Snow burial protects plants and insects from extreme winter cold and we expect that in may cases, beetle survival is better in years with longer snow cover. It may be that the positive relationship between annual precipitation and beetle range expansion is caused by increased snow cover in the winter.
Using CUP logger data from a site to determine how long the site is buried in snow each year.
The days that the CUP loggers are buried in snow is usually easy to determine, as illustrated in the figure above. It may be important that the CUP loggers are not at ground level but rather are 1-3' above ground on a willow branch. To examine this effect, we put out a series of BASE loggers at ground level. In spring 2010 we will compare to see if the CUP logger data accurately reflects snow cover at the base of the plants. Excel file summarizing 11 years of snow cover data
Graph showing results of regression analysis in which the days of snow cover at each site are plotted as a function of elevation, for each of 12 years. The relationship may be broken down into two components: (1) the year may be wet or dry, causing the snow cover values to be higher or lower on the graph, and (2) the common slope indicates that for the average year, the cover increases by 13 days for each 100m of elevation (slope = 0.128 +/- 0.016 days per meter).
By taking the Y-intercept of the regression lines in the graph above, we record a snow cover index for any given year. For example, 2006 was a year of extensive long lasting snow cover, while 2007 was a severe drought year with snow cover only at the highest elevations. 2009 was unusual in that the snow melted early, but June was cool and wet as if there was a later snowmelt.
The graph below plots 12 years of snow cover as a function of regional precipitation during that year (Owens River water year, % of average value).
Beginning in 2009 we began collecting unshielded temperatures from the base of the willows (BASE logger data). This could potentially give a somewhat clearer picture of the timing of snowfall and snowmelt, because the sensor is on the ground rather than suspended off ground from a willow branch. Below is a graph comparing the timing of spring snow melt between the the two types of logger, CUP and BASE.
While the two measurements are very comparable, the BASE sensors yield a elevation vs melt-date relationship with reduced variance. Note also that the value of 60 days per km elevation is about half that estimated for the duration of snow cover shown in the graphs above. This is to be expected since the duration of snow cover is compounded from the date of fall snow burial and the date of spring snow melt, and sites with late snow melt are likely to have early snow burial. When the duration of snow cover is compared between CUP and BASE loggers they are also very comparable. The timing of snow burial in the fall is harder to calculate for the BASE data because the base is repeatedly buried and exposed before the "big" snow fall arrives.
For the early years we have not recorded the timing of burial and melt, only the duration of cover. This could be easily done since the raw data files are available.