SELF-ORGANIZING MAPS CLASSIFICATION OF SOIL PROPERTIES IN LEVEE USING DIFFERENT GEOPHYSICAL DATA

Abstract

  We propose a classification method to determine the soil type of a river levee using Self-Organizing Maps and k-means method. In order to classify the soil of the levee, electromagnetic waveform, S-wave velocity and resistivity data are acquired by ground penetration radar, surface wave survey, and electrical prospecting, respectively. Envelop and semblance were also calculated from electromagnetic wave reflected waveform. The multi-dimensional data sets are then classified using Self-Organizing Maps and k-means method. The classification results agree very well to the observed nature of the soil properties in the levee. Moreover, we were able to estimate the soil properties from the output of the classification.