Downward Continuation Techniques for Geophysical Potential Fields

Downward continuation is a fundamental inverse technique for extrapolating geophysical potential field measurements from the observation level to subsurface target planes. It serves to sharpen small-scale anomalies associated with shallow density or susceptibility contrasts by mathematically reversing the natural upward diffusion of fields. Formulated as the solution to Laplace’s equation with boundary data, the process is inherently unstable: minor observational noise amplifies dramatically as continuation depth increases. To mitigate ill-posedness, classical approaches employ spatial or spectral domain filters and regularisation strategies such as Tikhonov smoothing. Numerical schemes based on finite-difference or integral formulations have long underpinned gravity and magnetic depth imaging, but they often face trade-offs between resolution and robustness. Recent developments have introduced predictor–corrector algorithms to extend stable continuation to deeper levels, while machine-learning architectures now offer data-driven regularisation that adapts to noise characteristics without explicit parameter tuning. Advances in filter design and optimisation have further enhanced the stability of iterative continuation in the frequency domain. These methodological innovations have broad applications in mineral and hydrocarbon exploration, crustal structure mapping and environmental monitoring, delivering higher-fidelity subsurface images across global geophysical surveys.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Downward Continuation Techniques for Geophysical Potential Fields publication trend

The graph below shows the total number of articles in downward continuation techniques for geophysical potential fields across all publications each year (not limited to Nature Index journals).

Technical terms

Downward continuation: Inverse transformation of potential field measurements from a higher observation plane to a lower target plane, amplifying short-wavelength features and prone to instability.

Potential field: A scalar field governed by Laplace’s equation, such as gravity or magnetic anomalies, arising from subsurface mass or magnetisation contrasts.

Regularisation: Stabilisation technique for ill-posed inverse problems, introducing constraints or smoothing operators to limit noise amplification.

Convolutional neural network: A deep-learning architecture designed for grid-based data, employing convolutional layers to learn spatial patterns and non-linear mappings.

Low-pass filter: Signal-processing operator that suppresses high-frequency noise components while preserving long-wavelength signals.

Predictor–corrector method: Numerical integration strategy combining explicit prediction and implicit correction steps to improve stability and accuracy.

References

1. Two New Methods Based on Implicit Expressions and Corresponding Predictor-Correctors for Gravity Anomaly Downward Continuation and Their Comparison. Remote Sensing (2023).
2. Stable downward continuation of the gravity potential field implemented using deep learning. Frontiers in Earth Science (2023).
3. Magnetic Field Downward Continuation Iterative Method Based on Low-pass Filter. Journal of Physics Conference Series (2023).