By the year 2047 the mean air temperature around the planet will shift completely out of the range seen in recent history. From that point on, even a cold year will be warmer than any warm year from 1860 to 2005 if nations continue to emit carbon dioxide the way they do now. And the new extreme temperatures—the new normal—will first occur not in the Arctic but in the tropics, where people, plants and wildlife are least equipped to adapt. That disquieting analysis comes from a massive new study led by Camilo Mora at the University of Hawaii at Mnoa, published today in Nature. (Scientific American is part of Nature Publishing Group.)

The report begins with the recognition that the annual mean global air temperature fluctuates from year to year, even though it has been climbing overall since the industrial revolution began. If the world does nothing new, then the temperature by 2047 will exceed even the highest annual temperature from 1860 to 2005. If the world aggressively cuts back on emissions, then the date at which the temperature fully departs from modern history will be delayed until 2069. Mora and his colleagues crunched data from 39 different climate models that feed two future scenarios: business as usual (leading to the 2049 date) and aggressive mitigation (the 2069 date). The same two scenarios are cornerstones of the new climate assessment released two weeks ago by the International Panel on Climate Change.

Mora’s team also ran the exercise for six other variables: sea-surface water temperature, evaporation, heat flux, ocean pH, precipitation and transpiration (moisture emitted by plants). They then generated a map for each variable showing when the date of departure occurs for all places around the world (see above and below).

The surprising revelation is that “the earliest and most extreme effects will happen in the tropics,” Mora says, and not in the Arctic, where so much concern is focused. That’s because the natural variation in air and ocean temperatures and other climate-related variables in the tropics is very small, so only relatively minor changes will push these conditions out of bounds. As a result, Mora says, tropical plants and animals will be stressed the most, because species accustomed to living within such narrow fluctuations will not be able to adapt quickly enough to such unprecedented and rapid climate changes. More of the world’s species live in the tropics than anywhere else, so global biodiversity will be hit hard.

Many of the world’s low-income nations are also located at tropical latitudes, and they are likely to struggle to adapt to the new extreme climate. As the paper concludes: “This suggests that any progress to decrease the rate of ongoing climate change will require a bigger commitment from developed countries to decrease their emissions but will also require more extensive funding of social and conservation programs in developing countries to minimize the impacts of climate change.”

One question for policy makers is whether the years 2047 and 2069 are firm transition points. After all, those departure dates are compared with only a 145-year period, from 1860 to 2005. Mora says the uncertainty of the calculations for 2047 is plus or minus three years and for 2069 is about four years. To add further rigor Mora’s team ran the same analysis against the last 1,000 years of global temperatures and found that each of the two temperature departure dates only increases by two years. The reason, Mora says, is that in 145 years “you pick up enough extremes to give a reliable long-term curve.”

Although air temperature will cross over into a new normal of extremes in only a few decades, ocean surface temperature will cross over soon as well (see map above). Alarmingly, ocean pH already crossed over in 2008, both because pH has varied extremely little, historically and the oceans are absorbing a large percentage of man-made CO2 emissions, which turns the water more acidic.

The level of data generated by the study is substantial. One map specifies when local temperature will depart for various cities around the world, and it reflects the early extremes in the tropics. For example, the departure date is 2029 in Jakarta and Lagos, 2031 in Mexico City, 2047 in New York City, 2066 in Reykjavik and 2071 in Anchorage.

Another map posted on Mora’s lab page allows the reader to pick any one of 10,000 spots on the planet (Chicago, New Delhi, the South Pacific) and see the entire air temperature curve from 1860 to 2013, its projection to 2100 and the date at which climate departure will occur. The paper, data and graphs can be downloaded from the lab page, too.

Maps reprinted with permission from Nature and Camilo Mora.