大地电磁数据静位移影响及校正方法研究
|
摘要
MT是以自然界分布广泛的宽频带(10-4～104Hz)天然变化电磁场为场源,对深部地质构造进行研究的一种频率域电磁测深法。
与其它频率域电磁测深法一样,MT的测量数据也经常受到静位移畸变的影响。静位移效应会在频率域给全频段的测量数据造成负面影响,使得后续的反演结果及地质解释与实际情况有较大偏差。
研究者们认为：静位移效应的产生机理,主要是因为在外部电场作用下,浅部三维异常体与区域构造介质的界面上会产生积累电荷,积累电荷的电场使得正常的感应电场发生畸变,但对正常的感应磁场影响却非常小。若用畸变后的电场计算视电阻率,计算视电阻率在频率域双对数坐标系中,相对于正常场视电阻率曲线会在模值坐标方向上发生平移,但曲线形态保持不变,同时,频率域的相位曲线不受影响。
静位移校正是MT数据处理中必须解决,却很难彻底解决的问题。研究者们针对不同的实际应用情况提出了很多切实可行的校正方法,如：曲线平移法,各种空间滤波法,联合TEM进行校正的方法,等等。
虽然,上述校正方法都具有一定的实用性,但只用单个测点的采集数据对其自身静位移效应进行校正的方法却仍是个有待研究的问题。
目前,MT的三维／二维反演模型是人们的研究热点,本论文中的新方法也是针对三维／二维反演模型设计的。
设区域构造主轴坐标系中,测量的阻抗模值张量Za,测量数据受到的静位移畸变可用静位移效应张量C与正常场阻抗模值张量Z02相乘表示,其数学形式如下：
新方法结合适当的限制条件,对通过穷取法所提出的静位移效应元素估计值组合进行筛选,最终获得目的数位上的唯一近似值组合,筛选过程主要分四步骤：
步骤一：由式(1)可得：Z1/Z3=C2/C4, Z2/Z4=C1/C3式(2)
若为一维构造,则有：Z2/Z3=-C1/C4, Z1/Z4=-C2/C3式(3)
基于上述限定关系,建立一维及二维限定条件,对待选估计值组合进行筛选。
步骤二：将上步筛选剩余的组合通过下式：Z0=C(-1)*Za式(4)计算出正常场阻抗模值张量的可能值,通过观察Zte和Ztm曲线是否与区域构造情况相符合,并以此对估计值组合进行筛选。
步骤三：根据四个静位移效应张量元素的16种进位情况,做出相应的M-D图板,再根据估计值与图版上点的重合情况进行筛选。
步骤四：再次利用式(4),观察阻抗模值张量的对角线和副对角线元素是否符合区域构造的特征,或者借鉴已有地质资料,最终筛选出唯一的估计值组合。
通过对模型及实测数据进行分析处理,取得了比较理想的校正效果,但此新方法仍存在不足,需要后续不断对其进行完善。
The MT is a kind of electromagnetic sounding methods used for the study the deep geological structures in the frequency domain, of which the field resource is the natural changeable electromagnetical field which is widely distributed in the nature and has the very broad frequency range (10^(-4)-10^(4) H z).
Like the other frequency domain electromagnetic sounding methods, the MT measured data are always subject to the static shift distortion. The static shift effect will make negative impact on the data in the whole frequency range and then the subsequent inversion results and the geological interpretation will have a large deviation with the actual situation.
The researchers consider that the main mechanism of the generation of the static shift effect is the distortion of the normal electric field and it is made by the accumulated charge on the boundary between the mediums which have different conductivity. Under the action of the external electric field, the accumulated charge will be accumulated on the boundary between the shallow three-dimensional anomalies and the regional structural medium while the affect to the induced magnetic field is very weak. If the values of apparent resistivity are calculated by the values of the distorted electric field, the curves of the apparent resistivity will translate in the direction of the modulus value, comparing with the curves of the apparent resistivity calculated by the normal value, while the shape of the curves and the curves of the phases will not change.
The correction of the static shift effect is a problem which must be solved in the MT data processing, but it is also hard to solve the problem completely. For the practical applications in different situations, researchers have made a lot of useful methods, such as:the curve translation method, the spatial filtering methods and jointing the TEM method into the correction, etc.
Although the correction methods above are useful in the certain ranges, making the correction by only using the data from the measuring pot itself is still a problem which is unsolved.
At present, the study of the MT3D/2D inversion model is a hot spot. In this paper, the new method of the correction is designed for the3D/2D inversion model.
In the regional tectonic axis coordinate system, the impedance modulus tensor will be marked as:Za and the distortion of the measured data caused by the static shift can be expressed in the form of the static shift tensor C multiplying the normal impedance modulus tensor Zoa and the mathematical form of the calculation is showed below:
The combinations of the static shift effect elements which are made by the brute-force method will be filtered by combining the new method with the appropriate restrictions and finally the unique combination of the approximations on the target digits will be got.
The filtering process consists of the four steps.
Step1:we can get the relationships below by formula(1): Z1/Z3=C2/C4, Z2/Z4=C1/C3formula(2)
In the1D structure, the forms of formula(2) can be deformed into: Z2/Z3=-C1/C4, Z1/Z4=-C2/C3formula(3)
Based on the limit relations above, the combinations of the estimates will be filtered by1D and2D restrictions.
Step2:to calculate the remaining combinations by the formula below: Z0=C-1Za formula (4)
In this step, the estimates of the impedance modulus tensors of the normal fields will be calculated at first and then the curves of Zte and Ztm will be observed. If the forms of the curves do not fit the characteristics of the regional tectonic structure, the relative combinations of the estimates will be filtered.
Step3:according to16approximating situations of the4elements of the static shift tensor,16related M-D boards will be made and the estimates will be filtered by watching the effects whether the estimates will coincide with the points on the M-D board or not.
Step4:by observing the effects of the diagonal elements and vice diagonal elements of the impedance modulus tensors respectively fitting regional structural characteristics or by learning from the existing geological data, the unique combination of the estimates will be got at last.
By processing and analyzing the relative models and the actual measured data, the ideal effect of the correction has been got, however, this new method is still not perfect and it is necessary to do a lot study to improve the method.
与其它频率域电磁测深法一样,MT的测量数据也经常受到静位移畸变的影响。静位移效应会在频率域给全频段的测量数据造成负面影响,使得后续的反演结果及地质解释与实际情况有较大偏差。
研究者们认为：静位移效应的产生机理,主要是因为在外部电场作用下,浅部三维异常体与区域构造介质的界面上会产生积累电荷,积累电荷的电场使得正常的感应电场发生畸变,但对正常的感应磁场影响却非常小。若用畸变后的电场计算视电阻率,计算视电阻率在频率域双对数坐标系中,相对于正常场视电阻率曲线会在模值坐标方向上发生平移,但曲线形态保持不变,同时,频率域的相位曲线不受影响。
静位移校正是MT数据处理中必须解决,却很难彻底解决的问题。研究者们针对不同的实际应用情况提出了很多切实可行的校正方法,如：曲线平移法,各种空间滤波法,联合TEM进行校正的方法,等等。
虽然,上述校正方法都具有一定的实用性,但只用单个测点的采集数据对其自身静位移效应进行校正的方法却仍是个有待研究的问题。
目前,MT的三维／二维反演模型是人们的研究热点,本论文中的新方法也是针对三维／二维反演模型设计的。
设区域构造主轴坐标系中,测量的阻抗模值张量Za,测量数据受到的静位移畸变可用静位移效应张量C与正常场阻抗模值张量Z02相乘表示,其数学形式如下：
新方法结合适当的限制条件,对通过穷取法所提出的静位移效应元素估计值组合进行筛选,最终获得目的数位上的唯一近似值组合,筛选过程主要分四步骤：
步骤一：由式(1)可得：Z1/Z3=C2/C4, Z2/Z4=C1/C3式(2)
若为一维构造,则有：Z2/Z3=-C1/C4, Z1/Z4=-C2/C3式(3)
基于上述限定关系,建立一维及二维限定条件,对待选估计值组合进行筛选。
步骤二：将上步筛选剩余的组合通过下式：Z0=C(-1)*Za式(4)计算出正常场阻抗模值张量的可能值,通过观察Zte和Ztm曲线是否与区域构造情况相符合,并以此对估计值组合进行筛选。
步骤三：根据四个静位移效应张量元素的16种进位情况,做出相应的M-D图板,再根据估计值与图版上点的重合情况进行筛选。
步骤四：再次利用式(4),观察阻抗模值张量的对角线和副对角线元素是否符合区域构造的特征,或者借鉴已有地质资料,最终筛选出唯一的估计值组合。
通过对模型及实测数据进行分析处理,取得了比较理想的校正效果,但此新方法仍存在不足,需要后续不断对其进行完善。
The MT is a kind of electromagnetic sounding methods used for the study the deep geological structures in the frequency domain, of which the field resource is the natural changeable electromagnetical field which is widely distributed in the nature and has the very broad frequency range (10^(-4)-10^(4) H z).
Like the other frequency domain electromagnetic sounding methods, the MT measured data are always subject to the static shift distortion. The static shift effect will make negative impact on the data in the whole frequency range and then the subsequent inversion results and the geological interpretation will have a large deviation with the actual situation.
The researchers consider that the main mechanism of the generation of the static shift effect is the distortion of the normal electric field and it is made by the accumulated charge on the boundary between the mediums which have different conductivity. Under the action of the external electric field, the accumulated charge will be accumulated on the boundary between the shallow three-dimensional anomalies and the regional structural medium while the affect to the induced magnetic field is very weak. If the values of apparent resistivity are calculated by the values of the distorted electric field, the curves of the apparent resistivity will translate in the direction of the modulus value, comparing with the curves of the apparent resistivity calculated by the normal value, while the shape of the curves and the curves of the phases will not change.
The correction of the static shift effect is a problem which must be solved in the MT data processing, but it is also hard to solve the problem completely. For the practical applications in different situations, researchers have made a lot of useful methods, such as:the curve translation method, the spatial filtering methods and jointing the TEM method into the correction, etc.
Although the correction methods above are useful in the certain ranges, making the correction by only using the data from the measuring pot itself is still a problem which is unsolved.
At present, the study of the MT3D/2D inversion model is a hot spot. In this paper, the new method of the correction is designed for the3D/2D inversion model.
In the regional tectonic axis coordinate system, the impedance modulus tensor will be marked as:Za and the distortion of the measured data caused by the static shift can be expressed in the form of the static shift tensor C multiplying the normal impedance modulus tensor Zoa and the mathematical form of the calculation is showed below:
The combinations of the static shift effect elements which are made by the brute-force method will be filtered by combining the new method with the appropriate restrictions and finally the unique combination of the approximations on the target digits will be got.
The filtering process consists of the four steps.
Step1:we can get the relationships below by formula(1): Z1/Z3=C2/C4, Z2/Z4=C1/C3formula(2)
In the1D structure, the forms of formula(2) can be deformed into: Z2/Z3=-C1/C4, Z1/Z4=-C2/C3formula(3)
Based on the limit relations above, the combinations of the estimates will be filtered by1D and2D restrictions.
Step2:to calculate the remaining combinations by the formula below: Z0=C-1Za formula (4)
In this step, the estimates of the impedance modulus tensors of the normal fields will be calculated at first and then the curves of Zte and Ztm will be observed. If the forms of the curves do not fit the characteristics of the regional tectonic structure, the relative combinations of the estimates will be filtered.
Step3:according to16approximating situations of the4elements of the static shift tensor,16related M-D boards will be made and the estimates will be filtered by watching the effects whether the estimates will coincide with the points on the M-D board or not.
Step4:by observing the effects of the diagonal elements and vice diagonal elements of the impedance modulus tensors respectively fitting regional structural characteristics or by learning from the existing geological data, the unique combination of the estimates will be got at last.
By processing and analyzing the relative models and the actual measured data, the ideal effect of the correction has been got, however, this new method is still not perfect and it is necessary to do a lot study to improve the method.
引文
[1]孟银生. CSAMT中的静位移现象研究：[硕士学位论文].北京：中国地质大学（北京），2010
[2]黄兆辉，底青云，侯胜利. CSAMT的静态效应校正及应用.地球物理学进展，2006，21（4），1290~1295
[3]陈辉.可控源音频大地电磁法静态效应校正技术及其方法研究：[硕士学位论文].北京：中国地质大学（北京），2007
[4]陈辉，王春庆，雷达，等. CSAMT法静态效应模拟及其校正方法对比.物探化探计算技术，2007，29（增刊）
[5]张旭，刘宽厚，李貅. CSAMT的静校正应用--联合反演法.西北地质，2010，43（2）
[6]王书明，林长佑，杨长福，等.电偶源频率电测深中的静位移现象及视电阻率曲线的畸变.西北地震学报，2000，22（1）
[7]梁生贤，张胜业，祁晓雨，等.基于空间滤波和相位换算的MT静校正方法比较.工程地球物理学报，2010，7（3）
[8]张翔，胡文宝，严良俊.小波变换在大地电磁测深校正中的应用.江汉石油学院学报，2002，24（2）
[9]吴长祥，刘苗，吴健生.克里格方法在大地电磁静校正中的应用.石油物探，2008，47（1）
[10]杨长福，林长佑.用TEM反演法进行MT静位移的识别和校正.地球科学--中国地质大学学报，2001，26（6）
[11]段波.校正大地电磁测深中静态效应的首枝重合法.长春地质学院学报，1994，24（4）
[12]王家映.关于大地电磁的静校正问题.地质科技情报，1992，11（1）
[13]张辉.大地电磁二维正演软件系统开发：[硕士学位论文].北京：中国地质大学（北京），2010
[14]蔡军涛.大地电磁三维畸变特征及校正新技术的应用研究：[博士学位论文].北京：中国地震局地质研究所，2009
[15]蔡剑华.基于Hilbert-Huang变换的大地电磁信号处理方法与应用研究：[博士学位论文].湖南省长沙市：中南大学，2010
[16]蔡剑华，汤井田.基于Hilbert-Huang变换的大地电磁信号谱估计方法.石油地球物理勘探，2010，45（5）
[17]胡玉平，鲍光淑，贾继标.基于畸变3个因子的MT畸变曲线改正方法与应用.地质与勘探，2005，41（6）
[18]王立凤，晋光文，孙洁，等.一种简单的大地电磁阻抗张量畸变分解方法.西北地震学报，2001，23（2）
[19]李艳青.张量阻抗分解的实现及在藏南构造走向研究中的应用：[硕士学论文].北京：中国地质大学（北京），2008
[20]晋光文，孔祥儒.大地电磁阻抗张量的畸变与分解.北京：地震出版社，2006.
[21]李金铭.地电场与电法勘探.北京：地质出版社，2005.
[22] Karsten Bahr.Interpretation of Magnetotelluric Impedance Tensor:RegionalInduction and Local Telluric Distortion.Journal of Geophysics,1988,62:119-127
[23] M.N.Berdichekvskiy,V.I.Dmitriev.Distortion of Magnetic and Electrical Fieldsby Near-surface Lateral Inhomogeneities.Acta Geodaet.,Geophys.et Montanist.Acad.Sci.Hung.Tomus,1976,11(3-4):447-483
[24] Philip E. Wannamaker, Gerald W. Hohmann and Stanley H. Ward.MagnetotelluricResponses of Three-dimensional Bodies in Layered Earths. Geophysics, Vol.49, No.9,September1984:1517-1533
[25] P.Zhang, R.G.Roberts and L.B.Pedersen.Magnetotelluric Strike Rules.Geophics,March1987,52(3)
[26] J.C.Larsen. Removal of Local Surface Conductivity Effects from Low FrequencyMantle Response Curves. Acta Geodaet.,Geophys.et Montanist.Acad.Sci.Hung.Tomus,1977,12(1-3):183-186
[2]黄兆辉，底青云，侯胜利. CSAMT的静态效应校正及应用.地球物理学进展，2006，21（4），1290~1295
[3]陈辉.可控源音频大地电磁法静态效应校正技术及其方法研究：[硕士学位论文].北京：中国地质大学（北京），2007
[4]陈辉，王春庆，雷达，等. CSAMT法静态效应模拟及其校正方法对比.物探化探计算技术，2007，29（增刊）
[5]张旭，刘宽厚，李貅. CSAMT的静校正应用--联合反演法.西北地质，2010，43（2）
[6]王书明，林长佑，杨长福，等.电偶源频率电测深中的静位移现象及视电阻率曲线的畸变.西北地震学报，2000，22（1）
[7]梁生贤，张胜业，祁晓雨，等.基于空间滤波和相位换算的MT静校正方法比较.工程地球物理学报，2010，7（3）
[8]张翔，胡文宝，严良俊.小波变换在大地电磁测深校正中的应用.江汉石油学院学报，2002，24（2）
[9]吴长祥，刘苗，吴健生.克里格方法在大地电磁静校正中的应用.石油物探，2008，47（1）
[10]杨长福，林长佑.用TEM反演法进行MT静位移的识别和校正.地球科学--中国地质大学学报，2001，26（6）
[11]段波.校正大地电磁测深中静态效应的首枝重合法.长春地质学院学报，1994，24（4）
[12]王家映.关于大地电磁的静校正问题.地质科技情报，1992，11（1）
[13]张辉.大地电磁二维正演软件系统开发：[硕士学位论文].北京：中国地质大学（北京），2010
[14]蔡军涛.大地电磁三维畸变特征及校正新技术的应用研究：[博士学位论文].北京：中国地震局地质研究所，2009
[15]蔡剑华.基于Hilbert-Huang变换的大地电磁信号处理方法与应用研究：[博士学位论文].湖南省长沙市：中南大学，2010
[16]蔡剑华，汤井田.基于Hilbert-Huang变换的大地电磁信号谱估计方法.石油地球物理勘探，2010，45（5）
[17]胡玉平，鲍光淑，贾继标.基于畸变3个因子的MT畸变曲线改正方法与应用.地质与勘探，2005，41（6）
[18]王立凤，晋光文，孙洁，等.一种简单的大地电磁阻抗张量畸变分解方法.西北地震学报，2001，23（2）
[19]李艳青.张量阻抗分解的实现及在藏南构造走向研究中的应用：[硕士学论文].北京：中国地质大学（北京），2008
[20]晋光文，孔祥儒.大地电磁阻抗张量的畸变与分解.北京：地震出版社，2006.
[21]李金铭.地电场与电法勘探.北京：地质出版社，2005.
[22] Karsten Bahr.Interpretation of Magnetotelluric Impedance Tensor:RegionalInduction and Local Telluric Distortion.Journal of Geophysics,1988,62:119-127
[23] M.N.Berdichekvskiy,V.I.Dmitriev.Distortion of Magnetic and Electrical Fieldsby Near-surface Lateral Inhomogeneities.Acta Geodaet.,Geophys.et Montanist.Acad.Sci.Hung.Tomus,1976,11(3-4):447-483
[24] Philip E. Wannamaker, Gerald W. Hohmann and Stanley H. Ward.MagnetotelluricResponses of Three-dimensional Bodies in Layered Earths. Geophysics, Vol.49, No.9,September1984:1517-1533
[25] P.Zhang, R.G.Roberts and L.B.Pedersen.Magnetotelluric Strike Rules.Geophics,March1987,52(3)
[26] J.C.Larsen. Removal of Local Surface Conductivity Effects from Low FrequencyMantle Response Curves. Acta Geodaet.,Geophys.et Montanist.Acad.Sci.Hung.Tomus,1977,12(1-3):183-186