Study of soil moisture sensor for landslide early warning system: Experiment in laboratory scale

Abstract

The high rate of rainfall is the main trigger factor in many cases of landslides. However, each type of soils has unique characteristics and behavior concerning the rainfall infiltration. Therefore, early warning system of landslide will be more accurate by monitoring the changes of ground water condition. In this study, the monitoring of ground water changes was designed by using soil moisture sensor and simple microcontroller for data processing. The performance of soil moisture sensor was calibrated using the gravimetric method. To determine the soil characteristic and behavior with respect to water content that induce landslides, an experiment involving small-scale landslide model was conducted. From these experiments, the electric resistance of the soil increased as soil water content increases. The increase of soil water content led to the rise of the pore pressure and soil weight which could cause soil vulnerability to the movement. In addition, the various soil types were used to determine the responses of soils that induce the slope failure. Experimental results showed that each type of soils has different volumetric water content, soil matrix suction and shear strength of the slope. This condition influenced the slope stability that trigger of landslide.