|Title||Observed changes in Upper Rio Grande snowpack water storage and snowmelt timing: 1980 to 2018|
|Publication Type||Conference Paper|
|Year of Publication||2019|
|Authors||Elias E, James D.K., Heimel S|
|Conference Name||Universities Council on Water Resources|
|Conference Location||Snowbird, UT|
|ARIS Log Number||364395|
One of the most widely anticipated impacts of climate change on the hydrological cycle of western watersheds is earlier snow melt. Observational studies and hydrological modeling efforts predict diminished snowpack and earlier peak streamflow in watersheds across the western United States. Here we use snow course data of the snowpack telemetry (SNOTEL) network from 16 stations in the Upper Rio Grande watershed with the longest data record (~40-years). Trend analyses with various methodologies (parametric, mann-kendall and regional kendall) and time-windows (1980 to 2018 and 1984 to 2018) augment previous modeling studies. We anticipate that since the observable impacts of climate change are occurring in many sectors, observed data may show a decreasing trend in maximum water equivaled of the snowpack (SWE), earlier snowmelt, shifts in the snowmelt window and a more rapid snowmelt (fewer days between peak snowpack and no snowpack). Trend analyses between 1980 and 2018 show the maximum annual water content in snowpack is decreasing, the snowpack is melting earlier and the date of the maximum SWE is earlier in the year. However, the number of days between the peak snowpack and no snowpack did not change – indicating that the snowpack was not melting faster. The early 1980’s was an anomalous wet time, so we tested trends from 1984 to 2018 and found trends similar to those reported using linear regression. Maximum SWE declined in 11 of 16 watersheds, maximum SWE was reached earlier in 10 of 16 watersheds and the snowmelt window was earlier in 10 of 16 watersheds. Trends for the 1980 to 2018 and the 1984 to 2018 timeframes were similar, despite controlling for an anomalously wet early 1980’s. The regional kendall test reflects these trends with significantly decreasing maximum annual SWE, earlier arrival of maximum SWE and an earlier snowmelt window. Over the 40 years of measured snowpack data results indicate that snowmelt is 16 days to 5 weeks earlier, similar to modeling studies comparing a 40 to 70-year timeframe (1999 to midcentury).