Abstract Developments in Interferometric Synthetic Aperture Radar (InSAR) and GNSS (Global Navigation Satellite System) during the past decades have promoted significant advances in geosciences, providing high-resolution ground deformation data with dense spatio-temporal coverage. This large dataset can be exploited to produce accurate assessments of the primary processes occurring in geologically active areas. We present a new, original methodology to carry out a multi-source inversion of ground deformation data to better understand the subsurface causative processes. A nonlinear approach permits the determination of location, size and three-dimensional configuration, without any a priori assumption as to the number, nature or shape of the potential sources. The proposed method identifies a combination of pressure bodies and different types of dislocation sources (dip-slip, strike-slip and tensile) that represent magmatic sources and other processes such as earthquakes, landslides or groundwater-induced subsidence through the aggregation of elemental cells. This approach has the following features: (1) simultaneous inversion of the deformation components and/or line-of-sight (LOS) data; (2) simultaneous determination of diverse structures such as pressure bodies or dislocation sources, representing local and regional effects; (3) a fully 3D context; and (4) no initial hypothesis about the number, geometry or types of the causative sources is necessary. This methodology is applied to Mt. Etna (Southern Italy). We analyze the ground deformation field derived from a large InSAR dataset acquired during the January 2009 – June 2013 time period. The application of the inversion approach models several interesting buried structures as well as processes related to the volcano magmatic plumbing system, local subsidence within the Valle del Bove and seaward motion of eastern flank of the volcano.

中文翻译：

---

弹性火山地表变形多源3D模型

摘要 干涉合成孔径雷达 (InSAR) 和 GNSS（全球导航卫星系统）在过去几十年的发展推动了地球科学的重大进步，提供了具有密集时空覆盖范围的高分辨率地面变形数据。可以利用这个大型数据集对地质活动区发生的主要过程进行准确评估。我们提出了一种新的、原始的方法来进行地面变形数据的多源反演，以更好地了解地下的成因过程。非线性方法允许确定位置、大小和三维配置，而无需对潜在源的数量、性质或形状进行任何先验假设。所提出的方法识别了压力体和不同类型的位错源（倾滑、走滑和张性）的组合，这些位错源代表岩浆源和其他过程，如地震、滑坡或地下水通过元素细胞的聚集引起的沉降。该方法具有以下特点： (1) 变形分量和/或视线（LOS）数据的同时反演；(2) 同时确定压力体或位错源等多种结构，代表局部和区域效应；(3) 全 3D 上下文；(4) 不需要关于原因来源的数量、几何形状或类型的初始假设。这种方法适用于山。埃特纳火山（意大利南部）。我们分析了从 2009 年 1 月至 2013 年 6 月期间获取的大型 InSAR 数据集得出的地面变形场。反演方法的应用模拟了几个有趣的埋藏结构以及与火山岩浆管道系统、Valle del Bove 内的局部沉降和火山东侧向海运动相关的过程。