On global and regional spectral evaluation of global geopotential models

Ustun A., Abbak R. A.

JOURNAL OF GEOPHYSICS AND ENGINEERING, vol.7, no.4, pp.369-379, 2010 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 7 Issue: 4
  • Publication Date: 2010
  • Doi Number: 10.1088/1742-2132/7/4/003
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.369-379
  • Keywords: global geopotential models, terrestrial gravimetry, spectral evaluation, Turkish territory, CHAMP, GRACE, GRAVITY MODELS, GRIM5-S1
  • Kocaeli University Affiliated: No


Spectral evaluation of global geopotential models (GGMs) is necessary to recognize the behaviour of gravity signal and its error recorded in spherical harmonic coefficients and associated standard deviations. Results put forward in this wise explain the whole contribution of gravity data in different kinds that represent various sections of the gravity spectrum. This method is more informative than accuracy assessment methods, which use external data such as GPS-levelling. Comparative spectral evaluation for more than one model can be performed both in global and local sense using many spectral tools. The number of GGMs has grown with the increasing number of data collected by the dedicated satellite gravity missions, CHAMP, GRACE and GOCE. This fact makes it necessary to measure the differences between models and to monitor the improvements in the gravity field recovery. In this paper, some of the satellite-only and combined models are examined in different scales, globally and regionally, in order to observe the advances in the modelling of GGMs and their strengths at various expansion degrees for geodetic and geophysical applications. The validation of the published errors of model coefficients is a part of this evaluation. All spectral tools explicitly reveal the superiority of the GRACE-based models when compared against the models that comprise the conventional satellite tracking data. The disagreement between models is large in local/regional areas if data sets are different, as seen from the example of the Turkish territory.