Electromagnetic data inversion

A system, comprising:

• a data engine to receive electromagnetic (EM) data acquired during towed streamer acquisition;
• a sensitivity engine to determine an electric field within a predetermined sensitivity area around each of a plurality of source positions associated with the towed streamer acquisition;
• an inversion engine to:
• iteratively invert the electric field using a plurality of nodes of the system until a convergence criterion is met;
• split the iterative inversion into portions demanding memory amounts below a threshold;
• and
• distribute each of the portions to a different node within the plurality of nodes for inversion, wherein each different node has memory available above the threshold;
• and
• a resistivity engine to determine a seafloor resistivity based on the iteratively inverted electric field and an iteratively solved linear system of equations.

A non-transitory machine-readable medium storing instructions executable by a processing resource to:

• build an inversion grid of a predetermined size surrounding a source and comprising a plurality of controlled source electromagnetic (CSEM) electrical resistivity data cells;
• determine an electric field for each of the plurality of CSEM electrical resistivity data cells;
• iteratively invert, using a plurality of nodes, each of the electric fields using a Gauss-Newton inversion algorithm;
• generate a linear system of equations comprising a Jacobian matrix based on the plurality of determined electric fields and the iterative inversion;
• iteratively solve, using the plurality of nodes, the linear system of equations until a convergence criterion is met;
• combine results of the iterative inverting and the iterative solving;
• and
• determine a seafloor resistivity associated with the each of the CSEM electrical resistivity data cells based on the combined results.

A method to manufacture a geophysical data product, the method comprising:

• obtaining geophysical data, wherein obtaining the geophysical data comprises:
• determining an electric field associated with electromagnetic (EM) data collected using a plurality of sources and receivers during an EM survey and within a predetermined sensitivity area around each of a plurality of source positions;
• processing the geophysical data, comprising:
• iteratively inverting the electric field for a sub-surface resistivity EM model indicative of a subterranean formation for each of a plurality of EM electrical resistivity data cells within each of the predetermined sensitivity areas;
• generating a linear system of equations comprising a Jacobian matrix based on the iterative inversion;
• storing the linear system of equations;
• and
• iteratively solving the linear system of equations at each iteration of the iterative inversion to update the sub-surface resistivity EM model until a convergence criterion is met,
• wherein performance of portions of the iterative solving is distributed over a plurality of nodes demanding memory amounts below a threshold;
• and
• wherein the updated sub-surface resistivity EM model is better indicative of the subterranean formation as compared a sub-surface resistivity EM model of a previous iteration;
• and
• recording the geophysical data product on one or more non-transitory machine-readable media, thereby creating the geophysical data product.