Evaluation of saltation flux impact responders (Safires) for measuring instantaneous aeolian sand transport intensity

Abstract

The measurement of aeolian sand transport rates on small scales is of interest to the development and testing of detailed models of sand movement by wind. This paper reports on laboratory evaluations and preliminary field tests of a new design of a piëzo-electric impact responder, called a ‘Safire’, capable of measuring saltation impacts at a frequency of 20 Hz. The advantages of the Safire are: (1) that it provides high-frequency measurements, (2) that it presents a minimal obstruction to the wind flow (no scour observed in the field), and (3) that it is of a (relatively) low-cost. Laboratory calibrations were performed with a vertical gravity flume generating known sand grain fluxes using both mixed sand and specific size fractions. Initial tests investigated three fundamental characteristics: correspondence between digital and analogue signals generated by the instrument, directional response of the probe, and linearity of instrument response to mass flux. Instrument calibration included determination of the momentum threshold required for the sensor to register a grain impact. Based on this lower limit and the known distribution of grain size and speed at different fall heights, a prediction is made as to the sand grain flux the Safire ought to measure, which is then compared with the signal response. The result of this comparison is an assessment of the instrument's efficiency in counting saltating grains. These Safires were also deployed in the field as part of a larger investigation of spatio-temporal transport variability. This experiment provided the opportunity to compare the instrument's performance with traditional sand traps, and this paper develops methods and assumptions required to convert measurements from impact responders to traditional mass transport rates. The evaluations indicate that improvements to the instrument production process are required to ensure a standard momentum threshold among individual instruments. Furthermore, the sensor design needs to be reconsidered in order to eliminate the variation in response depending on azimuth direction, so that the sensor is uniformly omni-directional.