GRACE Fast Mascons From Spherical Harmonics and a Regularization Design Trade Study.

Mass concentration (mascon) solutions have become a prominent medium for investigating time‐variable gravity recovered by GRACE and GRACE Follow‐On. While GRACE Level‐2 spherical harmonic products require various postprocessing techniques to eliminate correlated noise, mascon formulations employ spatial regularization strategies during the estimation step to improve signal recovery. However, mascon estimation has traditionally required large computing resources and GRACE Level‐1B processing capabilities. In this study, we show that a typical mascon estimation system can be reformulated to allow for the estimation of regularized mascons from Level‐2 spherical harmonics. Provided that spherical harmonic solution covariances are available, the computed mascons will be mathematically equivalent to similar mascons estimated from Level‐1B observations. This method is computationally efficient, better leverages GRACE spherical harmonics than past methods, and matches the performance of typical mascon solutions without locking scientists into predetermined regularization designs. We develop a proof‐of‐concept solution using ITSG‐Grace2018 and compare results with traditional mascons from the Jet Propulsion Lab (JPL) and NASA Goddard Space Flight Center (GSFC). We then assess the effects of spherical harmonic truncation and use of regularization correlations on basin signal recovery. We find that spherical expansions to degree and order 60 provide the minimum expansion necessary to study most basins, while larger expansions help further localize signals. We also find that diagonal regularizations (i.e., regularizations that do not contain inter‐mascon correlations) are adversely affected by leakage, especially across boundaries such as coastlines where signals are not highly correlated, whereas including inter‐mascon correlations and regional boundaries in the regularization greatly improves signal recovery. Plain Language Summary: This study derives a new method for producing surface mass change maps in the form of so‐called "mascons" using GRACE and GRACE Follow‐On data that is computationally simpler but equally as rigorous as typically used methods. Using this method, scientists no longer need specialized orbit determination and estimation software to produce these maps, and therefore this type of work is now accessible to scientists outside of the few major centers that currently produce these maps. At the same time, centers with these specialized software packages can incorporate the computational efficiencies presented here to improve their own processes. As a result, this new method can be used to study individual problems of interest outside the scope of the globally focused solutions available today (for example, to specifically study targeted river basins), and these new studies can additionally incorporate other known information about those problems. We provide a proof‐of‐concept solution showcasing the new method and apply the new method to highlight key design decisions that must be made when producing mascon solutions. Key Points: Computationally efficient mascons computed from GRACE Level‐2 spherical harmonics are equivalent to those from Level‐1B ranging dataThese mascons can be tailored to specific problems of interest and do not require special orbit determination softwareA presented trade study shows the effects of spherical expansion truncation and compares diagonal regularization to that with correlations [ABSTRACT FROM AUTHOR]