台灣屬於複合型災害頻率高的國家,且位置正好座落於颱風易生成區域的附近,每年夏季必然遭受颱風豪雨侵襲。因降雨時間、強度及空間分布不均,大雨成災與無雨缺水交替的情況愈趨明顯。另一方面,臺灣有超過75%以上是山坡地,各類型的坡地災害亦經常發生。 本研究區域位於臺北市文山區猴山岳步道附近,於2008年到2009年之間,因為連續颱風所夾帶的豪大雨,造成研究區域約10公頃的邊坡滑動,研究利用前人所訂定的臺北市崩塌警戒標準(Lin et al., 2003),進行雨量指標計算的程式開發。雨量指標程式計算結果顯示在2008年因辛樂克颱風影響下,木柵雨量測站之歷史雨量資料所計算出來的雨量指標是明顯超過崩塌警戒基準的。 研究中也參考萬用土壤流失公式(Universal Soil Loss Equation)進行猴山岳崩塌區域的沖蝕量推估,沖蝕量計算結果顯示在發生大規模崩塌之前每年約有53公噸的土壤沖蝕量(2007)。2008年為崩塌過後,研究區域地表植生遭到嚴重破壞,裸露地占總計算面積的31%,土壤沖蝕量每年高達2958公噸。2009年由於地表植生的自然恢復,許多原本為裸露地的區域長出了芒草,土壤沖蝕量從每年的2958公噸降至690公噸。 由於USLE公式的誤差範圍甚大,且上述結果明顯高估,為了評估沖蝕量計算的準確性,研究中使用地面雷射掃描儀進行各獨立邊坡掃描,得到不同時期之數值地形模型,進而計算真實土方量差異,本文中利用單一獨立邊坡的掃描資料進行土壤沖蝕量的準確性評估,結果顯示USLE推估的土壤沖蝕減去真實土方差所得到的誤差值,約為真實土方差的-12%至45%。
Taiwan has high frequency of compound disaster and is located nearby the area where typhoon forms easily, so it definitely encounter attack of typhoons and heavy rainfall in summer every year. Due to inconsistent rainfall duration, intensity, and space, the trend of alternation between rainfall caused disaster and rainfall deficit is becoming obvious. Moreover, since slopeland occupies over 75% of Taiwan’s area, various types of slopeland disasters happen often. Subject of this study is the trail located in Mt. Houshanyue in Wenshan District in Taipei. During 2008, heavy rainfall caused by successive typhoons created about 10 hectares of landslide on the slope. This study adopted the “Standard of Landslide Alert Taipei City (Lin et al., 2003) regulated by former researchers to develop a program for rainfall index calculation. The result of rainfall index program’s calculation showed that under the effect of Typhoon Sinlaku the rainfall observed and the index calculated by rainfall measuring stations in Mucha District obviously exceeded the reference threshold of landslide alert. This study also adopted Universal Soil Loss Equation to estimate erosion of Mt. Houshanyue’s landslide area. Calculation of erosion indicated that this area has had 53 tons of soil erosion every year before the large-scale landslide (2007). The landslide in 2008 seriously damaged surface vegetation of the subject area which caused 31% share of exposed area of the total area and soil erosion sums up to 2958 tons every year. During 2009, the surface vegetation recovered naturally in this area where much of the exposed area appeared miscanthus and volume of soil erosion reduced from 2959 tons to 690 tons. In order to calculate volume of soil erosion, this study detected each slope with terrestrial laser scanner. Then it assessed accuracy of soil erosion volume for each slope’s data detected. The result indicated that estimated error of soil erosion volume was between 12% and 45% of difference of soil volume estimated.