Comparison of collocation and kriging statistical methods in the interpolation of InSAR time series displacement fields
|
Hamid Mehrabi * , Mina Nadian , Fateme DallalZade Atoofi |
University of Isfahan |
|
Abstract: (1292 Views) |
This article compares two statistical interpolation methods, least squares collocation (LSC) and kriging, for analyzing time series radar images based on permanent scatterers to investigate subsidence in Isfahan City. The study processed 73 images from Sentinel 1-A satellite between 10/19/2014 and 12/10/2020. The LSC method used Hirvanen, Gaussian, and exponential models of analytical covariance functions with RMSE values of 4.52, 5.31, and 4.28 respectively. The Kriging method used spherical analytical variogram functions models in addition to segmental, exponential, Gaussian, and combined effects with RMSE values of 4.35, 4.31, 9.70, and 4.40 respectively. Due to the lower RMSE value for the exponential model function in the Kriging method, ordinary kriging and ordinary Co-kriging methods were used for further interpolation using this model. The results showed that both LSC and Kriging methods can be effectively used for interpolating permanent scatterer data to investigate subsidence in urban areas such as Isfahan City; however, the exponential model function was found to be more effective when using Kriging interpolation methods than other models tested in this study due to its lower RMSE value for this dataset.
|
|
Keywords: Kriging interpolation method, least squares collocation method, Synthetic aperture radar interferometry, statistical interpolation methods, InSAR permanent scatterers |
|
Full-Text [PDF 1950 kb]
(343 Downloads)
|
Type of Study: Research |
Subject:
Geodesy Received: 2023/04/26 | Accepted: 2023/09/12 | ePublished ahead of print: 2023/10/10 | Published: 2023/10/10
|
|
|
|
|
Send email to the article author |
|