Precise and Efficient Evaluation of Gravimetric Quantities at Arbitrarily Scattered Points in Space

Gravimetric quantities are commonly represented in terms of high degree surface or solid spherical harmonics. After EGM2008, such expansions routinely extend to spherical harmonic degree 2190, which makes the computation of gravimetric quantities at a large number of arbitrarily scattered points in space using harmonic synthesis, a very computationally demanding process. We present here the development of an algorithm and its associated software for the efficient and precise evaluation of gravimetric quantities, represented in high degree solid spherical harmonics, at arbitrarily scattered points in the space exterior to the surface of the Earth. The new algorithm is based on representation of the quantities of interest in solid ellipsoidal harmonics and application of the tensor product trigonometric needlets. A FORTRAN implementation of this algorithm has been developed and extensively tested. The capabilities of the code are demonstrated using as examples the disturbing potential T , height anomaly ζ, gravity anomaly ∆g, gravity disturbance δg, northsouth deflection of the vertical ξ, east-west deflection of the vertical η, and the second radial derivative Trr of the disturbing potential. For the range from the surface of the Earth up to 544 kilometers above that surface the Approved for public release, 17-477. K. G. Ivanov Institute of Mathematics and Informatics, Bulgarian Academy of Sciences, 1113 Sofia, Bulgaria, E-mail: [email protected] N. K. Pavlis National Geospatial-Intelligence Agency (NGA), Research Directorate (R), Springfield, VA 22150, USA, E-mail: [email protected] P. Petrushev Department of Mathematics, University of South Carolina, Columbia, SC 29208, USA, E-mail: [email protected] current version of the software runs on a standard PC at speeds between 20,000 and 40,000 point evaluations per second, depending on the gravimetric quantity being evaluated, while the relative error does not exceed 10−6 and the memory (RAM) use is 9.3 GB.