Everyone is talking about snowpack this year. The news boasts epic-sounding statistics for snowpack that is “649 percent of normal,” “128 percent of average,” “440 percent of median,” or, my personal favorite, “1,776 percent of last year.” One need not speak math to know that big snowpack is generally a good thing. But unpacking the statistical lingo can help in understanding what those numbers actually indicate.
People generally understand snowpack to refer to snow depth. But in the context of water supply, snow depth doesn’t matter as much as snow-water equivalent. SWE is the amount of liquid water released when snow is melted instantaneously. SWE data mainly comes from automated sites across Colorado that capture, record and report daily data.
SWE is a more useful measurement for understanding water supply than snow depth, because water content in snow layers can vary — consider the difference between two inches of wet, heavy snow and dry powder. And when discussing snowpack, what most people care about is not cubic inches of snow in the mountains; it’s how much water will flow through our rivers, ditches and reservoirs.
This year, peak SWE for the Upper Colorado Basin was 20 inches. To add context, last year, which was dry, the SWE was 16 inches, while in the 2011 flood year it was 25 inches.
Snowpack varies throughout the year. Imagine a mountain profile where the left side is an easy, long slope that steps up to a series of peaks, and the right side is a steep, jagged descent. This is what the actual SWE data looks like plotted for each day as snowpack starts to build in October, slowly accumulates through the winter, peaks in April, then starts to melt (with some intermittent, spring-storm gains) though June or July when it’s fully depleted.
When most people bandy snowpack statistics around, they are referring to a point on that graph and “average” is probably the most confused term that gets used. We typically understand “average” to mean the arithmetic “mean,” which is the sum of numbers in a set divided by the number of numbers used in the set. (For example, the mean of 4, 6 and 20 is 10). But when discussing snowpack, the “average” typically refers to the median, which is the number that falls in the middle of the set. (So, the median of 4, 6 and 20 is 6.) In other words, there are different ways of describing what is “normal.”
Snowpack averages most often compare a relevant date of the current year to the median value on that same date within the study period (for western snowpack, the dataset is from 1981 to 2010). The median is more appropriate because it is not affected by outlier years.
You can see on an SWE graph that the median snowpack is less than the mean snowpack. This means there have been more low-snow years than high-snow years, but the high-snow years were really snowy, thus skewing the mean higher and the median lower.
Anyway, math schmath. … What most people really care about isn’t the difference between median and mean. People care about how much water there is. We want to know when the rivers will really start to flow, whether they’ll flood, and when runoff will peter out. And the relationship between snowpack and runoff has almost as much to do with the timing and intensity of spring and summer weather as the timing of peak SWE.
For example, because of spring-weather timing, runoff this year is late and expected to be sustained for longer. Peak snowpack also occurred later, and temperatures are staying cooler, so runoff is expected to peak in late June. Compare that to 2011, another heavy snow year, where warmer weather and rain caused runoff to peak in May, with more flooding.
This year’s snowpack-generated runoff is good news for water users now, but how good is it for the future? It will recharge the soil and aquifers, refill reservoirs, make for long recreation and irrigation seasons, and help mitigate Colorado’s long-term drought.
However, this region has been in a 16-year drought and needs several more years of increased snowpack and cool temperatures to alleviate the pressures on regional water supplies. According to Colorado River District engineer John Currier, we will need at least seven consecutive 2019s to fully fill lakes Powell and Mead, which are currently at historic lows.
Understanding snowpack, in sum, has more than some depth to it.
Erika Gibson is a contributor to this monthly column for the Middle Colorado Watershed Council, which works to evaluate, protect and enhance the health of the middle Colorado River watershed through the cooperative effort of watershed stakeholders: anyone standing in the watershed. To learn more about the MCWC, visit https://www.midcowatershed.org. You can also find the Council on Facebook at https://www.facebook.com/midcowatershed.
