氣象災害影響下山村區域土地承載力及產業資源空間配置整合研究 Spatial-based analysis of resource carrying capacity and rural industries distribution.

本計畫研究期間為105至108年。105年以水里溪流域、106年以南投縣、107年以台中市、彰化縣和南投縣、108年以高屏溪流域為研究範圍，進行水文及土石災害、農林資源調查、環境土壤調查之成果資料整合。利用多期歷史崩塌圖資，以及依據內政部國土利用調查及人口普查資料建立之統計推估模型，於地理資訊系統分析人口分布對於崩塌災害之暴露程度，並找出可能的高風險山村地點。此外，針對崩塌發生區位進行環境因子分析，瞭解崩塌事件好發地點之環境與地質特徵。另針對歷史崩塌地進行現生植被及物種組成調查，瞭解森林的自然回復能力與時間序列，以物種組成建立生態回復力評估系統，亦針對不同立地條件及干擾強度，提出適宜之適林經營或復育策略。研究工作共計完成：（1）土地利用資料清查與整理，建構空間資料庫；（2）運用資料探勘方法，完成高解析度人口密度空間預測模型建置，作為災害風險分析之人口暴露度指標；（3）與崩塌歷史紀錄及潛勢區域結合，篩選可能暴露於高度災害風險的山村聚落；（4）提出生態回復力評估系統之構想與分析方法，並依既有野外調查資料，完成植被演替分析及適宜性復育造林策略研提。本資料集內容與資料為林業試驗所108農科計畫項下之「氣象災害影響下山村區域土地承載力及產業資源空間配置整合研究。」。

Several scientific reports have demonstrated that the frequency and intensity of extreme climatic events such as heatwaves, droughts, and heavy rains have increased in Taiwan during the past 100 years. Some model projections indicate that substantial precipitation would be more concentrated in the wet season, and the lack of water availability would be more evident in the future. It also predicts that the average increase in extreme daily rainfall will be 10% to 30% in almost all regions of Taiwan by the end of the twenty-first century. Taiwan is an island with high disaster risks due to its steep mountains, active geological events, and the intense anthropogenic land use. The increased extreme climatic events might make the rural areas being exposed to high risk of meteorological disasters, especially for indigenous villages and poor economically developed counties.
This project took Taichung City, Changhua County, Nantou County, and the Kaoping river watershed, as an example area to develop a GIS-based model, which was composed by anthropogenic and environmental factors such as present and historical land use, historical landslide events, geological and topographical factors, limitation of slope land use and social-economical inventory data, to explore the exposure of rural hamlets and agroforestry industry to landslide disasters. We developed a land use-based model for estimating the distribution of population in rural areas, which can be utilized as a social-economical factor overlaid with the historical landslide events to reveal the high-risk rural villages. Furthermore, a time-series species composition of secondary forests on historical landslide regions was investigated and studied to reveal the successional series of forests at different altitudes. Based on our results, an adapted strategy for re-establishing forest in the landslide region, which includes the selection of tree species and its suitable altitudinal/climatic environments.