貝氏最大熵法應用於雲林地區地層下陷之時空分析

雲林縣位於濁水溪沖積扇南翼，由於雲林地區各項用水長期超出地面水源供水能力，致使各項產業包括民生、工業、養殖及部分農業用水過度抽取地下水，加上本區富含細砂及黏土之地質特性，因此發生地層下陷現象。根據經濟部水利署歷年水準測量資料分析，顯示雲林為全臺顯著下陷面積及下陷速率最大地區，其中高鐵經過雲林縣包括二崙鄉、西螺鎮、虎尾鎮、土庫鎮、元長鄉及北港鎮等6 鄉鎮，仍持續出現嚴重下陷趨勢。針對雲林地區地層下陷問題，本區雖然長期進行地層下陷之監測工作，但由於監測點的種類與建置時間不同，導致實際監測資料在時間和空間上具有不確定性。貝氏最大熵法(Bayesian Maximum Entropy Method, BME)是一種近代時空模式與繪圖(Modern spatiotemporal modeling and mapping)的方法，可自動整合非線性的推估技巧與非高斯的機率密度函數理論，也可以處理不同種類與尺度的資訊，以及不確定性的資料型態。研究結果顯示，在不同年份之地層下陷分布在時空分析上並非具有相同之高斯分布與指數分布特性，雲林地區之地層下陷之時空變化具有相當之複雜性。

關鍵字

地層下陷；貝氏最大熵法；時空分析

並列摘要

Yunlin County, which is located in the south of Choushui River Alluvial Fan, is currently facing the serious land subsidence problems. Due to the lack of sufficient surface water, the supply of agriculture, aquaculture, industry and domestic water are from heavy withdrawal of groundwater. In recent years, the main subsidence areas gradually move inland, causing the security issues of Taiwan High Speed Rail, which route passes through such as Tuku Township, Yuanzhang Village, etc. According to the leveling network was established by Water Resource Agency, the maximum subsidence area and rate of Taiwan in 2010 is Yunlin area. Because of some restrictions of the situation in reality, only limited amount of continuously time serial data can be used. Because of the uncertainty due to a lack of knowledge about the behavior of the system that is conceptually resolvable. The Bayesian Maximum Entropy (the BME method) is a modern method in the time and spatiotemporal statistics problem. This study applied the BME method to estimate the land subsidence distribution and different kinds of scale and uncertainty subsidence data in Yunlin County. According to our research, subsidents in different years have different Gaussian and Exponential characters on spatiotemporal geostatistics. It shows that variety of subsidence in Yunlin area have considerable complexity.

並列關鍵字

land subsidence ； Bayesian Maximum Entropy Method ； Taiwan high speed rail

國際替代計量

貝氏最大熵法應用於雲林地區地層下陷之時空分析

全文下載

主題瀏覽