考慮地表動態條件下之都市地區高程控制系統維護策略

Abstract

我國位處環太平洋火山地震帶，受到歐亞大陸板塊與菲律賓海板塊的板塊運動，基準坐標系統更容易受到地表作用影響，長期誤差累積會造成控制坐標系精度降低，在高程控制坐標系統中，相較平面坐標系統受到板塊作用影響下，例如地下水抽取、工程施工位移等更多的複雜因素牽扯其中，能夠維護良好的高程坐標系統實屬不易。本研究以建構都會地區高程控制系統的維護為目標，以我國首都臺北市作為實驗區域，以臺北市內連續運行基準站進行解算，包含TPE-CORS和GPSLAB的資料，其中TPE-CORS使用GAMIT/GLOBK以獲得高精度的高程時間序列成果，並使用克利金及反距離權重空間內插方法建構臺北市區域速度場模型，在平差計算的部分，以臺北市105年和107年的水準測量資料，進行時變模型平差，考慮地表速度場的內驗精度和外部精度都較不考慮地表速度場的平差成果來的佳，顯示時變模型更符合真實的情形，可避免高程控制點誤差累積的問題，可持續增加不同年度的觀測量逐步進行平差，使高程控制測量成果隨時間持續保存，實現動態坐標系統的概念。本研究更依據水準測量相關行政規則，計算垂直、相對速度場結合精度規範分析，在相對速度場推算容許誤差所需改正的年數，一等的水準測量精度內需每5年改正維護一次，而二等為9年，加密控制為15年，工程精度約36年才需改正，可見臺北市連續運行基準站可維持長時間的觀測。

With the rapid development of urbanization in recent years, urban land price has continued to rise. Urban areas measurements become an important issue. In order to manage the land effectively, vertical datum in metropolitan areas usually use leveling method to fix the height of control points. Since the Continuously Operating Reference Stations (CORS) technology has been invented, three-dimension coordinates of control points can be measured continuously. With the frequencies of measurement become higher, more surface behavior information is provided. However, the accuracy of vertical datum in cities often declines because of plate movement. For example, Asian cities is located on circum-Pacific belt. Plate movement causes significant surface changes, and vertical datum may cause discontinuous movements due to the earth's crust. After considering the dynamic behavior of the surface, the vertical datum in urban areas must consider dynamic behavior to maintain a high-quality land management. This study integrates GNSS data with annual urban leveling data. The experiment would be implemented in Taipei city, the capital city of Taiwan. In order to obtain high quality time series result, TPE-CORS (Taipei CORS) data would be processed by GAMIT/GLOBK.GPSLAB data is processed by GIPSY. The vertical velocity field interpolated model is constructed by Kriging and IDW method. For the results from time-variant model, the precision would increase after considering vertical velocity field. From these results, time-variant model can apply on annual leveling network data precisely, and ensure the reliability and stability of vertical datum in urban areas.