Hotplate-derived Wind Speed and Snowfall Rate

Retrieval of snowfall based on National Weather Service radars is calibrated using a network of precipitation gauges. Problems with this include, 1) capping of gauges by snowfall, 2) delay in gauge registration of the snowfall, and 3) wind-induced undercatch. A gauge developed by NCAR and DRI measures the electrical power required to vaporize snow that falls on an upward-facing heated surface. This instrument, here referred to as the hotplate, converts the measured power to a liquid-equivalent snowfall rate. Because a hotplate’s upward- and downward-facing surfaces are maintained at 70 ^oC, this type of precipitation gauge is immune to problems #1 and #2. In addition, a hotplate evaluates wind speed and uses that to account for snow particle undercatch. Two publications reveal substantial bias in a hotplate’s estimate of wind speed. In one of these the bias is positive, in the other it is negative. Our premise is that bias occurs because the instrument’s calibration is incapable of accounting for site-specific characteristics encountered in a field deployment. We have developed a new way of processing data acquired by a hotplate system. Our method is based on two Nusselt-Reynolds relationships. One is for a hotplate’s upward-facing surface. This relationship, with the measured input power and a power budget model, are used to calculate the liquid-equivalent precipitation rate. Using an input power measurement and a power budget for the downward-facing surface, the second Nusselt-Reynolds relationship is used to calculate wind speed. Our approach requires simultaneous logging of data from a hotplate and a reference wind measurement system. Using our approach, the hotplate wind speed bias is corrected, unbiased snow particle catch efficiencies are derived, and bias in estimation of liquid-equivalent rate is minimized. In contrast with the NCAR/DRI method, ours derives the wind-dependent catch efficiency at gauge height (2 m). Thus, our approach does not require measurement of wind at 10 m. The latter, or an estimate of the latter, is needed in the NCAR/DRI method; this is because the 10-m wind speed is the quantity in the NCAR/DRI catch efficiency function. This talk will explain our method for obtaining wind speed and snowfall rate from a hotplate system and will contrast that with the NCAR/DRI method.