curiesphere v2.0.0

curiesphere v2.0.0

Outputs differ from v1.0.0. Forward-model amplitudes are typically 10–15% higher than with v1.0.0 (global RMS of Br); pattern correlations are almost unchanged. To reproduce the published figures and numbers of Williams et al. (2025, JGR), use v1.0.0.

What's new

• Bug fixes. Errors in the forward transform and in the units of the induced magnetization have been corrected. Details are in docs/review-2026-09-units-and-geometry.md.
• Ocean depth. Optional forward models place the oceanic sources at a realistic depth: below the sea floor and sediments, on the WGS84 ellipsoid, with each layer of the oceanic lithosphere at its own depth. See notebooks/depth_models.py.
• Benchmarks. The code agrees to rounding error with an independent equivalent-source (dipole sum) calculation and an independent spherical-harmonic quadrature. See docs/benchmark-dipole-sum.md and docs/benchmark-sh-quadrature.md.
• Tests. Run python -m pytest tests.
• Packaging. setup.py has been repaired, so pip install . now works and installs the input data. The package is named curiesphere; it is still imported as remit.

JGR submission

Version of the code associated with submission to JGR, and initial generation of zenodo doi