Development of a Low-cost Ultrasonic Sensor for Groundwater Monitoring in Coastal Environments: Validation using Field and Laboratory Observations

Abstract

Gonzaga, B.A.; Alves, D.C.L.; Albuquerque, M.G.; Espinoza, J.M.A.; Almeida, L.P.M., and Weschenfelder, J., 2020. Development of a low-cost ultrasonic sensor for groundwater monitoring in coastal environments: Validation using field and laboratory observations. In: Malvárez, G. and Navas, F. (eds.), Global Coastal Issues of 2020. Journal of Coastal Research, Special Issue No. 95, pp. 1001–1005. Coconut Creek (Florida), ISSN 0749-0208.This study presents the development and validation of a low-cost ultrasonic sensor (HCSR 04), coupled to an open source Arduino microcontroller platform, for groundwater monitoring in coastal environments. Given the fact that low-lying coastal zones are regions naturally prone to flooding (due to oceanic forcing and pluviosity) and densely occupied, monitoring groundwater variations is crucial for the management of flood-related events. A groundwater monitoring well was built, (made of a PVC pipe with 100 mm in diameter and 2 m long) and installed on the ground of a dune field, located in Cassino beach, southern of Brazil. The system was deployed for 84 consecutive days and programmed to perform water level measurements every 10 min. In laboratory, three units with the same hardware setup of the unit installed in the field (ultrasonic sensor coupled to the Arduino microcontroller) were installed on a 20 liter graduated PVC tube. Varying water levels were tested in the laboratory along the time (changing 1 cm every 2 minutes during 4 hours) in order to reproduce the synthetic signal. Eighty measurements were obtained for each sensor, a total of 240 for each validation process. For the laboratory validation, the results for nonparametric statistical tests (Kruskal and Wallis, and Dunn) presented a correlation of 99%, with Root mean square error (RMSE) 0.4113 and bias 0.0418. For the field tests, 3262 data collected by an ultrasonic sensor during 240 hours showed a means registration of 0.543 m. The minimum and maximum cataloged is 0.40 and 0.70 m, respectively, with 0.1076 m of standard deviation. Groundwater classical measurements, using a measuring tape, showed a mean and standard deviation of 0.544 and 0.1069 m, respectively. These preliminary tests showed that the developed monitoring system performs observations with an accuracy and precision within standard methods, therefore can be applied to monitoring changes in the level of groundwater.