Signal processing and calibration of a low-cost inductive rain sensor for raindrop detection and precipitation calculation

Abstract

Precipitation measurements in urban areas are becoming more and more important due to the increasing number of heavy rainfall events. In order to establish a dense measurement network within a city, low-cost sensors are necessary. A low-cost inductive rain sensor was developed for this purpose. In this study, the signal processing of the sensor is presented. To collect measurement data, the sensor was deployed in the field next to a high-precision reference sensor. An algorithm has been designed to detect the drops falling on the sensor. Subsequently, two methods have been developed to determine the amount of precipitation from the detected drops. The first approach is based on the statistical drop size distribution, which can be described by a gamma function. In the second approach, drop sizes are assigned to the detected drops based on their peak height. The drop detection algorithm has an accuracy of 89.25 %. The two methods for precipitation calculation have deviations of 11.11 % and 11.51 % from the reference measurement.