Re-weighting of GPS baselines for vertical deformation analysis

H de Heus, M Martens, H van der Marel

Research output: Chapter in Book/Conference proceedings/Edited volumeChapterScientificpeer-review

Abstract

In the Netherlands, GPS is used to monitor subsidence caused by extraction of natural gas. GPS campaigns are organized on a yearly, basis. Each campaign consists of GPS observations 30 to 60 points, using typically 5 dual frequency GPS receivers. The duration of a session is between one and two hours, but as every point is observed several times, up to 56 sessions have been observed for each campaign. The area is about 75 by 60 km, resulting in an average baseline length of 10 km.

The GPS data is processed, using an iterative stepwise approach. First the GPS data is processed session by session. Then the sessions are combined, using their full covariance matrix in a network adjustment. The network adjustment is used to estimate the proper variance of the GPS observations. This involves removal of observation errors using statistical procedures, estimation of variance components, re-scaling and reprocessing of the GPS sessions. The best results are obtained using L1 and L2 phase data with an ionospheric constraint determined from the network adjustment, and using distinct weighting factors for the horizontal and vertical components.

The precision of the heights after the network adjustment is on the average 6 mm for individual campaigns. This has been confirmed by independent levellings.
Original languageEnglish
Title of host publicationGeodesy Beyond 2000
PublisherSpringer Science
Pages381-386
Number of pages6
Volume121
ISBN (Print)3-540-67002-5
DOIs
Publication statusPublished - 2000

Publication series

NameInternational Association Of Geodesy Symposia

Keywords

  • GPS
  • ionosphere
  • subsidence monitoring
  • variance component estimation

Fingerprint

Dive into the research topics of 'Re-weighting of GPS baselines for vertical deformation analysis'. Together they form a unique fingerprint.

Cite this