The Spatial Footprint and Frequency of Historic Snow Droughts in Yellowstone
by Bethany L. Coulthard, Gregory T. Pederson, & Kevin J. Anchukaitis
In the face of climate change and increasing human pressures, monitoring and characterizing environmental change is increasingly important in national parks and protected areas (Hansen and Phillips 2018). Regional measurements of snowpack are a critical vital sign (see “Vital Signs Monitoring is Good Medicine for Parks,” this issue) both for monitoring ecosystem health and anticipating future water availability. Snowpack represents accumulated cool-season water storage that, if kept cool, is slowly released as summer meltwater which maintains ecosystems and replenishes reservoirs that sustain society. Western snowpacks have been shrinking since around 1950 due to warmer temperatures resulting from human-caused climate change (Fyfe et al. 2017). Future climate projections indicate this trend will continue throughout the 21st century (Mankin and Diffenbaugh 2015). In Yellowstone National Park (YNP), deep snowpacks and subsequent meltwaters contribute substantially to the flow of the Missouri, Colorado, and Columbia rivers, making a “snow drought” or period of abnormally low snowpack in this region a serious threat to ecosystems and water supplies for much of the West (Tercek et al. 2015).
Multi-year Snow Droughts: How Often and How Long Can They Last?
Two important questions for water managers, ecosystems, and communities are: “how many years can snow droughts last?” and “how often do multi-year snow droughts happen?” A good example of why this is important is the recent multi-year drought (2011-2016) in California that resulted in the over-extraction of groundwater, agricultural irrigation restrictions, and hydroelectricity cutbacks. Ultimately, the drought culminated in an unprecedented statewide water use restriction (Mann and Gleick 2015). A single snow drought year can often be managed through using water stored in major reservoirs, but multi-year snow droughts become increasingly challenging as banked water supplies are depleted (Mote et al. 2018). In this article, we address the question: how long have snow droughts historically lasted in Yellowstone?
Multi-decadal snow measurements like those recorded across the United States by the Natural Resources Conservation Service (NRCS) Snow Telemetry (SNOTEL) and snow course networks are the main resource for understanding changing snow dynamics. However, these relatively short observational records typically around 30-40 years for SNOTEL records or 60 to nearly 100 years for the longest snow course records only capture a few multi-year snow drought events. Our understanding of the natural frequency and severity of snow droughts from such short records is therefore limited. This, in turn, restricts our ability to contextualize recent and likely future snow drought events.
Tree Rings Measure Snow Droughts Over Space and Time
One approach to answering questions about past snow droughts is by using snow-sensitive tree-ring records, which are increasingly being capitalized upon to reconstruct historical snow dynamics (Belmecheri et al. 2016, Pederson et al. 2011, Woodhouse 2003). Snow sensitive trees in the U.S. West tend to exhibit two types of growth responses, those whose growth benefits from snow meltwater (typically growing in low to mid-elevations) and those whose growth is inhibited by deep and late-lying spring snowpack (typically growing at high elevations). These growth responses are recorded at the stand (or population) level from numerous species and can provide annual historical records, or “reconstructions”, of year-to-year snow variability going back centuries and even millennia. Tree-ring based snowpack reconstructions have already provided important insights about past snow dynamics in YNP regarding the influence of modern temperature increases on snowpack declines, and the recent widespread and synchronized spring snowpack declines since the 1980s along the Rockies (Pederson et al. 2011). Here we present a set of new snowpack reconstructions for YNP that offer enhanced spatial resolution (4 km2 grid versus large watersheds) and have been extended in length (beginning in AD 700 compared to AD 1200) relative to the original Pederson et al. (2011) reconstructions. We use these new spatially gridded snowpack reconstructions to identify and map YNP’s past severe multi-year snow drought events.
The new reconstructions are unique in that they are spatially gridded, which allows us to examine snow drought events over both space and time. Using a gridded spatial field reconstruction approach (Point-by-Point Regression; see Cook et al. 2004), each individual gridpoint is reconstructed and represents its own multi century- to multi millennia-long snow moisture (snow moisture is expressed as snow water equivalent or SWE) reconstruction describing local, yearly snow levels and spanning the age of available tree-ring records. We used a 2.5 mi x 2.5 mi (4 km x 4 km) gridded modeled snow moisture dataset (see Hostetler and Alder 2016) and a large network of more than 500 curated, snow-sensitive tree-ring records (i.e., chronologies) in our analysis. Snow moisture content or SWE data as measured on April 1st of each year was used since it historically serves as a reliable estimate of maximum cool-season snow water content, which is directly relevant to forecasting spring runoff and warm-season water supply. The final reconstruction model was employed over the length of available tree-ring records to provide annual estimates of local, gridpoint snow moisture (i.e., SWE) back through time. Further methodological details can be found in Cook et al. (2004).
A 1300-year Gridded Snowpack Record for Yellowstone
The top panel in figure 1 shows a plot of reconstructed mean snow moisture (in mm of SWE) for YNP (the average of all gridded reconstruction means from YNP). Similar to a reconstruction by Pederson et al. (2011) developed specifically for the Greater Yellowstone Region (GYR), this record highlights unusually low snow moisture (SWE) during the early 1500s, during the 1930s dustbowl, and at the turn of the century drought in the 2000s. This longer record, however, also documents unusually low snow moisture conditions in YNP during the mid-700s and in AD 1014.
Notable Major Snow Drought Events
The bottom panel in figure 1 shows maps of gridded average reconstructed snow moisture during known historical snow droughts, demonstrating the utility of the gridded tree-ring records to provide spatially relevant information on snow moisture conditions of the past. The ability of the reconstructions to faithfully record known historical events like the 1930s dustbowl drought is a useful check on the model accuracy, while the spatial fingerprint of pre-historic snow droughts in the 700s and early 1500s can also be examined. Notably, the early 1500s were wetter in YNP than across the GYR, and SWE deficits during the 700s appear more severe than other major pre- and post-industrial snow droughts in YNP and the across the region.
The Frequency of Multi-year Snow Droughts
To identify multi-year snow droughts in YNP we conducted an analysis (see González and Valdés 2003) on the reconstructed April 1 snow moisture data. We defined a snow drought year as a reconstructed SWE value below the 25th percentile or value below the lowest 25 percent of observations for the entire reconstruction. We also calculated how many snow droughts occurred at each grid point, testing all possible snow drought-lengths from one year long to twelve years long. For snow droughts greater than three years long, a single year of high snow moisture (i.e., SWE ≥25th percentile) was not permitted to interrupt a multi-year snow drought. This was done because a single year of above average snow moisture is often not sufficient to replenish stored water at major reservoirs during a multi-year snow drought from a water resource management perspective.
The maps in figure 2 show the number of 3- to 12-year long snow droughts in YNP over the past 1,300 years. Almost the entire region has experienced more than 200 three-year snow droughts in this time. In other words, during a total of ~190 years of the 1,300-year reconstructed record, YNP was in a three-year snow drought. Longer snow droughts are also more common in YNP than has been previously recognized. Beyond known decadal-scale droughts like the dustbowl, tree-ring records suggest most of YNP has experienced between one and forty 12-year long snow droughts since AD 700. Considering all of the snow droughts identified in this analysis, 37% lasted between 3 and 6 years, and 16% lasted between 7 and 12 years. Note that to provide a complete picture of both short and long snow droughts that occur in YNP, shorter droughts contained within longer droughts were counted as events in this analysis. These findings indicate long and persistent snow droughts were a natural part of the snowpack system in YNP during the pre-industrial era. Such droughts could have serious consequences for water supply and natural resource management in YNP, especially when exacerbated by the host of other water-related and ecological changes that are likely with continued warming.
Belmecheri, S., F. Babst, E.R. Wahl, D.W. Stahle, and V. Trouet. 2016. Multi-century evaluation of Sierra Nevada snowpack. Nature Climate Change 6(1):2.
Cook, E.R., and P.J. Krusic. 2004. The North American drought atlas. Lamont-Doherty Earth Observatory and the National Science Foundation. http://iridl.ldeo.columbia.edu/SOURCES/.LDEO/.TRL/.NADA2004/.pdsi-atlas.html
Fyfe, J.C., C. Derksen, L. Mudryk, G.M, Flato, B.D. Santer, N.C. Swart, N.P. Molotch, X. Zhang, H. Wan, V.K. Arora, J. Scinocca, and Y. Jiao. 2017. Large near-term projected snowpack loss over the western United States. Nature Communications 8:14996.
González, J., and J.B. Valdés. 2003. Bivariate drought recurrence analysis using tree ring reconstructions. Journal of Hydrologic Engineering 8:247-258.
Hansen, A. J., and L. Phillips. 2018. Trends in vital signs for Greater Yellowstone: application of a WildlandHealth Index. Ecosphere 9(8):e02380.
Hostetler, S.W., and J.R. Alder. 2016. Implementation and evaluation of a monthly water balance model over the US on an 800 m grid. Water Resources Research 52:9600-9620.
Mankin, J.S., and N.S. Diffenbaugh. 2015. Influence of temperature and precipitation variability on near-term snow trends. Climate Dynamics 45(3-4):1099-1116.
Mann, M.E., and P.H. Gleick. 2015. Climate change and California drought in the 21st century. Proceedings of the National Academy of Sciences 112:3858-3859.
Mote, P.W., S. Li, D.P. Lettenmaier, M. Xiao, and R. Engel. 2018. Dramatic declines in snowpack in the western US. npj Climate and Atmospheric Science 1:2.
Pederson, G.T., S.T. Gray, C.A. Woodhouse, J.L. Betancourt, D.B, Fagre, J.S. Littell, E. Watson, B.H. Luckman, and L.J. Graumlich. 2011. The unusual nature of recent snowpack declines in the North American Cordillera. Science 333:332-335.
Tercek, M., A. Rodman, and D. Thoma. 2015. Trends in Yellowstone snowpack. Yellowstone Science 23:20-27.
Woodhouse, C.A. 2003. A 431-yr reconstruction of western Colorado snowpack from tree rings. Journal of Climate 16:1551-1561.
Bethany Coulthard is an Assistant Professor of Geoscience at the University of Nevada Las Vegas. As a hydroclimate and paleo scientist, she explores how climate change influences global water cycles, human populations, and ecosystems. She specializes in developing and analyzing tree-ring records of drought, flood, snow, runoff, and ice melt—and the climate patterns that drive these processes—especially in mountains. She is dedicated to providing practical datasets and information that can be used by resource managers, policy makers, and communities confronting climate change.
Series: Yellowstone Science - Volume 27 Issue 1: Vital Signs - Monitoring Yellowstone's Ecosystem Health
Last updated: September 16, 2019