英文摘要 |
Test pitting in archaeological excavation is time consuming and not keeping pace with urban development, therefore, the prehistoric sites are virtually destroyed before they are found. The technique of remote sensing can be applied to large area survey for analyzing and predicting possible site distributions. Based on the technique, this study analyzes potential site areas with the assistance of on-site test pitting. It also examines the feasibility of the technique and geological information system in determining alternative test pitting in archeological excavation. The city of Taichung is a highly developed area which might interfere with site exploration and consequently lead to the limitation of satellite image analysis. Therefore, orthophoto base maps are included to evaluate the rivers and determine the correlation between site distributions and geographical environment. Suggestions about possible site distributions as reference for randomized manual digging. Based on archeological experts' experiences, the suggested distributing lines be revised and the on-site archeological excavation and the loction of the pits sampling can be initiated. The study results indicate that the location of the chosen prehistoric cultural settlement is within the range of suggested line sampling which can be used as a reference for trial pitting. With archaeological experts' professional judgment the workload of digging can be reduced. Possibly due to the flat terrain, the most abundant exposed site is 60 meters from the suggested sampling line. However, the slope variation of the terrain and the underground sphere on the north and the east contains rich organic materials and fine soil texture, the evidence of a sediment environment. Li-ming ditch branch stretches across the site of 'public land number 3', therefore, the effects of water flowing through the area are enormous and further influence the accuracy of the predictions on site distribution. It is suggested that future study should include reconstructing information on archaic environment and complete site distribution to improve the accuracy of predicting potential site distributions. |