Why has global warming paused? Pacific Ocean's 'engine room' running cool.

Global warming has been put on a 15-year (and counting) hold by a prolonged period of cold ocean temperatures in the eastern tropical Pacific – part of a natural climate pattern that should allow the rate of warming to pick up when the pattern shifts, according to a new study.

During the past 15 years, warming has continued. Indeed, the decade from 2000 to 2009 was the warmest on record globally, with 12 of the 14 warmest years on record falling between 2001 and 2012. But the warming has occurred so slowly that, statistically, the rate of warming per decade could just as easily have been zero, researchers say. This real-world pace was far slower than the pace found in computer simulations of climate change for the same period

The hiatus triggered finger-wagging from some of the more strident climate-change skeptics, as well as chin-scratching among many climate scientists. With carbon-dioxide emissions rising relentlessly to levels not seen in hundreds of thousands of years, how could atmospheric temperatures fail to respond in a stronger way, many asked.

To tackle that question, two researchers at the Scripps Institution of Oceanography in La Jolla, Calif., turned to a sophisticated climate model, but used it in what they say is a novel way.

Typically, researchers feed an array of data into the model and then turn it loose to see how the climate system evolves with time. But for this study, researchers Yu Kosaka and Shang-Ping Xie did not let the model compute sea-surface temperatures in the eastern tropical Pacific – a patch of ocean representing less than 10 percent of the Earth's surface. Instead, they specified those data.

The reason is that the distribution of sea-surface temperatures along the tropical Pacific can have far-ranging effects on climate, Dr. Xie says.

"That's the engine room of the global climate system," he says. "If you mess up the engine room, you'll get a huge response."

The region undergoes both short- and long-term cycles in sea surface temperature. The short term cycles are known as El Niño and La Niña, and they exchange dominance a few times times a decade. They fall within longer-term swings known as the Topical Pacific Decadal Oscillation, during which one phase can hold sway for several decades. The cool phase currently is in place, prompting the researchers to test its influence on the hiatus.

When the scientists let the model run on its own, without specifying the ocean temperatures, the global average temperatures tracked the rise in real-world temperatures fairly well until around 2003. Then the two diverged. The trend in modeled temperatures kept rising, but the trend in real-world temperatures reached a rough plateau.

When the duo added observed sea-surface temperatures to the model, the model's temperature record matched well with the real-world temperature record, even in the '00s.

They also ran the model with specified sea-surface temperature, but they held rising greenhouse gases to 1990 levels with no additional variations to the data. Under these conditions, the ups and downs in temperatures over time closely tracked the ups and downs in tropical eastern Pacific sea-surface temperatures – illustrating fairly tight control the eastern tropical Pacific holds over the climate system in the absence of rising greenhouse-gas levels.

That still leaves a conundrum, Xie suggests.

"For the past 11 years, global average temperatures remain flat, but yet you hear the news of a Russian heat wave, a US heat wave, and this year, a Chinese heat wave," he says.

The duo found that the hiatus was most evident as cooler-than-normal winter temperatures in the northern hemisphere.

When summer comes, however, that influence wanes. "We're still under siege by greenhouse-gas warming. We're still baked during summer," he says.

The work confirms an aspect of a 2011 study, which aimed to hunt for heat that wasn't showing up in the atmosphere, but also wasn't appearing at the top of the atmosphere in satellite measurements. The team's modeling study suggested that the missing heat actually was getting stored in the deep ocean – a byproduct of cooler oceans, which have a higher capacity for storing heat than warmer water. The team, led by Gerald Meehl of the National Center for Atmospheric Research in Boulder, Colo., also suggested that cooling due to natural variability in the Pacific was the agent responsible for the hiatus.

For Judith Curry, a climate scientist at Georgia Tech who heads the School of Earth and Atmospheric Sciences, the new study accomplishes an important feat.

The model Dr. Kosaka and Xie used is a type that usually does "a poor job of simulating the natural internal variability," particularly with natural climate swings that span several decades, as the Tropical Pacific Decadal Oscillation does, Dr. Curry writes in an e-mail. Kosaka and Xie have devised a way to represent those hard-to-model swings in a realistic way.