松辽盆地深层火山岩岩性预测方法

PREDICTING METHOD OF THE LITHOLOGIES OF THE DEEP VOLCANIC ROCKS IN SONGLIAO BASIN

基于松辽盆地深层火山岩岩心的密度及磁化率数据,通过模型正演分析可以看出,深层火山岩产生的重、磁异常不大,大多表现为叠加在强背景之上的次级异常,不易识别火山岩体.为此提出了积分-迭代延拓平化曲方法,增强火山岩磁异常信息,均衡磁异常,消除深度影响.通过应用该方法进行深层火山岩岩体圈定及岩性识别表明：该方法能有效增强深层火山岩的磁异常信息;磁异常的斜导数、欧拉反褶积能有效圈定火山岩边界;重、磁反演的视密度、视磁化率及两者的相关系数是识别火山岩的最佳三参数组合;井约束下的神经网络能有效判别火山岩岩性.

Based on the data of the core density and magnetic susceptibility of the deep volcanic rocks in Songliao Basin and moreover through the forward analyses of the model, the gravity and magnetic anomalies of the rocks are not so serious, they show subordinate anomalies that are superimposed on the strong background and result in the difficulty in the recognition of the volcanic rock bodies. Therefore a new method of integral-iterative continuation changed from flat to curve is put forward to enhance and balance the magnetic anomaly information of the volcanic rocks, and finally eliminate the influences of the depth. By means of the application of the method, the delineation of the rock bodies and identification of the lithologies of the deep volcanic rocks show that the approach can effec- tively enhance the magnetic anomaly information; the oblique derivative, Euler deconvolution of the anomaly can effectively delineate the boundary of the rocks ; the apparent density, magnetic susceptibility obtained from the grav- ity and magnetic reversion and their correlation coefficients are the best three combination parameters; the neural network constrained by wells is an effective method to recognize the lithologies of the volcanic rocks.