Abstract -- This project is “Analysis and study of the observation data of debris flow”. It’s purpose is mainly to investigate various hydrological and physiographical factors which might trigger debris flows. In this study, relevant data or information of the factors were collected, tested and analyzed to evaluate the feasibility of the observation method. In addition, they may be used to establish the criteria of the critical rainfall lines for debris flow warning or may be used for reference of other relevant research.
In this study, the test site was in Nan－Tou county. In this area, 28 watersheds were selected as the samples. The selected factors affecting debris flow occurrence included physiographic and hydrological factors which consisted of rainfall and electrical conductivity of groundwater. The physiographic factors were：watershed area, landslide area, stream length, mean width of watershed, form factor of watershed, mean slope steepness of stream, the landuse factor, soil dry unit weight, soil porosity(or void ratio), percentage of the soil particle greater than sieve No.4, percentage of the soil particle finer than sieve No.200, internal friction angle of the soil, cohesion of the soil, soil plasticity index, etc..
To evaluate the feasibility by observing the electrical conductivity (EC) of groundwater to predict debris flow occurrence, a number of tests by using seepage box and water quality tests were conducted, The results showed that there were mainly four compositions which induced the EC to increase. They were：Cl^-, SO4^- , Ca²⁺ and NO3^-. Aside from these, the digitalization of rainfall data was improved so that it was more useful for debris flow warning. In addition, considering the factors of environment and power supply, an automatic observation station of rainfall and EC of groundwater was install at the Hsin-Hsin bridge. Using this station, the real-time data in this area can be observed. Including this station, in this study area, there were five stations for rainfall observation and three stations for EC observation. Thus, a monitoring system for real-time data collection was then installed.