Measurement of Snowfall by Optical Attenuation

View More View Less
  • 1 Department of Meteorology, McGill University, Montreal, Canada
© Get Permissions
Full access

Abstract

A transmissometer has been used to provide a continuous record with good time resolution of falling snow. The pulsed light, of wavelength 0.45μ, traversed a path 71 m long about 20 m above ground level A total snow amount of 160 millimeters of water (mmw) from 20 storms through the 1966-67 winter season was recorded and analyzed, Attenuation by snow was found to be proportional to rate of snowfall, with the constant of proportionality 11 (db km-1) / (mmw hr-1). A previous experiment by Lillesaeter yielded 18 for this constant.

The attenuation from precipitation-free air was found to increase as the relative humidity increased above about 60%. Relative humidity increase between the beginning and end of a storm could lead to an increase in "clean air" attenuation of 4-5 db km-1. Lack of correction for this together with the effect of thermal fluctuations on Lillesaeter's narrower transmitted beam probably account for his higher value of the constant of proportionality.

Snow amounts for individual storms deduced from the attenuation record agreed with amounts measured by standard instruments to within a factor of 2. When depths on the ground were compared, agreement was within a factor of 1.5. Over the 20 principal storms of the season, the total snow amount from the attenuation records agreed to within 2% with the accumulation in a standard Nipher gage.

Abstract

A transmissometer has been used to provide a continuous record with good time resolution of falling snow. The pulsed light, of wavelength 0.45μ, traversed a path 71 m long about 20 m above ground level A total snow amount of 160 millimeters of water (mmw) from 20 storms through the 1966-67 winter season was recorded and analyzed, Attenuation by snow was found to be proportional to rate of snowfall, with the constant of proportionality 11 (db km-1) / (mmw hr-1). A previous experiment by Lillesaeter yielded 18 for this constant.

The attenuation from precipitation-free air was found to increase as the relative humidity increased above about 60%. Relative humidity increase between the beginning and end of a storm could lead to an increase in "clean air" attenuation of 4-5 db km-1. Lack of correction for this together with the effect of thermal fluctuations on Lillesaeter's narrower transmitted beam probably account for his higher value of the constant of proportionality.

Snow amounts for individual storms deduced from the attenuation record agreed with amounts measured by standard instruments to within a factor of 2. When depths on the ground were compared, agreement was within a factor of 1.5. Over the 20 principal storms of the season, the total snow amount from the attenuation records agreed to within 2% with the accumulation in a standard Nipher gage.

Save