A preprint posted to arXiv in March 2026 by researcher Naveen K. offers a targeted but meaningful advance in geophysical forward modeling. The paper tackles a persistent headache in direct-current (DC) resistivity surveys: how to accurately simulate electric potentials when the ground surface is bumpy, sloped, or sharply angled rather than conveniently flat.

Electrical resistivity tomography (ERT) works by injecting current into the earth through electrodes and measuring voltage differences at other points. These data are inverted to build images of subsurface resistivity variations, aiding everything from groundwater mapping and mineral exploration to archaeological site assessment and landslide risk analysis. The 'forward model' — simulating expected measurements for a known resistivity distribution — sits at the heart of this inversion process. When terrain undulates, the math gets messy.

Most singularity-removal techniques subtract the singular part of the potential analytically so that numerical methods only solve for a smoother secondary field. The conventional analytical primary potential assumes a flat homogeneous half-space. The new work demonstrates that this assumption breaks down at sharp corners, slope breaks, or even at the linear facets created by ordinary finite-element surface meshes. The geometric mismatch between the assumed flat surface and the actual solid angle subtended by the real topography injects systematic errors.

The authors derive a new closed-form primary potential based on a V-shaped wedge geometry. This analytical expression matches the local surface slope discontinuities that appear in both natural landscapes and discretized models. They embed this primary field inside a 2.5-D finite-element forward operator (assuming strike invariance along one horizontal axis, a common compromise between computational cost and realism). Numerical tests on three synthetic cases — perfectly flat ground, a V-shaped trench, and a sinusoidal hill-and-valley profile — show relative errors consistently below 0.1 percent even when using coarse linear tetrahedral meshes. Traditional flat half-space primary potentials produced errors orders of magnitude larger under identical mesh conditions.

This preprint, which has not yet undergone peer review, builds directly on foundational singularity-removal work (Lowry et al., Geophysics, 1989) and classic 2.5-D finite-element formulations (Dey & Morrison, Geophysics, 1979). It also connects to more recent efforts that handle topography through either massive local mesh refinement or fully 3-D unstructured meshes (e.g., Rücker et al., Computers & Geosciences, 2017, describing the open-source pyGIMLi library). Those approaches incur high computational penalties; the wedge analytical trick achieves comparable or better accuracy at far lower cost.

What the paper itself under-emphasizes is the downstream impact on inversion stability. Inversion algorithms run the forward model thousands of times; even small systematic forward errors can translate into spurious resistivity artifacts or slow convergence. By enabling reliable results on coarser meshes, the method could cut inversion runtimes substantially, an important practical gain for large 3-D surveys now common in engineering geology. The study is purely synthetic — no field data validation is presented — and remains limited to 2.5-D geometry. Real-world 3-D topography with closed depressions or overhangs would require further generalization, possibly to polyhedral or curved analytical elements.

The deeper pattern here mirrors developments in other computational sciences: instead of brute-force grid refinement, researchers are returning to analytical insight tuned to the dominant local geometry. Similar strategies appear in aerodynamic boundary-layer corrections and seismic wavefield singularity handling. For geophysicists surveying steep volcanic flanks, glaciated valleys, or open-pit mines, the practical takeaway is clear — topography should no longer be treated as a minor perturbation to be smoothed away. It is a first-order geometric boundary condition that deserves an analytically faithful primary field.

If adopted and extended, this wedge formulation could quietly improve the reliability of thousands of resistivity surveys conducted annually in rugged terrain, yielding clearer pictures of what lies beneath our feet.