Gradient-boosted equivalent sources

The equivalent source technique is a powerful and widely used method for processing gravity and magnetic data. Nevertheless, its major drawback is the large computational cost in terms of processing time and computer memory. We present two techniques for reducing the computational cost of equivalent source processing: block-averaging source locations and the gradient-boosted equivalent source algorithm. Through block-averaging, we reduce the number of source coefficients that must be estimated while retaining the minimum desired resolution in the final processed data. With the gradient-boosting method, we estimate the sources coefficients in small batches along overlapping windows, allowing us to reduce the computer memory requirements arbitrarily to conform to the constraints of the available hardware. We show that the combination of block-averaging and gradient-boosted equivalent sources is capable of producing accurate interpolations through tests against synthetic data. Moreover, we demonstrate the feasibility of our method by gridding a gravity data set covering Australia with over 1.7 million observations using a modest personal computer.

[1]  Walter H. F. Smith,et al.  Gridding with continuous curvature splines in tension , 1990 .

[2]  F. Guspí,et al.  Reduction to the pole and transformations of scattered magnetic data using Newtonian equivalent sources , 2009 .

[3]  John D. Hunter,et al.  Matplotlib: A 2D Graphics Environment , 2007, Computing in Science & Engineering.

[4]  W. Menke.  Geophysical data analysis : discrete inverse theory , 1984 .

[5]  Vanderlei C. Oliveira,et al.  Modeling the Earth with Fatiando a Terra , 2013, SciPy.

[6]  V. Barbosa,et al.  Fast iterative equivalent-layer technique for gravity data processing: A method grounded on excess mass constraint , 2017 .

[7]  João B. C. Silva.  Reduction to the pole as an inverse problem and its application to low-latitude anomalies , 1986 .

[8]  P. Wessel,et al.  PyGMT: A Python interface for the Generic Mapping Tools , 2021 .

[9]  R. O. Hansen.  Interpretive gridding by anisotropic kriging , 1993 .

[10]  M. Fuchs,et al.  Satellite gravity gradient grids for geophysics , 2016, Scientific Reports.

[11]  P. Argentiero,et al.  On least squares collocation , 1978 .

[12]  Gary Barnes,et al.  Processing gravity gradient data , 2011 .

[13]  A fast equivalent source method for airborne gravity gradient data , 2019, GEOPHYSICS.

[14]  V. Barbosa,et al.  Convolutional equivalent layer for gravity data processing , 2020 .

[15]  Cericia Martinez,et al.  Denoising of gravity gradient data using an equivalent source technique , 2016 .

[16]  L. Uieda,et al.  Polynomial equivalent layer , 2012 .

[17]  T. Fukushima.  Speed and accuracy improvements in standard algorithm for prismatic gravitational field , 2020, Geophysical Journal International.

[18]  D. Nagy,et al.  Corrections to “The gravitational potential and its derivatives for the prism” , 2002 .

[19]  Anderson Banihirwe,et al.  Pooch: A friend to fetch your data files , 2020, J. Open Source Softw..

[20]  William J. Hinze,et al.  Spherical earth gravity and magnetic anomaly analysis by equivalent point source inversion , 1981 .

[21]  Carlos Alberto Mendonça,et al.  The equivalent data concept applied to the interpolation of potential field data , 1994 .

[22]  J. Friedman.  Greedy function approximation: A gradient boosting machine. , 2001 .

[23]  Yi Zhang,et al.  Transforming Total‐Field Magnetic Anomalies Into Three Components Using Dual‐Layer Equivalent Sources , 2020, Geophysical Research Letters.

[24]  Nils Olsen,et al.  An equivalent source method for modelling the global lithospheric magnetic field , 2014 .

[25]  Xiong Li,et al.  Ellipsoid, geoid, gravity, geodesy, and geophysics , 2001 .

[26]  Lindrith Cordell,et al.  A scattered equivalent-source method for interpolation and gridding of potential-field data in three dimensions , 1992 .

[27]  Douglas W. Oldenburg,et al.  Rapid construction of equivalent sources using wavelets , 2010 .

[28]  J. Friedman.  Stochastic gradient boosting , 2002 .

[29]  et al.,et al.  Jupyter Notebooks - a publishing format for reproducible computational workflows , 2016, ELPUB.

[30]  C. N. G. Dampney,et al.  The equivalent source technique , 1969 .

[31]  K. Jarrod Millman,et al.  Array programming with NumPy , 2020, Nat..

[32]  Gábor Papp,et al.  The gravitational potential and its derivatives for the prism , 2000 .

[33]  Reduction of Magnetic Anomaly Observations from Helicopter Surveys at Varying Elevations , 2006 .

[34]  Stephan Hoyer,et al.  xarray: N-D labeled arrays and datasets in Python , 2017 .

[35]  P. Wessel,et al.  Global Bathymetry and Topography at 15 Arc Sec: SRTM15+ , 2019, Earth and Space Science.

[36]  João B. C. Silva,et al.  Discrete linear transformations of potential field data , 1989 .

[37]  D. Sandwell.  BIHARMONIC SPLINE INTERPOLATION OF GEOS-3 AND SEASAT ALTIMETER DATA , 1987 .

[38]  Gravity‐enhanced representation of measured geoid undulations using equivalent sources , 2004 .

[39]  D. A. Emilia.  Equivalent Sources Used As An Analytic Base For Processing Total Magnetic Field Profiles , 1973 .

[40]  A. N. Tikhonov,et al.  Solutions of ill-posed problems , 1977 .

[41]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[42]  William J. Hinze,et al.  Improved inversion of geopotential field anomalies for lithospheric investigations , 1988 .

[43]  Siu Kwan Lam,et al.  Numba: a LLVM-based Python JIT compiler , 2015, LLVM '15.

[44]  Leonardo Uieda.  Verde: Processing and gridding spatial data using Green's functions , 2018, J. Open Source Softw..