Abstract：
We present a new filtering method for potential fields, based on modelling the fields in terms of very compact solutions, i.e., the sources are expected to occupy the smallest allowable volume in the source domain. The selected solutions, which we call \"Extremely Compact Sources\" (ECS) form a sort of atomized model, which still satisfies the non-unique inverse problem of gravity and magnetic fields. The ECS model is not only characterized by sparsity, but also by large values of the physical property (density or magnetic susceptibility). The sparse nature of the model allows for the definition of a highly localized filter, which can be obtained by simply specifying the atoms to be selected in a given area. This feature allows managing tasks normally impossible with traditional filters, such as the separation of interfering anomalies having a similar wavenumber content. In addition, the procedure can perform a very effective regional/residual separation. We demonstrate the method on synthetic cases and apply it in the real case of gravity data of Campi Flegrei volcanic area (Italy), where we use the ECS filtering to isolate the gravity effect of the Mount Olibano dome.
摘要：
我们提出了一种新的势场滤波方法，基于非常紧凑的解决方案对字段进行建模，即，预计源将占用源域中的最小允许体积。选定的解决方案，我们称之为“极其紧凑的源”(ECS)形成一种原子化模型，仍然满足重力和磁场的非唯一反问题。ECS模型不仅具有稀疏性，但也通过大的物理属性值（密度或磁化率）。模型的稀疏特性允许定义高度局部化的滤波器，这可以通过简单地指定要在给定区域中选择的原子来获得。此功能允许使用传统过滤器管理通常不可能的任务，例如分离具有相似波数内容的干扰异常。此外,该程序可以执行非常有效的区域/残余分离。我们在综合案例中演示了该方法，并将其应用于坎皮·弗莱格雷火山地区（意大利）的重力数据的真实案例中，我们使用ECS过滤来隔离奥利巴诺山穹顶的重力效应。
