical thermostat, then half of the world's biodiversity – over half of the world's population, the tropical rainforests, the reefs, India, Brazil – these populous and very important countries have nothing to prevent them from warming up substantially above conditions that humans have been used to," said Huber.

Additional research could still provide data supporting the thermostat hypothesis, but for now, it looks like we'll need to base our predictions of the future on tropical waters that could get a lot warmer than 31 degrees Celsius. 

This research was published in Science Advances.