Optimal Interpolation of Spatially Discretized Geodetic Data

Zheng-Kang Shen, Min Wang, Yuehua Zeng, & Fan Wang

Published July 30, 2015, SCEC Contribution #2077

We present an algorithm to calculate horizontal strains (or strain rates) through interpolation of geodetically derived displacements (or velocities). This is an under-determined inverse problem to derive smoothly distributed strains (or strain rates) using spatially discretized geodetic observations. A priori information, in the form of weighted smoothing, is critical to facilitate the solution. At a given site, the horizontal displacement (or velocity) field in its vicinity is approximated by a bi-linear function and represented by rigid block translation, rotation, and strains (or their rates). The weighted displacement (or velocity) data in the neighborhood are used to estimate the field parameters through a least-squares inversion procedure. Optimal weightings are prescribed for the neighboring data, based on their distances to the interpolation site and their spatial coverage. Non-elastic strains resulted from surface fault rupture and creep may also be excluded from the solution.

We apply this method to the Southern California Earthquake Center (SCEC) Crustal Motion Map Version 4.0 (CMM4) velocity field and derive the strain rate field in southern California. Our result shows that: (1) Distance dependent weighting can be optimally achieved by employing either a Gaussian or quadratic decay function, with the former offering a slightly sharper result than the latter. (2) Spatially dependent weighting is important to improve the interpolation, and can be done by invoking either an azimuthal weighting or a Voronoi cell areal weighting function. The result shows detailed strain rate pattern in southern California, which is dominated by dextral shear of the San Andreas fault (SAF) system. The secondary faults surrounding the “Big Bend” of the SAF strike at oblique angles with respect to the maximum shear direction, suggesting that tectonic deformation field on and off the SAF is dominated by mechanic processes of the SAF.

Citation
Shen, Z., Wang, M., Zeng, Y., & Wang, F. (2015). Optimal Interpolation of Spatially Discretized Geodetic Data. Bulletin of the Seismological Society of America, 105(4), 2117-2127. doi: 10.1785/0120140247. http://www.bssaonline.org/content/early/2015/07/10/0120140247.abstract

Related Projects & Working Groups
Tectonic Geodesy