Litcius/Paper detail

Multiframe Transformation With Variance Component Estimation

Yu Hu, Xing Fang, Wenxian Zeng, Hansjörg Kutterer

2023IEEE Transactions on Geoscience and Remote Sensing17 citationsDOI

Abstract

The modern GNSS technique is one of the most effective geoscience and remote-sensing tool to observe crustal motions and quantify plate tectonics dynamics. Given multiple installed continuously operating GNSS observing stations, the multi-frame transformation is implemented to connect the timevarying GNSS coordinates by the traditional step-wise method. Compared with the step-wise treatment of each pair of frames, the proposed structured total least-squares method considers the combined estimation for all frames, guaranteeing unique and consistent results for the multi-frame symmetric transformation. Furthermore, we introduce the variance component as the nutshell and flexible indicator for land movement. The variance components can quantify the movement coordinate-wise, regional-wise, or frame-wise if the variance components are estimable as we analyze. The simulated experiment shows that the multi-frame symmetric transformation is statistically superior to the traditional stepwise treatment. For the application, the deformation caused by Tohoku earthquake that happened in 2011 in northeast Japan is analyzed.

Topics & Concepts

GNSS applicationsComputer scienceTransformation (genetics)Variance (accounting)Frame (networking)Principal component analysisComponent (thermodynamics)Variance componentsCoordinate systemReference frameGeodesyRemote sensingAlgorithmGeologyGlobal Positioning SystemArtificial intelligenceMathematicsStatisticsTelecommunicationsGeneBiochemistryThermodynamicsBusinessChemistryAccountingPhysicsStatistical and numerical algorithmsGeophysics and Gravity MeasurementsSoil Geostatistics and Mapping
Multiframe Transformation With Variance Component Estimation | Litcius