基于深度学习数据集的鲁西隆起区域层析成像

Tomography of the Luxi Uplift Based on the Deep Learning Dataset

ES评分 0

DOI 10.14075/j.jgg.2025.08.281
刊名
Journal of Geodesy and Geodynamics
年,卷(期) 2026, 46(6)
作者
作者单位

1 山东省地震局, 济南, 250014

摘要
采用深度学习方法构建一个高精度地震数据集, 基于双差层析成像方法反演得到区域三维速度模型。研究表明, 地表速度结构与地质构造之间存在联系, 即山脉区多呈现P波高速异常, 而坳陷区多呈P波低速异常。这种特征可能源于, 褶皱区地表多为基岩, 波速相对高, 而坳陷区地表多为第四系土层, 波速相对低。地震震中密集分布在高速和低速异常的过渡区域。高速和低速过渡区域应力梯度变化显著, 更易发生破裂而导致地震活动性强。鲁西隆起区存在一些近垂直的高速和低速块体, 部分震中也呈高倾角分布, 推测为高倾角断层。这可能源于, 在太平洋板块俯冲和郯庐断裂带运动影响下, 鲁西隆起经历多个挤压和伸展的构造作用过程, 形成多个隆起和坳陷的周期, 期间岩石褶皱不断挤压和伸展, 褶皱的倾角逐渐升高, 形成近垂直的速度块体和高倾角断层。通过P波与S波速度联合讨论, 推测尼山和鲁山东北区域均存在地幔玄武质岩浆的侵入, 尼山地区以侵入岩为主, 而鲁山东北区域则存在地幔玄武岩浆突破地表, 形成古火山与喷出岩。
Abstract
A high-precision seismic dataset was constructed using deep learning methods, and a regional three-dimensional velocity model was obtained through double-difference tomography. The study reveals a correlation between surface velocity structures and geological features: mountainous areas generally exhibit high-velocity P-wave anomalies, whereas depression zones are characterized by low-velocity P-wave anomalies. This pattern may be attributed to the widespread presence of bedrock with relatively high wave velocity in folded regions, in contrast to the Quaternary sedimentary layers with lower velocity in depression areas. Seismic epicenters are densely distributed along the transition zones between high- and low-velocity anomalies. These transitional regions experience significant stress gradients, which facilitate rock fracturing and lead to heightened seismic activity.The Luxi uplift is characterized by several near-vertical high- and low-velocity blocks, with some earthquake epicenters also displaying steeply dipping distributions, suggesting the presence of high-angle faults. This structural configuration may result from multiple phases of compression and extension under the influence of Pacific Plate subduction and the tectonic activity of the Tan-Lu fault zone. These processes have led to cyclic uplift and subsidence in the region, during which progressive compression and extension of rock folds gradually increased their dip angles, forming near-vertical velocity blocks and high-angle faults.Integrated analysis of P- and S-wave velocities suggests the presence of mantle-derived basaltic magmatic intrusions beneath both the Nishan area and the northeastern region of the Lushan. In the Nishan area, the intrusions are predominantly plutonic, whereas in northeastern Lushan, mantle-derived basaltic magma likely reached the surface, forming ancient volcanoes and extrusive rocks.
关键词
鲁西隆起; 济阳坳陷; 双差层析成像(TomoDD); 深度学习方法; 地震数据集; 三维速度模型
KeyWord
Luxi uplift; Jiyang depression; double-difference tomography (TomoDD); deep learning method; seismic dataset; 3D velocity model
基金项目
页码 718-727
  • 参考文献
  • 相关文献
  • 引用本文

崔鑫, 董腾超, 张春鹏. 基于深度学习数据集的鲁西隆起区域层析成像 [J]. 大地测量与地球动力学. 2026; 46; (6). 718 - 727.

  • 文献评论

相关学者

相关机构