Ultrasonic snow depth sensors for measuring snow in the U.S

The measurement of snowfall and its water content has yet to be perfected and continues to be a constant challenge to the National Weather Service and others involved in climate monitoring. Since the late 1980s, a technology utilizing ultrasonic sound waves that reflect off the snow surface to estimate changes in snow depth has been tested. Depth sensors have been manufactured and sold for several years. The U.S. Department of Agriculture Natural Resources Conservation Service has successfully deployed hundreds of these depth sensors over the past few years to measure and report snow depth from remote, mountainous areas in the western U.S. A similar instrument has also been deployed recently at many weather stations in Canada. These sensors provide very useful information on snow depth at remote weather stations especially in areas with deep and continuous snow cover. However, up until now, these depth sensors have not been tested or used extensively to measure snow depth at any of the National Weather Service's U.S. weather stations.

In 2003 the National Weather Service funded a study to evaluate snow depth sensor technology for possible adaptation at ASOS and cooperative weather stations. A major collaborative intercomparison study was undertaken during the winter of 2004-2005 to quantitatively compare the results of ultrasonic depth sensors to traditional manual observations of both snowfall and snow depth. This paper presents results of this study from 14 sites across the coterminous U.S. spanning various climate types. Both deep, continuous and shallow, intermittent snow covered areas were monitored. In order to compare manual snowfall measurements to the sensors an algorithm was developed to estimate snowfall from the total depth of snow on the ground measured by the sensors. The estimation of snowfall from snow depth is important for the integrity and continuity of snowfall records across the country. Manual measurements of snowfall were taken at each site in 6 and 24 hour intervals for comparison to estimated sensor snowfall. The factors effecting sensor performance were also investigated and will be presented. While there are some limitations, considerable progress has been made and this technology is showing good promise as an objective tool for measuring snow accumulation where no human observers are available.