Time-Variant Adjustment for a Level Network

Abstract

It has been routine work to determine the elevations of points on Earth's surface for a wide range of purposes. However, surveyors are now facing new challenges because of the increasing demand for high-quality measurements despite potential errors caused by Earth's dynamics, which introduce time-dependent signals in the points being investigated. Classical level network adjustment works well when benchmark elevations are known precisely and remain constant over time. However, the traditional approach fails to consider possible ground motions in a single process of network adjustment. This study proposes a time-variant model and unified least-squares technique for level network adjustment. The main emphasis of this approach is to accommodate the time-dependent behaviors of benchmarks caused by ground motions and other unknown sources. The numerical results from a simulated test and a real case study show that the proposed approach can capture the time-dependent signals of benchmarks. Consequently, significant perturbations in benchmarks can be identified, and an improved analysis for level network adjustment thus can be achieved.