| Abstract 
              -- This project is a study on the warning model of 
              debris flow occurrence by using the rainfall and the electrical 
              conductivity of groundwater. The purposes are mainly to investigate 
              the hydrological and physiographic factors which trigger debris 
              flows, and to evaluate the critical rainfall lines of debris flow 
              occurrence and their revisions for the warning of debris flow after 
              the Chi-Chi earthquake. Besides, a preliminary study was conducted 
              on the observation and theoretical analyses of the electrical conductivity 
              (EC) of groundwater. In this study, the Nan-Tou county was selected as the test site 
              and 28 watersheds in this area were selected as the samples. After 
              examination by using the Mann-Whitney-Wilcoxon test, it was found 
              that among the 14 factors, four factors were relatively highly related 
              to debris flow occurrence. The four factors were : landside area, 
              landuse factor, percentage of the soil particle greater than sieve 
              No.4 and effective watershed area. Equations for predicting debris 
              flow occurrence in the Nan-Tou area were proposed by using Fisher’s 
              discriminant function of multivariate statistical analysis. After 
              the Chi-Chi earthquake, the potential of debris flow of the streams 
              increased. Accordingly, the peak ground acceleration (PGA) of the 
              earthquake and a function decreasing with time were used to revise 
              the critical rainfall line for each watershed after the great earthquake. 
              In regard to the relationship between EC and debris flow occurrence, 
              to evaluate the feasibility by using the EC of groundwater to predict 
              the occurrence of debris flow, a number of tests by using seepage 
              box and water quality tests were conducted. It was found that four 
              compositions, namely, Cl-, SO4-, Ca2+ and NO3-, induced the EC value 
              to increase.
 In the site, considering the factors of environment and electrical 
              power supply, a real-time analyzing system was installed at the 
              Hsin-Hisn bridge which was about 2 km away from the Long-Hua elementary 
              school. The real-time rainfall and EC data observed from the station 
              at the Hsin-Hsin bridge was transmitted to the system at the Long-Hua 
              elementary school by using the GSM (wireless transmission) system. 
              After analyzing, the information was transmitted to the monitoring 
              center in Taipei through the internet system. Aside from these, 
              the information could be used for on-line demonstration and could 
              be shared with other.
 
    |