Derived Snowfall Accumulation Using Optical Sensors

Measuring snow efficiently and accurately has proven to be difficult for modern day automated weighing precipitation gauges. Due to the slow fall velocities of snow particles, winds make it problematic for snow to fall vertically into these automated precipitation gauges. This causes the gauges to underestimate the true amount of snow that has fallen, with underestimates of 20-50% being common (Rasmussen, 2010). An alternative to using a weighing gauge, a disdrometer is an instrument that is capable of measuring precipitation rates, and therefore accumulation. In 2009, Sugiura was the first to use a disdrometer to measure solid precipitation (Zhang, 2015), and a number of new optical sensors have been introduced since that time.

Using optical sensors to measure solid precipitation could potentially lead to a new and easier way to measure snowfall. Since optical sensors use a light beam for precipitation monitoring rather than physically weighing how much precipitation has fallen into the instrument, like a standard automated weighing precipitation gauge, wind should no longer be affecting the snowfall amounts being reported by the instrument. This would mean that it would no longer require setting up a double fence intercomparison reference (DFIR) with a weighing gauge inside, a time consuming process, in order to get accurate snowfall measurements. The DFIR is needed to surround weighing gauges because without one, wind speeds are too high for snow to fall vertically into the gauge, and therefore the gauge is undercatching the true amount of snow that has fallen (Rasmussen, 2010). In order to see if new disdrometers are capable of measuring snow accumulations accurately, I will be examining the OTT laser-optical Particle Size Velocity 2 (PARSIVEL) and the Thies Laser Precipitation Monitor (LPM).

The OTT PARSIVEL uses a laser operating in the 780 nm wavelength, with output power of 0.5 mW (OTT, 2015). The dimension of the beam size that is uses is 180 mm long and 30 mm wide, resulting in a 54 cm2 measurement surface (OTT, 2015). It is capable of measuring solid precipitation in the range of 0.2 – 25mm with speeds ranging from 0.2 – 20 m/s and intensities ranging from 0.001 – 1200 mm/hr (OTT, 2015). As precipitation falls through the laser field, a portion of the laser beam is blocked to the receiver which then registers a drop in output voltage corresponding to the diameter of the particle size (OTT, 2015), and depending on how long this drop in voltage is registered for, particle speed is also calculated. The particle size and speed is then put into one of 32 separate size bins, which the OTT algorithm then uses to classify what type of precipitation is falling through the laser field (OTT, 2015). The rate of precipitation is also calculated during this time (OTT, 2015). After one minute, the instrument then outputs this information to the data logger (OTT, 2015).

The Thies LPM uses a laser-optical beaming source operating in the 785 nm wavelength, with output power of 0.5 mW (Thies, 2015). The beam dimensions are 228 mm long and 20 mm wide, which results in a 45.6 cm2 measurement surface with a 0.75 mm depth (Thies, 2015). Solid precipitation is measured from the range of 0.16 - > 8mm, with speeds ranging from 0.2 - 20m/s and intensities from 0.005 – 250 mm/hr (Thies, 2015). Precipitation is measured in the same way that the PARSIVEL is, but the LPM uses 22 classes for particle size and 20 classes for speed (Thies, 2015). In addition, the LPM has temperature constraints of anything greater than 9 °C being reported as liquid (except hail), and anything below -4 °C is reported as solid (Thies, 2015).

Due to the high resolution of data that these optical sensors are capable of, these instruments seem to be capable of detecting and classifiying solid precipitation quite well. Previous research done by Battaglia and Boudala both used the first version of OTT's PARSIVEL in their snow studies, so I am curious how OTT's second version does. The only research that I have seen comparing the two optical sensors in snow conditions comes from Zhang, and they seemed primarily interested in how the two instruments do in very light snow with windy conditions. Therefore, I will be focusing my research in deriving and comparing snowfall accumulation using the PARSIVEL and LPM to be done using data from the National Center for Atmospheric Research (NCAR) Marshall Instrument Field Site in Boulder, CO.