Changes in the Timing of Snowmelt and Streamflow in Colorado: A Response to Recent Warming

David W. Clow from the U.S Geological Survey out of Lakewood, Colorado investigated how SWE (snow water equivalent) and streamflow, are correlated to the rate and timing of snowmelt in Colorado. Monthly air temperatures, snowfall, latitude, and elevation were also used in multiple linear regression models to determine the controls on snowmelt. Interval/ratio data for SWE was collected from Natural Resource Conservation Service (NRCS) snowpack telemetry sites (SNOTEL sites) and the streamflow data was collected from headwater streams. Data from 1978-2007 was collected from the SNOTEL sites with 97% of sites had ≥21 years of data and all sites <18 years being excluded. Daily streamflow data were obtained for 58 headwater streams in Colorado with long-term gauges operated by the U.S. Geological Survey (USGS) or the Colorado Division of Water Resources. The regional Kendall test (RKT) was used to determine changes in trends to air temperature and SWE over the 27 year period. Multiple linear regressions were used to determine how and to what degree the changes impacted the timing of snowmelt.

A benefit of using the RKT is that by grouping data into geographic regions trend detection is increased. The multiple linear regressions were beneficial in determining which variables impacted the timing and degree of melt the greatest. Increasing springtime air temperature and declining SWE explained most, 45%, of the interannual variability in snowmelt timing. Regression coefficients for air temperature were negative, indicating that warm temperatures promote early melt. Regression coefficients for SWE, latitude, and elevation were positive, indicating that abundant snowfall tends to delay snowmelt, and snowmelt tends to occur later at northern latitudes and high elevations (Clow, 2010). The use of these methods demonstrates the strength of more traditional data analysis techniques and how they can be applied to data sets.

The conclusions found in this article will influence my research proposal by demonstrating which variables are most influential in generating melt in a snowpack. Similarly, this research demonstrates how to use variables similar to the variables I will be using a multiple linear regression. The article also validates a portion of my research topic by stating, “It may be useful to include other possible controls on snowmelt timing, such as dust deposition, in regression models in the future.” (Clow, 2010) This article also makes predictions about how Nov-May air temperatures increased by a median of 0.9°C decade−1, while 1 April SWE declined by a median of 4.1 decade−1 and maximum SWE declined 3.6 cm decade−1. This could be utilized in my proposal by investigating if these trends have stayed consistent over the past decade since the research was conducted. Another aspect of this research that I am possibly going to include in my proposal is grouping data by region because certain areas might be influenced by variables in different ways. Using this technique I could determine, for example, if higher temperatures in southwest Colorado and more influential than SWE.  

Clow, D. W. (2010). Changes in the Timing of Snowmelt and Streamflow in Colorado: A Response to Recent Warming. Journal of Climate,23(9), 2293-2306. doi:10.1175/2009jcli2951.1