A nonnegative constrained method for high-precision downward continuation of gravity field data

Abstract

Gravity downward continuation is an effective method for processing and interpreting gravity data. Finding a gravity downward continuation method with high accuracy and stability is a persistent and popular research topic. At present, there are two common strategies for reducing the ill-posedness inherent in the downward continuation of potential fields, namely, regularization, and the design of wavenumber domain filtering operators. However, these downward continuation strategies often result in unsatisfactory continuation accuracy due to excessive smoothing of the downward continued field. To improve the stability and accuracy of the gravity downward continuation calculation, we construct an equivalent mathematical model of gravity downward continuation. This model is not only equivalent to the traditional gravity downward continuation model but also introduces nonnegative constraints into gravity downward continuation. This approach effectively reduces the ill-posedness of gravity downward continuation. Based on the newly proposed gravity downward continuation model, we propose a constrained gravity downward continuation method. Comparison experiments based on a synthetic gravity model and real gravity anomaly data show that, compared with the other four high performance unconstrained potential field downward continuation methods, the constrained downward continuation method proposed in this paper produces a more stable and accurate downward continued gravity field when the observed data are corrupted by noise.