JSG T.38

Terms of Reference

It is over 170 years since George Gabriel Stokes published his seminal formula for geoid determination using gravity anomalies. The formula was derived in spherical approximation and is valid under some well-known assumptions. Since then, geoid modelling has revolved more or less around handling these assumptions. As a result, there are now various geoid and quasigeoid computation methods of both types, i.e., methods with and without requiring the Stokes integration. However, despite this long-elapsed time, the determination of a cm-precise and/or cm-accurate geoid and quasigeoid remains an ongoing quest, although it has been achieved in a few studies.

With the computation of cm-precise and/or cm-accurate geoid, a supposition can be formed that solutions from different geoid modelling methods should converge within a given threshold, with an ideal threshold value being one-cm. The rationale is that, for a region, all the methods can be used to calculate the geoid using the same data and underlying theory. Still, methods differ primarily due to different handling of the data, and assumptions and approximations. Different methods provide different solutions due to many aspects including but not limited to: 1. different modifications of Stokes’s kernel, 2. different prediction/interpolation/extrapolation methods for non-Stokes integrating geoid modelling methods, 3. use of geodetic versus geocentric coordinates, 4. different Global Geopotential Models, 5. different gridding and merging techniques, 6. different parameter sweeps (integration radius and kernel modification degree), and 7. different handling of topography, atmosphere, spherical approximation, and downward continuation.

Given these possible sources for differences in geoid models, it becomes inevitable to first create a rigorous definition of a “cm-precise” and “cm-accurate” geoid followed by a comparative study of intermediate steps of different geoid modelling methods, in addition to comparing only the final results from different methods separately. Comparative study of intermediate steps is essential given the fact that if using the same datasets in different methods, it is expected to have geoid differences less than one cm when the methods are designed to take into account all effects greater than one cm.

Further, in view of cm-precise and/or cm-accurate geoid, it is important to compare multiple methods and parameter sweeps in different areas. This is because it would form an ideal strategy for a consistently precise/accurate geoid model. The difference between the precise geoid and consistently precise geoid is that the precision, in the latter, should be preserved when a geoid model is validated region-wise in addition to the validation with the complete ground truth. Otherwise, cm-precise geoid may have limited meaning.

Objectives

• Develop a statistical definition of cm-precise and cm-accurate geoid/quasi-geoid.
• Study and quantify the differences in handling the topography, atmosphere, ellipsoidal correction, and downward continuation in different geoid/quasigeoid modelling methods.
• Study, quantify and reduce the assumptions and approximations in different geoid modelling methods to attain congruency within some threshold.
• Study the requirement for merging various components/steps of different geoid modelling methods.
• Develop external validation techniques to determine region- or nationwide

Program of activities

• Presenting research findings at major international geodetic conferences, meetings, and workshops.
• Preparation of joint publications with JSG members.
• Organizing a session at the Hotine-Marussi Symposium 2026.
• Organizing splinter meetings at major international conferences and a series of online workshop.
• Supporting and cooperating with IAG commissions, services, and other study and working groups on gravity modelling and height systems.

Members

Ropesh Goyal (India); Chair
Sten Claessens (Australia); Vice-Chair
Ismael Foroughi (Canada)
Jonas Ågren (Sweden)
Xiaopeng Li (USA)
Bihter Erol (Turkey)
Jack McCubbine (Australia)
Pavel Novák (Czech Republic)
Koji Matsuo (Japan)
Riccardo Barzaghi (Italy)
Michal Šprlák (Czech Republic)
Jianliang Huang (Canada)
Yan-Ming Wang (USA)
Cheinway Hwang (China-Taipei)
Neda Darbeheshti (Australia)

Associate Members

Jack McCubbine (Australia)