Journal of Surveying Engineering

We may obtain a unique parameter vector solution from a combination of individually estimated parameter vectors by adjusting different observation groups. This kind of combination method, the so-called step-by-step procedure, is widely used in geodesy. We derive Helmert-type variance-component estimates for several observation groups via this step-by-step approach. In the estimation step, only individual results are used; hence, this procedure is called step-by-step estimation of variance components. The novel approach is applied to a horizontal control network and a leveling network.

Deformation parameters are estimated based on a time series of monitoring campaigns. For each monitoring campaign, the measurements’ contribution is in determining the relative positions of the network points and defining the datum of the network. When part of the datum definition contained in the measurements is not constant in time, the estimated deformation parameters could be erroneous. Measured distance, for example, can be assumed to have scale. But fluctuations in the instrument oscillator, when using the same instrument for all campaigns or when using a different instrument in each campaign, could affect the scale between monitoring epochs. For a single monitoring campaign, datum parameters contained in the distance measurements may be used for estimating the coordinates of the network points. However, in four-dimensional (4D) geodetic networks, a part of the datum definition determined by the geodetic measurements may not remain consistent. For example, the estimated deformation parameters related to scale could be wrong in distance-based networks. In this study, the measurements from each campaign are stripped from their datum content by extended free network adjustment constraints. The datumless measurements are used to define the datum by preliminary coordinates and linear constraints that remain constant for all monitoring campaigns, as well as to define the position of the network points. Subsequently, the variations in the network geometry can be modeled by means of a physical model. Following an introduction of the concept of extended free network adjustment constraints and a development of theoretical tools, the paper presents the deformation analysis of a small electronic distance measurement (EDM) network measured six times between 1989 and 2008 using the proposed method.

Accurate GPS positioning requires correct modeling of tropospheric refraction, the phenomenon responsible for tropospheric delay (TD). The Niell mapping function, which is used in almost all processing packages to model TD, is being replaced by improved models. This study analyzes the influence of TD models and absolute antenna phase center variation (PCV) models. Solutions computed with the Niell mapping function and global mapping function TD models were compared at several European International Global Navigation Satellite Systems (GNSS) Service (IGS) stations. GPS sessions of 24 hours, observed on five different days in 2008, were used for this investigation. Two independent network-station constraint approaches are considered: a GPS network fixing a single station was implemented and was followed by a second network solution in which several IGS stations were constrained. The elevation-cutoff angle was set to 0° to obtain the maximum influence of the TD model at altitudes close to the horizon. Results of the final solutions depend more on the constrained stations within the study area than on the type of tropospheric model or antenna type.

The idea of integrated measurements, combining classic field observations and satellite observations, has recently been applied in geodetic work with increasing frequency. To draw an entire set of such observations numerically, it is necessary to reduce them to a common mathematical space (e.g., onto the surface of an ellipsoid WGS84). This paper presents a proposed adjustment of GPS observations in a local reference system without the necessity of transforming and mapping them. The adjustment process is carried out after a GPS vector network is transformed into a spatial angular-distance network. This allows an easy integration with classical observations (joint adjustment of all observations). Detailed formulas are provided and used to develop a mathematical adjustment model. The computational algorithm is verified in practice.

In accordance with the advancement of technological development, access to all types of digital data is becoming increasingly important. Various problems are commonly faced when obtaining digital data from existing maps. These include errors caused by the imperfection of the digitizing device, deformations in paper maps, errors in matching the data along the edges of adjacent paper maps, and systematic errors caused by the operator. Although the data obtained from photogrammetric methods are affected by pointing errors by the operator and errors occurring when analog photos are scanned, these errors are independent of the deformations that occur in paper maps. This study compared the raster maps produced from aerial photos with the data acquired from digitized vector maps. The statistical distributions of the discrepancies obtained as a result of this comparison were analyzed. The data from the scanned map and photogrammetric map were found to be consistent with each other, but the scanned map and newly produced map were not consistent with each other.