Case Histories of Successful Early Warnings

This chapter addresses the practice of slope monitoring and early warning by using tilt sensors developed by the authors. Based on case histories of successful early warning, the rainfall threshold has proven effective for regional early warning, even though assessing landslide risk for any specific slope remains beyond its scope. However, there are more demands by local communities that are afraid of the instability of particular slopes in their vicinity. This is the reason why the tilt sensors were developed. People generally do not complain about false-positive errors, where warnings are issued but no disaster occurs. Therefore, we should not insist on excessive accuracy in early warning. Instead, we should prioritize avoiding false-negative errors, where an unexpected disaster causes damage to a community. The authors have conducted numerous slope monitoring projects, including field tests involving artificial rainfall that induced natural slope failures. These experiences have demonstrated which values of monitored displacement and velocity are relevant for issuing warnings well in advance (several hours before) of an ultimate slope failure. The conclusion is that slope deformation velocity is a more reliable indicator than displacement. The recorded data are presented in this chapter. Notably, field monitoring addresses not only rainfall-induced failures of natural slopes but also gravity-induced slope instability and safety management for construction projects where workers operate below newly cut slopes. The study sites are internationally distributed, including locations in India, Bhutan, China, Taiwan, Japan and Australia. Additionally, an attempt has been made to install 60 sensors on a single slope to monitor the two-dimensional progression of slope failure in Japan. To date, no failures have occurred at these sites.