Dear Colleagues,
We are pleased to announce our second seminar for our series, hosted by the International Association of Geodesy (IAG) Working Group 3.1: Hydrologic Signature in Geodetic Observations, given by Wenjin Chen, Assistant Professor at Jiangxi University of Science and Technology, China. This seminar will be held on April 1 at 10:00 EDT/ 7:00 PDT/ 16:00 CET/ 23:00 CST.
Join the webinar:
https://iu.zoom.us/j/82304710308
Meeting ID: 823 0471 0308
Title: Algorithms and Software for efficient Gravity forward modeling of hydrological mass changes
Abstract: Gravimetric forward modeling is a fundamental tool in physical geodesy and gravimetric geophysics to model the gravitational potential and its derivatives from a prescribed subsurface density distribution. This paper presents a systematic review and unified formulation of forward modeling algorithms in Cartesian and spherical coordinate systems, defined in both spatial and spectral domains. In the Cartesian coordinate framework, analytical and numerical expressions are provided for a simple representation of volumetric mass elements that include point mass, vertical line mass, horizontal mass layer, and prism as well as for more complex and generalized representations involving volumetric mass bodies bounded by variable upper and lower interfaces and volumetric mass bodies with arbitrary geometries and heterogeneous density distribution. In the spherical coordinates, expressions are given for prism and tesseroid together with two spectral domain approaches involving tesseroid modeling combined with spherical harmonic analysis and concentric-shell method suited for arbitrary mass geometries. Numerical experiments, based on using digital terrain model of Tibet and Himalaya, demonstrate that compared to prism method, simple point and line mass approaches reproduce only the long-wavelength gravity field, with gravity differences typically up to about 12 mGal and gravity gradient discrepancies about 0.45 E, whereas the horizontal layer method agrees with the prism gravity solution within ±0.01 mGal and approaches almost zero differences for gravity gradient solutions. Models with variable interfaces or full arbitrary geometries can deviate by 8 mGal in regions with extremely complex topography, underscoring necessity of realistic geometric representation in high resolution applications. Spectral domain implementations in both Cartesian and spherical coordinates substantially reduce computational time when compared with direct spatial integration techniques, while preserving numerical accuracy that is effectively indistinguishable from spatial domain algorithms at regional scales. These results highlight the trade-offs between geometric realism, accuracy, and computational efficiency, and provide practical guidance for selecting suitable forward modeling strategies in numerical studies. Finally, we present several examples of hydrological mass changes derived from gravity forward modeling.
