Efficacy of M_split Estimation in Displacement Analysis

Area of Science:

• Geodesy and Geophysics
• Geodetic observation analysis
• Deformation analysis

Background:

• Geodetic observation sets often contain data with differing functional models or parameter values.
• Conventional methods like least squares estimation require pre-separation of observation subsets.
• Analyzing observations from various measurement epochs presents challenges for traditional approaches.

Purpose of the Study:

• To evaluate the efficacy of Msplit estimation for detecting point displacements in geodetic data.
• To compare Msplit estimation with classical deformation analysis methods.
• To investigate the performance of Msplit estimation on disordered observation sets.

Main Methods:

• Application of Msplit estimation, which assigns each observation to competitive functional models automatically.
• Utilizing example geodetic observation sets.
• Conducting Monte Carlo simulations to assess performance.
• Comparing results with classical deformation analysis.

Main Results:

• Msplit estimation demonstrates efficacy in detecting point displacements.
• The method shows promise as an alternative to conventional geodetic data analysis techniques.
• Msplit estimation exhibits a distinct advantage over conventional approaches for disordered observation sets.

Conclusions:

• Msplit estimation is a viable and potentially advantageous alternative for deformation analysis in geodesy.
• The automatic assignment of observations to functional models simplifies the analysis process.
• The method's effectiveness is particularly notable in scenarios with disordered observational data.