By 2050, global average temperature could be between 1.4°C and 3°C warmer than it was just a couple of decades ago, according to a new study that seeks to address the largest sources of uncertainty in current climate models. That's substantially higher than estimates produced by other climate analyses, suggesting that Earth's climate could warm much more quickly than previously thought.
Many factors affect global and regional climate, including planet-warming "greenhouse" gases, solar activity, light-scattering atmospheric pollutants, and heat transfer among the land, sea, and air, to name just a few. There are so many influences to consider that it makes determining the effect of any one factor—despite years and sometimes decades of measurements—difficult.
Daniel Rowlands, a climate scientist at the University of Oxford in the United Kingdom, and his colleagues took a stab at addressing the largest sources of short-term climate uncertainty by modifying a version of one climate model used by the United Kingdom's meteorological agency. In their study, the researchers tweaked the parameters that influence three factors in the model: the sensitivity of climate to changes in the concentration of carbon dioxide in the atmosphere, the rate at which oceans absorb heat from the atmosphere, and the amount of cooling from light-scattering aerosols in the atmosphere.
Then the team analyzed the results of thousands of climate simulations—each of which had a slightly different combination of parameters—that covered the years between 1920 and 2080, Rowlands says. All of the simulations assumed that future concentrations of greenhouse gases would rise from today's 392 parts per million to 520 ppm by 2050. Each of the runs also allowed for variations in solar activity (which would affect how much the sun's radiation warms Earth) and rates of volcanic activity (which would influence the concentrations of planet-cooling sulfate aerosols in the atmosphere).
The team discarded results of simulations that didn't match observations of regional climate in more than 20 land areas and ocean basins from 1960 to today. Of those that passed this test, those considered statistically most likely—the two-thirds of those that best matched previous climate observations—suggest that global average temperature in 2050 will be between 1.4°C and 3°C warmer than the global average measured between 1961 and 1990 . All of the simulations that matched recent climate patterns suggested warming would be at least 1°C, the researchers report online today in Nature Geoscience.
The higher end of the team's range of likely warming scenarios is between 0.5°C and 0.75°C warmer than the scenarios published in the last report of the Intergovernmental Panel on Climate Change, Rowlands says. "These sorts of numbers haven't been seen in other complex climate models."
"I think people will be very interested in taking a close look" at details of the simulations that yielded unusually high estimates of warming to ensure the results are credible, says Isaac Held, a climate scientist with the National Oceanic and Atmospheric Administration in Princeton, New Jersey. Nevertheless, he notes, the team's set of simulations is a valuable resource for further analysis of climate change. Eventually, he suggests in a commentary  in Nature Geoscience, results of the new study may help scientists not only quantify the uncertainty in climate analyses but also reduce it.