Taking correlations into account: a diagonal correlation model
- 302 Downloads
The true covariance matrix of the GPS phase observations is unknown and has to be assumed or estimated. The variance of the least-squares residuals was empirically shown to have an elevation dependency and is often expressed as a sum of a constant and an exponential function. Disregarding correlations that are for instance due to atmospheric effects, the variance covariance matrices are diagonal. This simplification leads to errors in the estimates, including the float ambiguity vector, as well as to an overoptimistic precision. Thus, results of test statistics such as the outlier or the overall model test are impacted. For the particular case of GPS positioning, an innovative proposal was made to take correlations into account easily, condensed in an equivalent diagonal matrix. However, the a posteriori variance factor obtained with this simplification is strongly underestimated and in most cases the inversion of fully populated matrices has anyway to be carried out. In this contribution, we propose an alternative diagonal correlation model based on a simple exponential function to approximate the developed equivalent model. This way, correlations can be included in a diagonal variance covariance matrix without computation burden. A case study with an 80-km baseline where the ambiguities are estimated together with the coordinates in the least-squares adjustment demonstrates the potential of the model. It leads to a proposal based on the autocorrelation coefficient for fixing its parameters.
KeywordsVariance model Weighting GPS Least-squares Exponential model
The authors gratefully acknowledge the EPN and corresponding agencies for providing the data freely. Three reviewers are warmly acknowledged for their constructive comments.
- Collins JP, Langley RB (1999) Possible weighting schemes for GPS carrier phase observations in the presence of multipath. U. S. Army Corps Eng Topogr Eng Cent 1999:1–40Google Scholar
- Dach R, Hugentobler U, Fridez P, Meindl M (eds) (2007) Bernese GPS Software Version 5.0. Astronomical Institute, University of Bern, SwitzerlandGoogle Scholar
- Euler HJ, Schaffrin B (1991) On a measure for the discernibility between different ambiguity solutions in the static kinematic GPS mode. In: Schwarz KP, Lachapelle G (eds) Kinematic systems in geodesy. IAG symposia, vol 107. Springer, Berlin, pp 285–295Google Scholar
- Kermarrec G, Schön S, Kreinovich V (2017) Possible explanation of empirical values of the matern smoothness parameter for the temporal covariance of GPS measurements. Appl Math Sci 11(35):1733–1737Google Scholar
- Leick A, Rapoport L, Tatarnikov D (2015) GPS satellite surveying, 4th edn. Wiley, New YorkGoogle Scholar
- Luo X, Mayer M, Heck B (2012) Analysing time series of GNSS residuals by means of AR(I)MA processes. In: Sneeuw N, Novák P, Crespi M, Sansò F (eds) VII Hotine-Marussi Symposium on Mathematical Geodesy. International Association of Geodesy Symposia, vol 137. Springer, Berlin, pp 129–134Google Scholar
- Rao C, Toutenburg H (1999) Linear models, least-squares and alternatives, 2nd edn. Springer, New YorkGoogle Scholar
- Talbot N (1988). Optimal weighting of GPS carrier phase observations based on the signal-to-noise ratio. In: Proceedings of international symposia, global positioning systems. Gold Coast, Queensland, pp 4.1–4.17, 17–19 OctGoogle Scholar
- Teunissen PJG (2000) Testing theory and introduction. Series on mathematical geodesy and positioning. Delft University Press, DelftGoogle Scholar
- Wei M, Schwarz KP (1995) Fast ambiguity resolution using an integer nonlinear programming method. In: ION GPS-1995, Institute of Navigation, Palm Spring, CA, USA, pp 1101–1110, Sep 1995Google Scholar
- Williams MN, Gomez Grajales CA, Kurkiewicz D (2013) Assumptions of multiple regression: correcting two missconceptions. Pract Assess Res Eval 18(11):2Google Scholar