地震研究

交叉梯度约束的地震走时与重力联合反演方法研究

谭智^1,5,姜金钟^1,2,石磊^3,4,李永华^3,4

(1.云南省地震局,云南昆明 650224; 2.中国地震局昆明地震预报研究所,云南昆明 650224; 3.中国地震局地球物理研究所,北京 100081; 4.中国地震局震源物理重点实验室,北京 100081; 5.云南省地震局下关地震监测中心站,云南大理 671000)

关键词：重震联合反演; 结构耦合; 交叉梯度; 顺序策略

Joint Inversion Method for Seismic Body Wave Traveltime and Gravity Data Based on Structural Constraint

TAN Zhi^1,5,JIANG Jinzhong^1,2,SHI Lei^3,4,LI Yonghua^3,4

(1.Yunnan Earthquake Agency,Kunming 650224,Yunnan,China)(2.Kunming Institute of Prediction,China Earthquake Administration,Kunming 650224,Yunnan,China)(3.Institute of Geophysics,China Earthquake Administration,Beijing 100081,China)(4.Key Laboratory of Earthquake Source Physics,China Earthquake Administration,Beijing 100081,China)(5.Xiaguan Earthquake Monitoring Center Station,Yunnan Earthquake Agency,Dali 671000,Yunnan,China)

Keywords：joint inversion of gravity and seismic data; structural coupling; cross-gradient; sequential strategy

DOI: 10.20015/j.cnki.ISSN1000-0666.2025.0056

备注

摘要

全文

图/表

参考文献

不同地球物理数据参与的联合反演,可有效减少反演的多解性,为地球内部结构研究提供多参数约束。重力与地震联合反演的开展常基于速度-密度经验关系,但由于地下结构复杂,部分地区岩石物理关系难以准确确定。针对这一问题,提出了一种基于交叉梯度约束的地震走时与重力联合反演方法。该方法将频率域重力反演和近震体波走时反演相结合,采用优化的顺序策略,实现了速度模型与密度模型的相似性耦合。理论模型和青藏高原东北缘实际观测数据试验都显示,联合反演有效提高了密度结构的垂向分辨率。重力数据的引入,更好地约束了地壳浅层(地下5 km以内)P波速度结构,明显改善了体波射线分布稀疏地区的横向分辨能力。

Joint inversion of different geophysical data can effectively reduce the non-uniqueness of results and provide more constraints for studying the Earth's interior structure.Joint inversion of gravity and seismic data is usually carried out based on velocity-density relationships.However,the actual subsurface structures are complex,making it difficult to accurately determine the velocity-density relationships in certain areas.To address this issue,we have developed a joint inversion method for seismic traveltime and gravity based on cross-gradient constraints.This method combines frequency-domain gravity inversion and body wave tomography,and uses an optimized sequential strategy to achieve the structural similarity coupling between the velocity model and the density model.Synthetic tests and real data tests of the Qinghai-Xizang Plateau suggest that this joint inversion can effectively enhance the vertical resolution capabilities of the density structure.Furthermore,incorporating gravity data better constrains the P-wave velocity structure in the shallow Crust(5 km underground)and significantly improves the horizontal resolution capabilities in areas with sparse body wave ray coverage.

引言
1 方法与原理
2 理论模型试验
3 实际数据测试
4 结论

图1 交叉梯度约束的体波走时与重力异常联合反演流程<br/><br/>Fig.1 Flowchart of the cross-gradient joint inversionof body wave traveltime and gravity data

图1 交叉梯度约束的体波走时与重力异常联合反演流程

Fig.1 Flowchart of the cross-gradient joint inversionof body wave traveltime and gravity data

表1 合成速度模型异常体几何和物性参数<br/>Tab.1 Geometrical and physical parametersof the simple velocity model

表1 合成速度模型异常体几何和物性参数
Tab.1 Geometrical and physical parametersof the simple velocity model

图2 合成理论速度模型(a)与实际加噪重力异常(b)<br/>Fig.2 The synthetic theoretical velocity model(a)and gravity anomalies with Gaussian noise(b)

图2 合成理论速度模型(a)与实际加噪重力异常(b)
Fig.2 The synthetic theoretical velocity model(a)and gravity anomalies with Gaussian noise(b)

图3 青藏高原东北缘地震台阵和MS≥1.5地震事件分布(a)与剩余重力异常(b)<br/>Fig.3 Distribution of seismic stations and MS≥1.5 earthquake events(a),and residual gravity anomalies(b)in the northeastern margin of the Qinghai-Xizang Plateau

图3 青藏高原东北缘地震台阵和MS≥1.5地震事件分布(a)与剩余重力异常(b)
Fig.3 Distribution of seismic stations and MS≥1.5 earthquake events(a),and residual gravity anomalies(b)in the northeastern margin of the Qinghai-Xizang Plateau

图4 交叉梯度与走时数据拟合误差(a)和重力数据拟合误差(b)L-curve<br/>Fig.4 Cross-gradient and traveltime data fitting error L-curve(a),cross-gradientand gravity data fitting error L-curve(b)

图4 交叉梯度与走时数据拟合误差(a)和重力数据拟合误差(b)L-curve
Fig.4 Cross-gradient and traveltime data fitting error L-curve(a),cross-gradientand gravity data fitting error L-curve(b)

图5 真实、单独反演和联合反演构建的P波速度模型(a)和密度模型(b)<br/>Fig.5 P-wave velocity models(a)and density models(b)obtainedfrom true data,solo inversion,and joint inversion

图5 真实、单独反演和联合反演构建的P波速度模型(a)和密度模型(b)
Fig.5 P-wave velocity models(a)and density models(b)obtainedfrom true data,solo inversion,and joint inversion

图6 真实、单独反演和联合反演构建的P波速度(a)和密度(b)结果垂直剖面<br/>Fig.6 Vertical profiles of P-wave velocity models(a)and density models(b)obtainedfrom true data,solo inversion,and joint inversion

图6 真实、单独反演和联合反演构建的P波速度(a)和密度(b)结果垂直剖面
Fig.6 Vertical profiles of P-wave velocity models(a)and density models(b)obtainedfrom true data,solo inversion,and joint inversion

表2 单独反演与联合反演结果对比<br/>Tab.2 Comparison between solo inversion resultsand joint inversion results

表2 单独反演与联合反演结果对比
Tab.2 Comparison between solo inversion resultsand joint inversion results

图7 单独反演结果(a)和联合反演结果(b)的速度-密度交会拟合曲线<br/>Fig.7 The intersection points of velocity and density for the results obtainedfrom solo inversion(a)and joint inversion(b)

图7 单独反演结果(a)和联合反演结果(b)的速度-密度交会拟合曲线
Fig.7 The intersection points of velocity and density for the results obtainedfrom solo inversion(a)and joint inversion(b)

图8 单独反演和联合反演后的P波走时残差(a)和重力残差分布(b)<br/>Fig.8 P-wave traveltime residual distribution histogram(a)and gravity residualdistribution maps(b)after solo inversion and joint inversion

图8 单独反演和联合反演后的P波走时残差(a)和重力残差分布(b)
Fig.8 P-wave traveltime residual distribution histogram(a)and gravity residualdistribution maps(b)after solo inversion and joint inversion

图9 青藏高原东北缘地区单独反演和联合反演构建的P波速度(a)和密度(b)结果水平切片<br/>Fig.9 P-wave velocity(a)and density(b)variations from solo and joint inversionin the northeastern margin of the Qinghai-Xizang Plateau

图9 青藏高原东北缘地区单独反演和联合反演构建的P波速度(a)和密度(b)结果水平切片
Fig.9 P-wave velocity(a)and density(b)variations from solo and joint inversionin the northeastern margin of the Qinghai-Xizang Plateau