Integration of satellite geodetic observations for regional geoid modeling using remove‑compute‑restore technique

Thumbnail Image
File version

Accepted Manuscript (AM)

Kalu, Ikechukwu
Ndehedehe, Christopher
Okwuashi, Onuwa
Eyoh, Aniekan
Griffith University Author(s)
Primary Supervisor
Other Supervisors
File type(s)

Advanced satellite geodetic systems have contributed to improving knowledge on changes in global gravity fields in recent years. These systems offer profound opportunities in the determination of a high-precision, high-resolution gravimetric geoid models in data deficient regions. However, due to the absence of research competence in some regions, these conventional datasets have not been implemented to update the obsolete reference frames which have been in use in these regions. This study introduces the well-known Remove-Compute-Restore (RCR) technique in modeling a gravimetric geoid model for a large data deficient region in West Africa (Nigeria) using two sets of long and short wavelength data (a) EGM2008 (long) + Airborne gravimetric observation (AGO) dataset (short) (b) EGM2008 (long) + Terrestrial gravimetric undulation (TGU) dataset (short). This therefore resulted in obtaining two sets of resultant RCR-gravimetric details whose statistics comparative assessments are produced in the body of this work. The RCR-determined gravimetric geoid model showed a strong relationship with the observed terrestrial data. The goodness of fit for the orthometric height correlation between the computed gravimetric geoid and the terrestrial data is 97.87 %. Strong relationships between the computed model and the other height models determined from the primary data are also observed. The difference between the Flury and rummel geoid-based model and the Heiskanen and moritz model in the geoid-quasi geoid separation phase shows the importance of implementing the ‘rigorous modeling technique’ for geoid quasi-geoid separation instead of the approximation method. However, the discrepancy between both models can be overlooked in a case of low accuracy geoidal operations. The RCR-based Fast Fourier Transform (FFT), and the least squares modified stokes (LSMS) was used in the quasi-geoid determination of the study region. Using the 1-parameter fit, the modified Stokes integral (FFT(b)) outperformed the LSMS model with a root mean square (rms) difference of 0.6 cm, while the LSMS outperformed both the unmodified (FFT(a)) and the modified (FFT(b)) Stokes integral using the 4-parameter fit with a rms difference of 0.5 cm. The use of both models in this study created a better perception of the quasi-geoid differential analysis, and a more refined understanding of their interaction with the satellite based quasi geoid.

Journal Title

Earth Science Informatics

Conference Title
Book Title
Thesis Type
Degree Program
Publisher link
Patent number
Grant identifier(s)
Rights Statement
Rights Statement

© 2021 Springer. This is an electronic version of an article published in Earth Science Informatics, 2021. Earth Science Informatics is available online at: with the open URL of your article.

Item Access Status

This publication has been entered in Griffith Research Online as an advanced online version.

Access the data
Related item(s)

Global and planetary environmental engineering

Geomatic engineering

Photogrammetry and remote sensing

Geomatic engineering not elsewhere classified

Spatial data and applications

Computational statistics

Persistent link to this record

Kalu, I; Ndehedehe, C; Okwuashi, O; Eyoh, A, Integration of satellite geodetic observations for regional geoid modeling using remove‑compute‑restore technique, Earth Science Informatics, 2021