Effects of Natural Variability...

Effects of Natural Variability on Projections of Changing Climate: A Study in Nuance

Humans and their emissions have an undeniable effect on global, regional, and local climate; however, natural climate variability introduces shades of gray into the prediction of future effects of climate change. Extreme events such as “Snowmageddon,” the series of blizzards that hit the U.S. East Coast in 2009-2010, or even Colorado’s cooler-than-normal 2013 summer demonstrate deviations from typical climate patterns. These out-of-the-norm events indicate what happens when natural climate variability and human-driven climate change intersect, explains Clara Deser, a senior scientist at the National Center for Atmospheric Research.

While average global temperatures continue to rise due to human emissions, changes in temperature and precipitation will not occur uniformly across regions. How future climate change manifests locally will differ from the global averages predicted by models, with some localities experiencing cooling rather than warming trends, for example, because of the effects of natural climate variability.

“Climate models do a good job of projecting the effects of human-made (anthropogenic) climate change, but natural variability limits the accuracy of these projections,” says Deser. “In other words, natural variability is the ‘X Factor’ that determines how future weather and climate events will manifest in specific locations or regions, for example Denver or the western United States.”

Regional and global model projections provide researchers and decision makers with best guesses of climate dynamics and what might be expected in the future, however, local and regional climate variations from one model projection to the next is largely a consequence of the chaotic nature of large-scale atmospheric circulation patterns. Given this reality, uncertainty around projections of regional climate will remain even as models improve or as greenhouse-gas trajectories become more certain.

Consequently, regional leaders will have to rely heavily on knowledge of the local environment, infrastructure capacity, and societal needs to best manage change in an uncertain climate future. For instance, within the next five decades, a region might win the climate change lottery with residents and ecosystems experiencing average winter temperatures that are slightly cooler than today’s. Alternatively, the effects of natural variability may leave residents of the region facing 4-degree warming in the next 50 years, or the result may be somewhere in the middle.

Given the breadth of climate possibilities, Deser and her colleagues explored these questions in a 2013 paper in Nature Climate Change. The scientists looked at a range of future-climate possibilities for North America, as well as specific cities (Seattle, Washington, Phoenix, Arizona, and Mazatlan, Mexico), factoring in natural variability. To test the variety of possible climate outcomes, Deser and her colleagues ran a global climate model 40 times, with each run beginning in 2000 and ending in 2060. For every run, only the atmospheric starting conditions changed (for example, air temperature or precipitation), while all other components – ocean, land, and sea ice characteristics, etc. – remained the same. The first time such a large set of climate change experiments has been run on a single state-of-the-art global climate model, the range of outcomes indicate the spectrum of possible future climate trends.

Next, to estimate future regional climate and climate change, the scientists took observed temperature and precipitation records from 1910 – 2010 for the three cities and the United States as a whole, extrapolating this information forward to 2060 using model projections. These projections took into account both the most and least extreme future climate trends for the locations considered.

Runs from the global climate model, Community Climate System Model version 3 (CCSM3), represented the spatial patterns and magnitudes of global and continental climate variability well on decadal and longer time scales, lending confidence to the model’s projections, says Deser. The scientists found that regional climate outcomes varied widely for both precipitation and temperature, with some locations projected to experience either drying or moistening, and either warming or cooling, depending on the particular coin-toss of the natural variability.  

While it is well established that models – both global and regional – frequently disagree on projected climate outcomes, based on their research, Deser and her colleagues suggest that to a large degree this occurs because the effects of natural variability that ride on top of anthropogenic climate change are present within both the natural and modeled systems. Because of this, even when averaging a series of modeled regional outcomes, effects of natural variability within the chaotic system may not allow the narrowing of possible future regional climate outcomes in the next 50 years.

These findings make clear that a range of future climate outcomes should be expected, particularly on regional and local scales for the next 50 years, rather than a single trajectory determined solely by human influences, says Deser.

“The models aren’t wrong – they provide insights on what to expect in the future and help untangle anthropogenic climate change from natural variability. But because of natural variability, local and regional decision makers will need to consider observed climate trends along with modeled climate projections to come up with common-sense options for managing effects of climate change,” she says.

Next up, Deser and her colleagues are running another large set of climate-model simulations with the newest version of the model, the Community Earth System Model version 1 (CESM1); these runs begin in 1920 and end in 2080. They hope that comparing observations with CESM model runs over the historical period will facilitate identification of natural and anthropogenic components to past climate trends, as well as those projected for the coming decades. Achieving this may help researchers, decision makers, and the public better understand nature’s patterns and could also assist in explaining regions of cooling in the face of increasing global average temperatures.