The snowpack over the western United States has declined about 23% since 1981. It is projected to decline more in the future.
I have written a number of posts about the looming water deficit in California due to a projected decline in the snowpack on the Sierra Nevada mountains. Is something similar projected to occur throughout the entire western United States?
Yes. Studies find that the water content of the snowpack throughout the West has already declined 23%, and it is expected to decline more, perhaps up to 30% by 2038.
This decline is not occurring via a decrease in precipitation. Indeed, to date precipitation across the West has actually increased slightly. The decline is occurring due to increased temperature. Some precipitation that used to fall as snow now falls as rain, and the snow that does fall melts more quickly.
Mote and Sharp studied the snow water equivalent* of the snowpack in April from 1955-2016 at SNOWTEL measuring stations operated by the U.S. Natural Resource Conservation Service. Figure 1 shows a map of the stations, with blue dots representing stations where the snowpack increased and orange dots representing stations where the snowpack declined. The size of the dots represent the magnitude of change.
It is easy to see that the vast majority showed declines in the snowpack, in many cases by as much as 80%. Overall, Mote and Sharp computed that there had been an average 23% decline in the western snowpack since 1955.
Fyfe and his colleagues conducted climate modeling to try to determine whether the decline in the snowpack was due to natural causes or human causes. Figure 2 shows the results in a rather complicated graph. Let’s unpack it. The computer models ran from 1950 to 2010. The dashed black line shows the observed trend in the snow water content. The solid blue line shows the projected snow water content if only natural climate causes are included in the model. It doesn’t fit the observed trend very well. The solid black line shows the projected snow water content if both natural and human climate causes are included in the model. It fits the observed data quite closely. (The pink and green lines show data from analyses using other sets of data and need not concern us here. The gray band and blue dotted lines show statistical confidence levels for the computer simulations, and also need not concern us here.)
The simulation that included both natural and human causes agreed with the observed data, but the one that included only natural causes did not. The authors concluded that natural causes could not explain the loss of snowpack in the West. A combination of human and natural causes could explain it.
Fyfe and his colleagues also conducted a suite of climate models to project snowpack loss into the future. The results are shown in Figure 3. In this graph, the y-axis represents the actual snow water content of the snowpack, not the change. The blue line represents the computer model that projected the least snowpack loss in 2030, and the red line represents the computer model that projected the most loss. It is common practice among climate modelers to run a suite of projections, and when this is done, the average of them is often also presented, and it is often taken as likely to be the most accurate. In Figure 3, the average of the projections is represented by the black line.
It is easy to see that the trend in all of the lines is down. There is considerable variation from point-to-point in the red and blue lines, indicating that the projections expect there to be considerable variability in the snowpack from year-to-year. The black line is pretty smooth, however, as might be expected from an average of several analyses, and it has a consistent downward trend. The losses in snowpack in some of the projections ran as high as 60%, though average loss across the suite of projections was about 30%.
A 30% decline in the snowpack does not sound so dire; after all the projections are for a 60% loss of snowpack in California (see here). However, that projection was for the end of the century. This projection is for 2038; that’s only 20 years from now.
Some may wonder about how little snow water equivalent is shown on the y-axis of Figure 3. In the 1990s, the snowpack maxed-out each year at only 6+ cm. of snow water equivalent. In thinking about this number, remember two things: first, a centimeter of water represents somewhere between 3 and 20 centimeters of snow, with an average value being somewhere around 10 cm. Thus, 6 cm. of snow water equivalent would roughly equal 60 cm. of snow, or 23.6 inches. Thus, the average depth of the snowpack was about 2 feet. Second, remember that the measurements were averaged across hundreds of locations; some were high and received a great deal of snow, but some were relatively low (low altitude means more rain, less snow), or were located in areas that don’t receive much precipitation of any kind.
Much of Missouri depends on the Missouri River as a water supply, including both Kansas City and St. Louis. The Missouri River gets much of its water from the western snowpack. A declining snowpack may, or may not, have implications for our water supply, depending on whether the reservoirs along the Missouri River can accommodate the shift toward earlier snowmelt and increased rain. I will look at this issue in the next post.
* Snow water equivalent: Different types of snow hold different amounts of water. Thus, scientists don’t just measure how deep the snow is. Rather, at a given location they take a representative sample of the snowpack and melt it, thereby determining how much water it holds. This is the snowpack’s snow water equivalent at that given location. April is generally when the snowpack is at its maximum.
Environmental Protection Agency. 2016. Climate Change Indicators in the United States: Snowpack. Retrieved online 5/22/2017 at https://www.epa.gov/sites/production/files/2016-08/documents/print_snowpack-2016.pdf.
Fyfe, John, Chris Kerksen, Lawrence Mudryk, Gregory Flato, Benjamin Santer, Neil Swart, Noah Molotch, Xuebin Zhang, Hui Wan, Vivek Arora, John Scinocca, and Yanjun Jiao. 2017. “Large Near-Term Projected Snowpack Loss Over the Western United States.” Nature Communications, DOI: 10.1038/ncomms14996. Retrieved online 5/14/2017 at https://www.nature.com/articles/ncomms14996.