FEASIBILITY OF DOL-VRS SERVICE FOR ESTABLISHING SURVEY CONTROL USING POST-PROCESSING METHOD

摘要

Since 2008, the Department of Lands (DOL), Thailand has operated a Network based Real Time Kinematic service to support its cadastral surveying applications. The service implements a Virtual Reference Station (VRS) concept based on a network of 11 CORS stations located mainly in the Central Thailand. Such service reduces several technical limitations of conventional RTK method and increases the productivity in a cost effective manner. This also opens the question how feasible DOL-VRS service may be for establishing survey control. The conventional way to set up high accuracy survey control for geospatial applications is in post processing mode using relative positioning method and double differenced carrier phase algorithms. Apart of the rover, the user needs a second receiver to collect data simultaneously and at the same epoch rate at a known point. VRS eliminates this requirement by creating synthetic observations for a non-physical, un-occupied, invisible reference station situated only few meters from the approximate location of the rover. On the other hand, previous studies in Thailand concluded that the ionospheric error is the main error source that degrades the VRS performance. Thus, this study investigates the feasibility from two points of view. The first point is the observation time. The second point is the influence of ionospheric conditions on the final network based solution. Therefore, we define four relevant cases: low, medium, high, and extreme cases as function of total electron content value. In addition, the ambiguity fixing rate, rate position jump, and root mean squared error are defined as performance indicators. The numerical analysis reveal that the reliable survey control may be obtained when at least 180 epochs of observation are considered on good ionospheric conditions (i.e., low TEC). On the other hand, 720 epochs or more are needed when the ionospheric activity increases. The increase in ionospheric intensity influences directly the spread of the solution on both horizontal and vertical directions. In addition, the VRS based solutions compared to an average least square adjustment solution show an offset of about 0.7 cm. Furthermore, the errors in the vertical direction for 720-epoch solutions display a peak-to-peak variation of 3.9 cm (low TEC) and 10.1 cm (high TEC), respectively. We can conclude that VRS based survey control may be established using DOL-VRS service precisely and accurately at 2-cm level especially in the low TEC scenario. When the ionospheric activity increases, the user is advised to use this approach with caution and take the average of multiple occupations on the same mark. The occupations should be separated in time to allow changes in the constellation geometry and ionospheric conditions. Such practice will enhance the reliability to establish high accuracy survey control.