Melting winter snowpack signals a growing threat to water security not captured by operational metrics

Abstract:

Across western North America, snow water resources are closely monitored to support seasonal water supply forecasts and long-term change assessments that commonly use the date of April 1^st to separate accumulation and melt seasons. As the climate warms, this seasonal distinction is expected to become increasingly blurred with unknown implications on forecast skill and climate change assessments. We present an analysis of daily snow water equivalent (SWE) measured at 1,097 snowpack monitoring stations in western North America. On average, maximum SWE occurred on April 1^st with 87% of seasonal snowfall yet to melt, supporting the metric’s operational utility. April 1^st SWE has declined at 12% of stations with long records. By comparison, melt occurring before April 1^st has increased at 42% of stations. Widely reported declines in April 1^st SWE are more strongly correlated with trends toward increased precipitation, dampening an underlying warming signal to which melt is more sensitive. The widespread melt increase despite lesser change in commonly used snowpack metrics indicates that this critical water resource is in steeper decline than is inferred from SWE trends alone. We evaluated 37 years (1983-2019) of Natural Resources Conservation Service (NRCS) statistical water supply outlook (April 1 – July 31) accuracy for relationships with snowpack metrics for 25 unimpaired headwater basins in the western U.S. We show that forecast skill in critical headwater basins is significantly lower in years when proportionately more melt has occurred before April 1^st. The results highlight climate-induced hydrologic change with operational implications across North America that is not being monitored with the commonly used April 1^st SWE metric.