Statistical analysis of astro-geodetic data through principal component analysis, linear modelling and bootstrap based inference

TL;DR

This paper applies advanced statistical methods including PCA, linear modeling, and bootstrap inference to analyze astro-geodetic data, revealing environmental factors influencing vertical deviation measurements.

Contribution

It introduces a combined statistical approach using PCA, linear models, and bootstrap techniques to analyze correlated astro-geodetic data effectively.

Findings

Pressure, temperature, and humidity influence vertical deviation measurements.

Bootstrap methods improve inference accuracy in correlated data.

Real data application demonstrates methodology effectiveness.

Abstract

The paper demonstrates the application of statistical based methodology for the analysis of the vertical deviation angle. The studied data set contains astro-geodetic observations. The Principal Component Analysis and the Multiple Linear Regression models are embedded within a bootstrap procedure, in order to overcome the difficulties related to data correlation, while taking advantage of all the information provided. The methodology is applied on real data. The obtained results indicate that the pressure, the temperature and the humidity are variables that may influence the measure of the vertical deviation.