The snowpack decline of the past 50 years in the Rocky Mountains is highly unusual in context of the past 800 years, according to findings being published today in Science.
The researchers used tree rings to reconstruct snowpack in river basins in the West during the last several centuries and show that past variations in snowpack were clearly attributable to natural factors affecting temperature and precipitation.
This work confirms work by others indicating that between 30 percent and 60 percent of the declines are likely from warming caused by greenhouse gas emissions from human activities with the rest arising from natural climate phenomena.
“Our results indicated that it’s due both to human-caused warming and ‘natural’ long-term fluctuations. The Northern Rockies in particular have been hit by a double whammy of climate change and natural variability in temperature,” said Lisa Graumlich, dean of the University of Washington College of the Environment and a co-author of the Science paper.
The study area includes three regions: the Northern Rockies, Greater Yellowstone and Upper Colorado with headwaters respectively of the Columbia, Missouri and Colorado rivers.
The researchers considered snowpack in Canada and the United States that feeds three major river basins, the Columbia, Missouri and Colorado. Among other things, the river basins provide water to more than 70 million people.
Snowpack losses across the West since the 1980s “may signal a fundamental shift from precipitation to temperature as the dominant influence on snowpack in the North America Cordillera, with significant consequences for regional water supplies,” the co-authors wrote. The term cordillera is applied to a continent’s principal mountain ranges.
If temperature is playing a larger role in snowpack, this could be more of an impact for the northern Rockies and the headwaters of the Columbia River compared to the other regions, said Jeremy Littell, research scientist with the UW’s Climate Impacts Group and a co-author of the paper.
That is because this area is low in elevation and temperatures already hover closer to freezing in winter and spring compared to the rest of the Rockies where it is colder. This means there’s an earlier transition to precipitation falling as rain instead of snow in the future.
The snowpack decline in the Rockies from 1950 to 2000 varies by region and elevation, from a 10 percent decline in the Central Rockies to 40 percent decline in the Oregon Cascades, he said.
Averaged over the Cascades Mountain range, the snowpack decline is about 30 percent. In the future in Washington state, the UW Climate Impacts Group has estimated that April 1 snowpack may decline on the order of 40 percent by the 2040s relative to the period 1916-2006.
Littell played a key role in developing the use of trees rings, which reveal a tree’s growth each year, as a way to document snowpack in the Pacific Northwest and Northern Rockies. At the highest elevations, heavy snows in this region decrease the length of time trees can grow each year, while low snowpack years make for a longer growing season.
This study is based on 66 tree-ring chronologies and, while a few previous papers have looked at single river basins, this study is the first to consider basins across the West, he says.
Comparing their tree-ring data with actual snowpack measurements – or snow water equivalent measurements – made April 1 by water resource managers since the late 1930s, gave the researchers the means to then calculate snowpack from just the tree-ring data going back centuries before humans started making measurements.
Natural variability in snowpack in the past has included decades-long shifts where winter storms were concentrated alternately over the northern Rockies for a period of time and then over the southern Rockies for a period of time. This shifting appears to have broken down after the 1980s, partly because of human-caused warming, the co-authors say.
Lead author on the paper is Gregory Pederson, U.S. Geological Survey; other co-authors are Stephen Gray, University of Wyoming; Connie Woodhouse, University of Arizona; Julio Betancourt, USGS; Daniel Fagre, USGS; Emma Watson; and Brian Luckman, University of Western Ontario. Funding was provided by National Science Foundation, USGS, National Oceanic and Atmospheric Administration’s Joint Institute for the Study of the Atmosphere and Ocean, NOAA’s Sector Applications Research program and the Natural Sciences and Engineering Research Council of Canada.