Network Adjustment Issues Using Mixed GPS Surveying Techniques

Modern GPS carrier phase-based survey techniques such as "rapid static" and "stop & go" are fast becoming the standard operational modes for building local geodetic control networks or for the densification of an existing network. However, such techniques do have some challenges when used together within a single survey project. For example, these mixed-mode GPS surveys do result in incompatible (and over-optimistic) variance-covariance matrices, are subject to correlations between baselines when two or more base GPS receivers are used, and can result in biased baseline vectors due to faulty ambiguity resolution compared with 'traditional' GPS static surveying techniques (based on the analysis of observation sessions of an hour or more in length). In this paper an example is presented which shows the impact on coordinate results if the standard network adjustment procedure is used. A new network adjustment procedure is proposed, which includes: (1) Processing all possible combinations of pairs of receivers as single baseline solutions. (2) Scaling all variance-covariance matrices in a session by a factor that is a function of the number of GPS receivers that have been used, the observation data sampling rate and the length of the observation session. (3) At the network adjustment step requiring that all baselines, with their apriori variancecovariance matrices, be treated as new (baseline vector) observations, and the least squares adjustment performed with minimal constraints. A methodology based on "reliability analysis", for detecting baseline vector solution biases caused by wrong ambiguity resolution, is proposed. On the basis of this study some practical suggestions for designing mixed-mode GPS survey networks are made.