An algorithm to estimate traditional snowfall measurements from ultrasonic snow depth sensors at U.S. observing sites

Since automation began in the 1990's, many U.S. observing sites ceased taking snowfall measurements because there was sensor that could perform the measurement in a way that would be climatologically homogeneous to what a human observer would report. From 2006-2008, a study was conducted that tested the Campbell Scientific SR-50® ultrasonic snow depth sensor in a triplicate configuration at 16 sites across the U.S. This study found the sensors to be quite useful when properly sited, but sensor noise was found to be a problem with the snowfall derivation algorithm. It was thought that using concurrent measurements from an automated precipitation gage would reduce false snowfall reports and improve the algorithm. During the winter of 2011-2012, three sites in the U.S. (Fort Collins, CO; Buffalo, NY and Aberdeen, SD) participated in a continuation study which added the U.S. Climate Reference Network Geonor gage setup with a double alter shield and wetness sensor to the triplicate configuration of the snow depth sensors. The algorithm and test results will be discussed.