台灣地質構造複雜,在颱風、豪雨季節,常導致山區土砂災害發生。為加強山區災害防治、長時間監測及災後重建等,需要更完善的三維資料。一般災前、後收集到三維資料時,常受限於經費或精度需求,而必須面臨使用不同三維資料來源之問題,如航空攝影測量、空載及地面光達等。而這些不同資料來源都具有不確定性,導致影響後續資料分析,如影響邊坡崩塌與堆積量之評估、邊坡變遷分析之高程差異量計算或土石變異量等。 而一般採用歷史資料分析變遷,所收集到的資料型態也有所不同,各類型的資料也都具有不確定性,因此本研究利用此特性,收集不同三維資料來源之2m數值等高線、4m網格之DTM及點雲。將所收集到不同三維資料組合得知變遷分析,而變遷分析結果也具有不確定性。因此本研究建立不確定性之方法,以評估變遷分析成果。 本研究選定台北市文山區猴山岳為研究區域,針對三個不同崩塌案例,來探討多時段多來源三維資料不確定性,所造成不同時期高程差異量之影響,並得知崩塌、堆積機率範圍及評估影響崩塌與堆積量。評估出多資料來源精度,並利用機率評估出土石變異量,以便提供決策者及後續相關單位研究等完整資料。
Taiwan is an island made up of highly complex geological structures. During seasons of typhoon and heavy rainfall, sediment disasters occur frequently in mountain areas. For the sake of enhancing disaster prevention, long-term monitoring and post-disaster reconstruction in mountain areas, more comprehensive three-dimensional data is of high necessity. Due to limited budget or accuracy requirements, pre/post-disaster three-dimensional data are often obtained from multiple resources, such as aerial photogrammetry and airborne LiDAR or ground-based LiDAR. However, uncertainty exists in data resulting from different methods, and this can affect subsequent data analysis, for instance during evaluation of slope failure and accumulation amounts, calculation of elevation differences in cases of slope change or debirs Variation. As for change analysis utilizing historical data, collected data also derive from multiple sources, and uncertainty remains to be an issue. This study therefore collected data including 2m contours, 4m grid DTMs and point clouds derived from different three-dimensional data sources. These data were then combined to perform change analysis, and results indicated existence of uncertainty as well. A method regarding uncertainty was therefore established to evaluate results of change analysis. The selected site of interest was at Houshanyue, located in the Wen-shan district of Taipei City. Three different cases of sediment collapse were taken into consideration in order to investigate the effect of inter-period elevation difference due to the uncertainty in multi-period, multi-source three-dimensional data. Calculation of collapse and accumulation probability range and evaluation of collapse and accumulation amounts were also performed. The accuracy of multi-source data was estimated, and collapse amount was estimated based on probability. These results can serve as reference in subsequent decision-making and research by relevant authorities.