探討都市土地利用型態與淹水潛勢之空間關聯──以原臺中市為例

Spatial Relationship between Land Use and Flood Potential-- A Case Study of Taichung City

臺灣由於都市化的影響，人口數量急遽上升，導致不透水地表面積持續增加，加上極端氣候導致更強的降雨量，造成地表逕流的增加，使得都市地區淹水災害的程度日漸嚴重。許多研究指出都市之空間型態為影響淹水危害度之重要因素，然而其實際關聯性較少在過去研究中被深入探討。因此，本研究嘗試整合空間型態分析與淹水模型，探討兩者之間的空間關聯。本研究首先針對1995年及2015年原臺中市各村里之六種土地利用類型，描述其面積、組成、破碎度、聚集度、連接度及離散度，接著透過FLO-2D淹水模型模擬都市淹水潛勢，最後，利用多元迴歸分析方法探討各村里的土地利用型態與淹水潛勢之間的關聯。研究結果顯示都市淹水主要以建成地影響最大；建成地聚集度越高，連接度越高且破碎度越低，導致淹水情形越嚴重，而淹水情形越嚴重之地區，其建成地聚集度和連接度亦較高，且破碎度較低；另外，當單一範圍內都市土地利用類型越多，淹水情形越不嚴重。未來土地利用型態規劃上，應避免建成地聚集度和連接度過高，並採區域內混合使用等方式以助於緩解該區的淹水情形。本研究期望所建立之模式可作為後續都市空間規劃上淹水減緩之參考依據。

Urbanization is leading to increasing impervious surface area and urban population. On the other hand, extreme weather leads to stronger rainfall events, resulting in increased surface runoff that generates more severe flooding events in urban areas. Since existing studies do not discuss the relationships between spatial patterns of urban land use and flooding disasters, which are important in successful planning strategies for flood risk mitigations, this study applies quantitative landscape indices to describe the spatial pattern of urban land use from the perspective of landscape ecology. In addition, this study simulates flooding situations in urban areas through a dynamic flood model, then applies statistical analysis to investigate the relationship between the two. The analysis indicates that the built-up areas have the highest degree of aggregation. Urban flooding is then simulated using the FLO-2D model. Simulation results reveal that the flooded area and the average flooding depth have increased from 1995 to 2015. Finally, multiple linear regression analysis is performed to examine the relationship between indices of land use patterns (including Percentage of Landscape (PLAND), Aggregation Index (AI), Patch Cohesion Index (COHESION), Landscape Division Index (DIVISION), Patch Density (PD) and Edge Density (ED)) and flooding potentials on the village scale. Analytical results show that the spatial patterns of built-up, agriculture and transportation uses affected flooding patterns the most. In conclusion, the relationship between flooding potential and urban spatial patterns is understood based on the framework developed in this research. According to the analytical results, flooding potentials are significantly affected by the spatial patterns of built-up areas. Flood risk can be mitigated in the future by comprehensively planning and configuring the spatial patterns of builtup areas, and avoided high degrees of aggregation and connectivity in built-up areas. The framework established in this work is presented as a tool for improving current flooding conditions in future urban spatial planning processes.