Evaluation and Tuning of Mascon Approach for Temporal Gravity Field Recovery

Satellite gravity missions are unique observation systems to directly observe mass transport processes in the Earth system. This study investigates innovative approaches for resolving Earth's gravitational field using mass concentration models (mascons) as an alternative to spherical harmonic synthesis. The primary objective is to systematically evaluate strategies and concepts to optimize temporal and spatial gravity field recovery.

This contribution investigates the theoretical foundations of mascon approaches, with detailed analysis of their mathematical and geometric properties. Functional and stochastic models are derived, focusing on three main approaches: Lumped Spherical Harmonics, point masses, and spherical cap mascons (primarily considered conceptually). This selection enables a comparative evaluation of different geometric structures, showcasing the advantages of discrete methods over continuous ones.

A key feature of this study is an automated simulation algorithm integrating spherical harmonic synthesis and mascon models. The algorithm executes a four-step process: orbit dynamics computation, observation simulation, noise handling, and adjustment. This enables global solution computation validated against a reference model, as well as precise subpixel analyses for targeted regional investigations. Results demonstrate the advantages of mascon approaches for regional resolution, including enhanced coastal densification to address leakage effects and detailed studies of regions like the Amazon basin.

Simulations based on a de-aliasing model and the observational geometry of a GRACE-like single-pair mission indicate that mascon approaches achieve near-complete signal recovery with high stability and accuracy. Deviations, expressed in millimeters of Equivalent Water Height (EWH), allow attribution to specific gravity field components. Global equiareal models exhibit minimal deviations (~15 mm), surpassing spherical harmonics in stability due to the combination of basis functions and regularization. Regional analyses reproduce structures with method-dependent deviations up to 40 mm, while highlighting challenges related to subpixel analyses. Incorporating additional grids and constraints to enhance signal recovery increases model complexity. The point mass approach proves especially robust, whereas Lumped Spherical Harmonics show quality declines in specific cases. The results meet current standards of conventional mascon models and reveal significant optimization potential. Future research should focus on refining regularization methods, expanding the data foundation, and integrating additional basis functions to further improve simulation accuracy and application outcomes.