Accurate identification of wet and dry weather periods at sub-hourly time intervals is important for the description and control of processes directly influenced by rainfall, such as infiltration into urban drainage systems, purification processes in wastewater treatment plants, or effective irrigation systems. It is also necessary for monitoring and modeling rainfall itself. Traditional instrumentation used to measure rainfall (rain gauges and radars) often fails to detect the transition between dry and wet weather at sufficient spatial and temporal resolution. Opportunistic sensing has become a promising approach in hydrology to overcome these deficits without drastically increasing the cost of measuring campaigns. In this study, we identify dry and wet weather periods using autonomous and inexpensive transmission line-type electromagnetic sensors, primarily intended for soil water content measurement.Four transmission line-type electromagnetic sensors, a tipping bucket rain gauge, and a laser precipitation monitor were installed in an urban catchment for an experimental period of 3 months during the summer. An algorithm for the reliable detection of the onset and end of precipitation episodes was developed for use with the sensors. Our analysis demonstrates that transmission line-type electromagnetic sensors provide results with accuracy similar to, and with five times greater sensitivity than a tipping bucket rain gauge. However, the sensors produced false-negative results more than 1.6% of the time (i.e., 25% of the received rain). Nevertheless, the low specificity of the sensors is not critical when they are used in combination with rain gauges or other sensors that are less prone to falsely detect wet periods.
Keywords: Laser precipitation monitor; Low-cost sensors; Rainfall detection; Urban hydrology.