Improving estimates of snowpack water equivalent using double sampling

TitleImproving estimates of snowpack water equivalent using double sampling
Publication TypeConference Proceedings
Year of Conference1993
AuthorsRovansek, R. J., Kane D. L., and Hinzman L. D.
Conference Name61st Annual Western Snow Conference
Series TitleProceedings of the 61st Annual Western Snow Conference
Date PublishedJune 1993
PublisherWestern Snow Conference
Conference LocationQuebec City, Quebec
KeywordsDouble sampling, Sampling, Snow surveys, SWE
Abstract

Equations are presented to estimate the average water equivalent of a snowpack and the variance of this estimate using double sampling. Double sampling refers to sampling the snowpack in two ways: measuring the depth at a number of points, and measuring the depth and the water equivalent at a smaller number of points. In a given sampling interval, doubling sampling can yield an estimate of snowpack water equivalent having a lower variance than is possible by measuring water equivalent only. The equations presented in this paper were tested during extensive snow surveys conducted on the North Slope of Alaska. The snowpack of the North Slope is similar to that of the Great Plains, thin and variable in both depth and density due to extensive redistribution of snow by wind. This variability makes estimation of an area difficult, and provides a rigorous test of the equations presented here.Two two-hour snow surveys conducted by sampling water equivalent only, and be sampling both depth and water equivalent yielded identical estimates of water equivalent (48.1 mm). However, the double sampling estimate had a variance of only 80% of the variance of the estimate using water equivalent only (4.9 mm² vs. 6.1 mm²). Equations are also presented for calculating the optimal ratio of depth to water equivalent measurements. By considering time as the only cost of collecting the two measurements it is possible to calculate the ratio of depth to water equivalent measurements that will result in the best estimate of average snowpack water equivalent in a given sampling time. The optimal ratio of depths to water equivalent measurements varies between individual surveys, so times measured during snow surveys in a number of different locations and snow conditions on the North Slope were used to estimate a regional optimal ratio of depth to water equivalent measurements of about 14 to 1.

URLsites/westernsnowconference.org/PDFs/1993Rovansek.pdf