傳統以來我們認為引發山崩的兩個主要因素為降雨強度和降雨量，因此大部分的山崩預警系統都使用降雨延時和強度來判斷，而往往忽略水滲入坡體的延時。在本研究中，我們藉由水筒模式中之蓄水指數(water storage index)歷線和即時地表位移之間的相關性，來說明蓄水指數與深層潛移山崩(deep-seated creeping landslide)的連動。以天池坡地區域2004年至2009年由GPS連續地表位移監測所記錄到的四個深層潛移山崩為例。結果顯示山崩位移與蓄水指數，比與降雨強度有更密切的時序發生相關性。
The amount and intensity of precipitation have long been considered to be two of the main factors that trigger landslides, so rainfall duration and intensity are used in virtually all landslide-warning systems. The effect of the infiltration of water and the increase in the pore water pressure in slopes are usually neglected in traditional approaches. This study demonstrates the effect of water storage content in a slope by using the correlation between the transient water storage index and the real time surface displacement. A tank model is used to model the water storage content and four deep-seated creeping landslides in the TienChih area, which recorded continuously using GPS, are studied as examples. The result demonstrates that the soil water index correlates more closely in time to the landslide motion than the rainfall intensity.
