Performance analysis of multi-GNSS static and RTK techniques in estimating height differences

Establishing reliable elevation differences is imperative for most geoscience and engineering applications. This work has traditionally been accomplished through spirit leveling techniques; however, surveyors have been utilizing satellite positioning systems in measuring height differences for more than a decade. Yet the quality of these heights needs to be evaluated in order to adopt them in different applications. In this article, we present the outcome of an accuracy assessment of height differences obtained with static and RTK surveys. Twenty control points with an average baseline length of 1 km were occupied with dual-frequency GNSS receivers for different time periods. Collected signals were processed using open-source software and verified with an online processing tool. Heights were estimated by processing the GPS and the GLONASS data individually, and combined (i.e. GNSS). Height differences were determined and compared with those measured by spirit levels and corrected through geoid models. Best results were achieved by combining GPS and GLONASS solutions for both static and RTK surveys. Solutions with either GPS or GLONASS satellites were comparable, but in most cases, the GPS solutions performed better. For the static surveys, longer occupation provided much accurate height differences. Inconsistencies among 10 different RTK surveys were minimum for the GPS + GLONASS solutions and worst for the GLONASS solutions. The ANOVA, LSD, F, and chi(2) statistical tests confirmed our findings at the 95% confidence level.