Measurement of Snowfall by Radar

Paul E. Carlson Dept. of Meteorology, McGill University, Montreal, Quebec, Canada

Search for other papers by Paul E. Carlson in
Current site
Google Scholar
PubMed
Close
and
J. S. Marshall Dept. of Meteorology, McGill University, Montreal, Quebec, Canada

Search for other papers by J. S. Marshall in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Using a CPS-9 radar (wavelength 3.2 cm, beamwidth 1°) in dry snow, radar returns from a layer at an average height of 5000 ft were converted to snowfall rates (taking ZR2.0) and summed over 36 hr to obtain a map of snowfall amount. This has been compared with a “climat” map based on depth measurements of new fallen snow at 140 climatological stations within 100 mi of the radar, said amounts ranging from under 2 to over 10 inches. For the 5550 mi2 area within 42 mi of the radar, the average amount by radar was set equal to the average climat amount. The radar/climat ratio was mapped, with the distribution being log normal. For ranges <42 mi, 68% of the data fell between values of 0.76 and 1.32. For ranges between 42 and 100 mi, 68% fell between 0.63 and 1.60. This scatter is about the same as other workers have found for rain. In the relation Z = aRb, a value for b of 2.0 proved appropriate to this particular storm, with some evidence that a slightly higher value might have been a little better.

Abstract

Using a CPS-9 radar (wavelength 3.2 cm, beamwidth 1°) in dry snow, radar returns from a layer at an average height of 5000 ft were converted to snowfall rates (taking ZR2.0) and summed over 36 hr to obtain a map of snowfall amount. This has been compared with a “climat” map based on depth measurements of new fallen snow at 140 climatological stations within 100 mi of the radar, said amounts ranging from under 2 to over 10 inches. For the 5550 mi2 area within 42 mi of the radar, the average amount by radar was set equal to the average climat amount. The radar/climat ratio was mapped, with the distribution being log normal. For ranges <42 mi, 68% of the data fell between values of 0.76 and 1.32. For ranges between 42 and 100 mi, 68% fell between 0.63 and 1.60. This scatter is about the same as other workers have found for rain. In the relation Z = aRb, a value for b of 2.0 proved appropriate to this particular storm, with some evidence that a slightly higher value might have been a little better.

Save