Abstract
Taiwan is one of the world’s worst natural disaster area and this situation tends more severe due to the impacts of global warming which made the extreme rain events not caused by typhoon more frequency than before not to mention that cause by typhoon. Heavy rainfalls are commonly occurred in local or large-scale areas due to the characteristics of spatial and temporal uneven distribution of rainfall and those rainfall amounts often exceed the design of mountain drainage facilities which results flooding, sand disasters and losing of life and property. Typhoon Morakot is a typical example of extreme weather event which caused unprecedented severe disasters in typhoon history of Taiwan. The frequency and scale of meteorological disasters have become more difficult to predict and the large scale of landslides and slope collapses induced by rainfall are prone to happened more and more frequent in mountainous area. Therefore, it is necessary to establish consistent interpretation criteria by analysis historical landslides for understanding the dynamic changes and causes of slope collapses in large and medium catchment area. Then according to the causes and characteristics of disasters, the relevant countermeasures and management plan can be made for governments at all levels. This project tries to conduct the interpretation of landslides in the study area from remote sensing data and satellite images through the geographic information system techniques combined with overlapping maps of environmental factors (roads, streams, geologically sensitive areas, rainfall intensity distribution, land use availability or land use patterns) to judge the main causes and mechanisms of slope failures and collapsed land and propose management countermeasures as a reference for relevant public affairs departments in prevention and treatment of disaster, rehabilitation of degraded land, and disaster prevention and relief. In addition, the investigation on the ability of water conservation for a watershed is another focal topic of this project. Although there are many approaches to estimate the ability of water conservation for a watershed such as soil porosity estimation, water budget balance method, baseflow separation technique and recession-curve- displacement method, all of them have using limitations and advantages and disadvantages and limited on static calculating amount of water within a certain period of time. In view of this, this study will compare different approach by using historical rainfall and discharge records of small upstream forested water to determine the most suitable approach for estimation the amount of conserved water and clarify the influence of rainfall characteristics on water conservation in the first year. The impact of natural disturbances (landslides) on water conservation will be explored in the second year. Then the functional relationship among rainfall characteristics (amount, intensity and distribution), natural environmental factors (geology, soil, topography and vegetation) and disturbance (landslides, land use patterns) will be established in the third year. Hopefully we can more understanding the key factors that affecting water conservation in forest watersheds and analyze the impact of extreme climate on water and soil resources.