- 學位論文
應用航空攝影測量於河川土石估算精度分析之研究
The Accuracy Evaluation of Quantity of River Sediment by Using Aerial Photogrammetry
摘要
台灣地區地質脆弱、地形陡峻且地震頻繁，山區易發生山崩及地滑。為減輕颱風豪雨可能帶來之土石災害，國內水利單位針對各河川定期規劃整體治理計畫，適時辦理疏濬，除持續提供砂石作為營建工程骨材使用外，亦可降低河川流域內人民生命財產損失，維護道路交通安全。因此，河道土石淤積量的評估，即成為土砂資源管理中必須掌握的重要數據。 本文試運用航空攝影測量與傳統大斷面測量的比較，計算兩種方式在各斷面高程均方根誤差與土石蘊藏潛量之差異。並以大甲溪不同時期及不同河段為案例，涵蓋馬鞍壩-篤銘橋(89年7月)、河口-后豐橋(89年10月)及河口-篤銘橋(94年4月)。 分析結果顯示，89年10月河口下游區域斷面高程均方根誤差0.76~4.41公尺，89年7月馬鞍壩以上之河段為3.05~12.67公尺，94年4月河口下游區域斷面高程均方根誤差0.80~1.99公尺，馬鞍壩以上之流域1.98~11.07公尺。 在土石蘊藏量差異比較方面，89年10月河口下游區域平均差異比為30%，94年4月河口下游區域平均差異比為9%；89年7月馬鞍壩以上之流域平均差異比為19倍，94年4月平均差異比4倍。 經檢討下游河口區域地勢較平坦，其誤差較小，但兩種方式之測量資料並非在同一時間下獲得，為影響分析結果之因素之一。馬鞍壩以上流域由於地形陡峻，左岸常須以輔助(延後)樁方式施測，故誤差大，影響分析精度。故本文建議兩種測量方法可依地形狀況在同一時期配合進行測量作業將可提高精度。
並列摘要
The landslide disasters in Taiwan are happened easily around the mountion areas due to the reasons including fragile geology, steep terrain and earthquakes. In order to reduce the occurrence of the debris flow hazards which get along with typhoons and heavy rainfall, the governing agency (i.e. the Water Resources Agency, Economic Affairs) conducts administion plans in the central govern rivers and dredges the deposits of the canal on a regular time schedule. The dreding programs besides prevent the people living and transportations in the basin from the floods and the deposit sadiments can provide for construction use constantly. Therefore, the evaluation quantity of river sediments is an important data for the sand and gravel administration policy. This study proposed a method to compare the aerial photogrammetry data with the conventional cross section surveying data of rivers, and to compute the root-mean-square errors on the items of the elevation of each river cross section and the reserves of deposit between two kinds of data. The above method is used to a case study, that is, various sections of the Da-Jia river including the section from the Ma-An Dam to the Du-Ming bridge in July, 2000 and the section from river-end to Hou-Feng Bridge in October, 2000 and the section from river-end to Du-Ming bridge in April, 2005. From the analysis of two kinds of data, the results show that, the root-mean-square errors of the elevation of river cross sections at the river-end area are between 0.76 to 4.41 meters in October, 2000 and the root-mean-square errors are between 3.05 to 12.67 meters at the upstream river of the Ma-An Dam in July, 2000. The root-mean-square errors are between 0.80 to 1.99 meters at the river-end area and are between 1.98 to 11.07 meters at the upstream river of the Ma-An Dam in April, 2005. To compute the differences on the reserves of deposit using the two kinds of data. The differences are 30% in October, 2000 and 9% in April, 2005 respectively at the river-end areas. The differences are 19 times in July, 2000 and 4 times in April, 2005 respectively at the upstream river of the Ma-An Dam. As a conclusion, we got good results at the downstream river of the river-end for its flat terrain and the surveying time of two kinds of data are not the same time will induce the differences. In addition, we got bad results at the upstream river of the Ma-An Dam for its steep terrain and the application of the conventional cross section surveying method in this mountainous area always induced big errors. This study would like to make a suggestion that it will get good results if we conduct the aerial photogrammetry surveying and the conventional cross section surveying at the same time.