雲林地區地下水位變動與地層下陷趨勢探討

Abstract

雲林地區擁有豐富的地下水資源，然而，長期以來對於地下水資源的依賴已導致部分地區面臨地下水位下降和地層下陷等嚴重問題。地層下陷可能會導致海水倒灌、淹水、建築物倒塌、交通設施受損等問題；海水倒灌會影響水資源之水質，淹水對於沿海地區、低窪地帶的居民將造成財產損失，更可能影響生命上的安全，而基礎建設受損，若加上颱風等天然災害的威脅，會導致重建工作更加困難。有鑑與此，本研究旨在探討不同含水層之地下水位變動與地層下陷之關聯性，透過分析不同地區之地下水位變動，找出地下水季節性變動、長期趨勢、區域特性等，對於雲林地區地下水資源狀況與地層下陷有更深入的了解。 本研究主要分為兩大部分，首先為不同深度地層年壓縮量之空間分佈探討；接著透過地下水位月變動量建立地層月壓縮量之線性回歸模型。不同深度地層壓縮量之空間分佈探討主要是透過經驗正交函數(Empirical Orthogonal Function, EOFs)來了解不同深度之地層年壓縮量的變異量時空間分佈。在線性回歸模型中，本研究首先針對所蒐集之四層地下水位進行空間內插，以對應地層下陷資料之空間位置；接著透過主成分分析法來探討不同含水層間地下水位變動之關聯性，透過主成分分析法之結果了解各主成分的相對貢獻，同時也可以了解不同層地下水位變動之共動變化，辨識主要變異性；最後將主成分分析之結果做季節與長期趨勢分解，做為線性回歸模型之自變數，以探討地層月壓縮與地下水位變化之關聯性。

The Yunlin region possesses abundant groundwater resources; however, the long-term reliance on these groundwater resources has led to serious issues such as declining groundwater levels and land subsidence in certain areas. Land subsidence can result in seawater intrusion, flooding, building collapses, damage to transportation infrastructure, and other problems. Seawater intrusion can affect the quality of water resources, and flooding can cause property losses for coastal and low-lying area residents, potentially endangering lives and making post-disaster reconstruction more challenging, especially in the face of natural disasters like typhoons. In light of these challenges, this study aims to investigate the relationship between groundwater level fluctuations in different aquifers and land subsidence in the Yunlin region. By analyzing groundwater level variations in different areas, we seek to gain a deeper understanding of groundwater resources and land subsidence in the region, considering seasonal fluctuations, long-term trends, and regional characteristics. The study is divided into two main parts. First, it explores the spatial distribution of annual aquifer compaction at different depths. Subsequently, a linear regression model is developed to establish the relationship between monthly groundwater level variations and monthly land compaction. The spatial distribution of compaction at different depths is investigated primarily through Empirical Orthogonal Functions (EOFs) to understand the spatial-temporal distribution of annual compaction at different depths. In the linear regression model, four layers of groundwater levels collected are first spatially interpolated to correspond to the spatial locations of land subsidence data. Principal Component Analysis (PCA) is then used to explore the relationship between groundwater level fluctuations in different aquifers, identify the relative contributions of each principal component, and understand the co-variability of groundwater level variations in different layers. The results of PCA are then used to decompose seasonal and long-term trends, serving as independent variables in the linear regression model to investigate the relationship between monthly land compaction and groundwater level fluctuations.