Horizontal Reference Network Densification by Multiple Free Stations

Abstract

Having available reference networks close to working areas minimizes error propagation and reduces costs to society when surveying or setting out engineering projects. The densification of reference networks using the Global Navigation Satellite System resulted in a more agile and cost-effective process. However, in locations with a signal obstruction or multipath—typical of urbanized areas—the solution is still to densify through direction and distance observations with total station. Traverse is the most widely applied among the conventional methods, although it has a higher error propagation. Triangulateration, in turn, is a method with higher redundancy and, therefore, less error propagation, but is also less productive, especially in urbanized areas. In this study, we present a new method for the densification of horizontal reference networks through multiple free stations, eliminating centering errors and allowing for unique and simultaneous adjustment of network points. To conduct an a priori quality analysis, the proposed method was compared with triangulateration and traverse methods considering the absolute confidence ellipses of network points; the relative confidence ellipses of the alignments; and the redundancy numbers of the observations in a simulated horizontal network. Although the redundancy of triangulateration guarantees higher reliability, the experiments showed that the multiple free stations method is more productive than traverse and triangulateration as it does not require the target points to be occupied. Besides that, it also provides better precision, that is, smaller confidence ellipses due to centering error elimination.