Gravity Inversion

This dataset contains 28,000 synthetic subsurface density fields, for each of which the surface gravity field (computed through a linear relation) has been calculated. This dataset can be used to apply machine learning methods to estimate density from gravity measurements.

Instructions:
Dataset Components:
• Part 1: Two-dimensional synthetic density fields representing Earth's density (Density_Train and Density_Test)
• Part 2: The matrix representing the relationship between density and gravity (A).
• Part 3: Resulting gravitational fields at Earth’s surface generated by subsurface density distributions (Gravity_Train and Gravity_Test)

Dataset Creation Parameters:
• The density fields are discretized into nxn(51x51) parts.
• The matrix A, with dimensions (6n,n^2), provides a linear mapping from density to gravity.
• The gravitational field on the surface is discretized into two components, vertical and horizontal, each with a size of 3×n.