應用地理統計於地下水觀測站網之建置分析

Abstract

地下水水文資訊為構成水資源規劃、監測與管理之最基本要素。規劃良好的地下水觀測站網可以提供代表該地下水區域特性之精確且可靠的地下水流流況資訊，影響地下水觀測站網之設立通常包含地形、含水層厚度及地下水文參數等因子，諸多因子將會影響到地下水觀測站網之位置與密度。過去雖投入大量人力、物力於集水區地下水觀測站站網設計之研究，但對於地下水觀測站設站之位置以及所需設站的最小站數仍無法得到確切的答案。故有必要進一步研究地下水區域設站之位置，以及可能反映該地下水區域之最少站數之問題。 本研究所使用之資料為1998-2007年十年間之年平均水位差，並使用克利金推估法進行水位差之空間推估，由空間推估結果得知雲林地區地層下陷主要抽水層為第二含水層，而濁水溪以北之彰化地區造成地層下陷之主要抽水層為第三含水層。本研究針對濁水溪沖積扇各分層利用控制面積百分比，進行評估濁水溪各分層地下水觀測井密度適宜性，並探討地下水觀測站是否可提供所需之地下水資訊，若不足則須增站。增站方式則使用次序佈井法做為依據，根據濁水溪沖積扇之1171個1公里 1公里之格網進行優選；並將傳統次序佈井法之順序從新排序計算，得到一新次序，此新次序排列可使得排序(rank)較小之新增井能具有較大之總變異量減少量(TVR)。然而重新排序後之新增井仍與傳統次序佈井法相同，所造成之總變異量也相同，為改善此一現象，本研究使用改良重新排序次序佈井法進行演算，使最後總變異量減少量多增加10%。

The basic elements of groundwater planning, monitoring, and management is groundwater hydro-informatics. A well-planned network of groundwater monitoring stations can provide accurate and reliable, regional characteristics of groundwater information of the establishment of groundwater monitoring networks stations typically include topography, thickness and groundwater aquifer parameters and other factors, many factors will affect to the groundwater monitoring network for the location and density. In this study, it proposed a geostatistic approach for underground water monitoring well network design. It was applying hydraulic head as the input data to obtain the spatial structure of the study area, Choushui river basin. In order to measure the uncertainty of the candidate wells, Kriging Variance is proposed to be the index in this study. First, deciding which monitoring wells should be removed from the original network due to its bad spatial location by using the Control –area algorithm. A sequential network design approach decides which candidate well should be added into the groundwater monitoring network. The result shows that 10 of 46 wells removed from the second layer,10 of 42 wells removed from the third layer, and 20 wells was suggested to add into the groundwater monitoring well network in the fourth layer.