Abstract
Adjustment has been based on the assumption that random errors of measurements are added to functional models. In geodetic practice, we know that accuracy formulae of modern geodetic measurements often consist of two parts: one proportional to the measured quantity and the other constant. From the statistical point of view, such measurements are of mixed multiplicative and additive random errors. However, almost no adjustment has been developed to strictly address geodetic data contaminated by mixed multiplicative and additive random errors from the statistical point of view. We systematically develop adjustment methods for geodetic data contaminated with multiplicative and additive errors. More precisely, we discuss the ordinary least squares (LS) and weighted LS methods and extend the bias-corrected weighted LS method of Xu and Shimada (Commun Stat B29:83–96, 2000) to the case of mixed multiplicative and additive random errors. The first order approximation of accuracy for all these three methods is derived. We derive the biases of weighted LS estimates. The three methods are then demonstrated and compared with a synthetic example of surface interpolation. The bias-corrected weighted LS estimate is unbiased up to the second order approximation and is of the best accuracy. Although the LS method can warrant an unbiased estimate for a linear model with multiplicative and additive errors, it is less accurate and always produces a very poor estimate of the variance of unit weight.
Similar content being viewed by others
References
Akima H (1996) Algorithm 760: rectangular-grid-data surface fitting that has the accuracy of a bicubic polynomial. ACM Trans Math Softw 22:357–361
Berg R, Ferguson J (2000) A practical evaluation of airborne laser mapping for highway engineering surveys. In: Proceedings of the ION GPS 2000, Salt Lake city, Utah
Crowder M (1995) On the use of a working correlation matrix in using generalised linear models for repeated measures. Biometrika 82:407–410
Dennis JE Jr, Schnabel RB (1996) Numerical methods for unconstrained optimization and nonlinear equations. In: SIAM Classics in Applied Mathematics, Philadelphia
Desmond AF (1997) Optimal estimating functions, quasi-likelihood and statistical modelling. J Stat Plan Inference 60:77–123
Detweiler ZR, Ferris JB (2010) Interpolation methods for high-fidelity three-dimensional terrain surfaces. J Terramech 47:209–217
Ewing CE, Mitchell MM (1970) Introduction to geodesy. Elsevier, New York
Firth D, Harris IR (1991) Quasi-likelihood for multiplicative random effects. Biometrika 78:545–555
Fitzmauric GM (1995) A caveat concerning independence estimating equations with multivariate binary data. Biometrics 51:309–317
Flamant PH, Menzies RT, Kavaya MJ (1984) Evidence for speckle effects on pulsed \(CO_2\) lidar signal returns from remote targets. Appl Optics 23:1412–1417
Goodman JW (1976) Some fundamental properties of speckle. J Opt Soc Am 66:1145–1150
Heyde CC (1997) Quasi-likelihood and its applications. Springer, New York
Hill AC, Harris M, Ridley KC, Jakeman E, Lutzmann P (2003) Lidar frequency modulation vibrometry in the presence of speckle. Appl Optics 42:1091–1100
Kidner DB (2003) Higher-order interpolation of regular grid digital elevation models. Int J Rem Sens 24:2981–2987
Kobler A, Pfeifer N, Ogrinc P, Todorovski L, Oštir K, Džeroski S (2007) Repetitive interpolation: a robust algorithm for DTM generation from Aerial Laser Scanner Data in forested terrain. Rem Sens Environ 108:9–23
Kraus K, Pfeifer N (1998) Determination of terrain models in wooded areas with airborne laser scanner data. ISPRS J Photogram Rem Sens 53:193–203
Kukusha A, Malenkoa A, Schneeweissb H (2010) Optimality of quasi-score in the multivariate mean-variance model with an application to the zero-inflated Poisson model with measurement errors. Statistics 44:381–396
Leigh CL, Kidner DB, Thomas MC (2009) The use of LiDAR in digital surface modelling: issues and errors. Trans GIS 13:345–361
López-Martínez C, Fàbregas X, Pipia L (2011) Forest parameter estimation in the Pol-InSAR context employing the multiplicative-additive speckle noise model. ISPRS J Photogram Rem Sens 66:597–607
MacDoran PF (1979) Satellite emission radio interferometric earth surveying series—GPS geodetic system. Bull Géod 53:117–138
Magnus JR, Neudecker H (1988) Matrix differential calculus with applications in statistics and econometrics. Wiley, New York
McCullagh P (1983) Quasi-likelihood functions. Ann Stat 11:59–67
McCullagh P, Nelder J (1989) Generalized linear models, 2nd edn. Chapman and Hall, London
Olive DJ (2012) A course in statistical theory. www.math.siu.edu/olive/irun.pdf
Oliver CJ (1991) Information from SAR images. J Phys D Appl Phys 24:1493–1514
Prentice RL, Zhao LP (1991) Estimating functions for parameters in means and covariances of multivariate discrete and continuous responses. Biometrics 47:825–839
Rao CR, Kleffe J (1988) Estimation of variance components and applications. Elsevier, North-Holland
Searle SR (1971) Linear models. Wiley, New York
Seeber G (2003) Satellite geodesy, 2nd edn. de Gruyter, Berlin
Seber G, Wild C (1989) Nonlinear regression. Wiley, New York
Tzavelas G (1998) A note on the uniqueness of the quasi-likelihood estimator. Stat Prob Lett 38:125–130
Ulaby F, Kouyate F, Brisco B, Williams T (1986) Textural information in SAR images. IEEE Trans Geosci Rem Sens 24:235–245
Veneziano D, Hallmark S, Souleyrette R (2004) Accuracy of light detection and ranging derived terrain data for highway location. Comput Aided Civil Infrastrust Eng 19:130–143
Wang JY, Pruitt PA (1992) Effects of speckle on the range precision of a scanning lidar. Appl Optics 31:801–808
Wedderburn R (1974) Quasi-likelihood functions, generalized linear models, and the Gauss-Newton method. Biometrika 61:439–447
Xu PL (1999) Despeckling SAR-type multiplicative noise. Int J Rem Sens 20:2577–2596
Xu PL, Shimada S (2000) Least squares estimation in multiplicative noise models. Commun Stat B29:83–96
Zimmerman D, Pavlik C, Ruggles A, Armstrong MP (1999) An experimental comparison of ordinary and universal kriging and inverse distance weighting. Math Geol 31:375–390
Acknowledgments
This work is partially supported by the National Foundation of Natural Science of China (No. 41204006) and a Grant-in-aid for Scientific Research (C25400449). The authors thank, with gratitude, the three reviewers for their very constructive and very detailed comments, which lead to a significant improvement and clarification of the paper.
Author information
Authors and Affiliations
Corresponding author
Appendix: Proof of Lemma 1
Appendix: Proof of Lemma 1
The proof of Lemma 1 consists of two steps. The first step is to use the first two terms on the left hand side of (28) to derive (29) of Lemma 1, and the second step is to prove:
Without loss of generality, let us first assume that (56) holds true. As a result, applying the mathematical expectation operation to (28) yields
or equivalently,
where \(\mathbf{b}(\hat{\varvec{\beta }})\) denotes \(E(\mathbf{b}_{\beta })\) and stands for the bias of the WLS estimate of \(\varvec{\beta }\). The term on the right hand side of (57), i.e., \(E(\mathbf{G}_2\varvec{\epsilon }_y)\), can be computed as follows:
because \(E( \varvec{\epsilon }_y\varvec{\epsilon }_y^\mathrm{T})=\varvec{\Sigma }_y\). By inserting (58) into (57), we can readily obtain
which is exactly (29) of Lemma 1.
Now we have to prove that the second step, i.e., (56), is indeed true. The proof of (56) can be further simplified to prove:
because the same method of proof can be used to prove
Actually, (60) can be rewritten as follows:
where \(\mathbf{e}_i\) is the ith row vector of the natural basis of dimension \(n\). Because the left and right hand sides of (61) have the same coefficient matrix \(\mathbf{A}^\mathrm{T}\varvec{\Sigma }_y^{-1} \mathbf{D}_{a\beta }\varvec{\Sigma }_m\), the proof of (61) can be further simplified to prove
In fact, for the ith element on the left side of (62), we have
In the similar manner, for the ith element on the right side of (62), we have
By comparing (63) with (64), we immediately conclude that (62) is true. As a result, we complete the proof of Lemma 1.
Rights and permissions
About this article
Cite this article
Xu, P., Shi, Y., Peng, J. et al. Adjustment of geodetic measurements with mixed multiplicative and additive random errors. J Geod 87, 629–643 (2013). https://doi.org/10.1007/s00190-013-0635-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00190-013-0635-2