Journal of Surveying Engineering

In this paper, the use of the parameter estimation method in a split functional model in the analysis of geodetic network deformation is presented. Special attention is drawn to the situation when competitive parameters fulfill certain independent conditions in the systems of observation equations. It is also proposed that the so-called reversal-point effect can be eliminated from the M_split estimation using conditional equations binding the competitive parameters. It is shown that the introduced extension of the M_split estimation method does not deprive the studied method of its qualities, including its robustness.

The effect of correlations in global navigation satellite system (GNSS) double difference observation has been investigated, and it has been empirically found that such effect is small in local networks (baseline up to a few kilometers). In this work, we show that if the extension of the network is very small (a few kilometers), if all receivers of the network observe the same satellites (no occlusion), and the network baselines are created by the star strategy, the discussed correlations do not have an impact at all on the estimates of the coordinates. The equivalence between the network solution and the single baseline solutions is first derived theoretically and then tested on a small real network. A method to derive the overall covariance matrix of the network from single baseline solutions is given.

The proposed method of optimal fitting of the existing lower-class leveling control networks to the modernized national higher-class networks is carried out in two stages. In the first stage, the network is fitted to the maximum possible congruent reference points using robust least-squares adjustment methods. At this stage, the outlier reference points are detected. The reason why outliers appear is usually the vertical displacement of the reference points in the long time period between measurement of the existing network and the modernization of the reference network. In the second stage, the network is fitted to the detected outlier reference points by remeasuring the height differences around these points. A sequential optimal method of design and an online adjustment of the remeasured height differences is proposed. In this way, the potential displaced points of the existing network located around the outlying reference point are also detected and updated. Finally, the network is adjusted and fitted to all reference points of the modernized national control network.

The solar eclipse on August 21, 2017, passed through North America and considerably affected the electron density in the ionosphere. By taking advantage of a global navigation satellite system (GNSS) for monitoring the ionosphere, the ionospheric response to the eclipse and its impact on GNSS positioning were investigated using data from the Oregon Real-Time GNSS Network (ORGN). From the dual-frequency GNSS observation, total electron content (TEC) was extracted along the signal ray path between a satellite and a station for observing the abnormality of the ionosphere. From the TEC observations during three sequential days—the day before, the day of, and the day after the eclipse—the rate of change of TEC (ROT) was derived to analyze the impact of solar eclipse on the ionospheric response. A distinctive reduction of the ROT was found on the day of eclipse. Moreover, the impact of the eclipse on GNSS positioning was examined by conducting an experiment of the baseline processing. Because of the unusual dynamics of the ionosphere during the eclipse, the GNSS positioning error during the event was notably increased. This experiment shows that the eclipse increased the ionospheric gradient between a station in the path and a station outside of the path of totality that significantly degraded the positioning performance.