استفاده از روش لي- اولدنبرگ در برگردان سه‌بُعدي داده‌هاي مغناطيسي زميني منطقه گزستان (استان يزد)

3D Inversion of ground magnetic data of Gazestan area based on Li-Oldenburg method

هدف از اين تحقيق، بررسي بي‌هنجاري‌هاي مغناطيسي موجود در منطقه گزستان در استان يزد با استفاده از برگردان سه‌بُعدي داده‌هاي مغناطيسي و عرضه يك راهكار جديد براي مدل‌سازي ساختارهاي زيرزميني است. با استفاده از نرم‌افزار Mag3D، كه براساس الگوريتم غيرخطي لي- اولدنبرگ نوشته شده است، امكان مدل‌سازي ساختارهاي زمين‌شناسي با توزيع هندسي پيچيده فراهم شده است. در پايان، توزيع خودپذيري مغناطيسي در هريك از سلول‌ها، شكل سه‌بُعدي توده زيرزميني را تعيين مي‌كند. رهيافت مطرح شده در اين تحقيق، متغير گرفتن ابعاد و تعداد سلول‌ها و پارامترهاي هموارسازي است. براي ارزيابي صحت مدل به‌دست آمده، نتايج با اطلاعات به‌دست آمده از گمانه‌هاي حفاري موجود در منطقه مقايسه شده است كه همبستگي قابل ‌قبولي بين نتايج وجود دارد. نتيجه تحقيق، تشخيص سه بي‌هنجاري نسبتاً بزرگ با خودپذيري بالاي 8/0 در دستگاه cgs در منطقه مورد بررسي است كه در عمق‌هاي 10 متر تا 120 متر واقع‌اند.

One of the appropriate methods for description of geological complexity of earth’s crust is modeling of magnetic data by inversion. After investigations performed about 1D and 2D modeling in recent decades, nowadays most of the investigations about potential field are based on 3D inversion modeling. The goal of the magnetic inversion is to obtain, from the extracted anomaly data, quantitative information about the distribution of the magnetic susceptibility in the ground. Thus it is assumed that the input data to the inversion program is the extracted residual anomaly and programs in the library are developed accordingly. In this study, we present a method for magnetic data inversion to make 3D susceptibility models of an area with a suitable potential of Iron-ore. We have used Mag3D program which uses the nonlinear Li, Y. and Oldenburg, D. W., algorithm. It determines the best model to show susceptibilities of bodies with an iterative approach. In This program susceptibilities are assumed to be small. This means that results will be wrong when susceptibilities are higher than that which causes self-demagnetization. The model is specified using a mesh of rectangular cells, each with a constant value of susceptibility and topography. Furthermore we assume that the remanent magnetization is zero and the magnetization effect is negligible. Thus in this paper the induced magnetization has been used. The inversion is solved as an optimization problem with the simultaneous goals of (i) minimizing an objective function on the model and (ii) generating synthetic data that match observations to within a degree of misfit consistent with the statistics of those data. To counteract the inherent lack of information about the distance between source and measurement, the formulation incorporates a depth or distance weighting term. For improving the results of the inversion method, priori geophysical or geological information have been incorporated. Even in Gazestan area there will be some geological information available in addition to the geophysical data. These constraints can be supplied to the inversion software, with adjustable levels of certainty, via a reference model of expected properties, bounds on the expected properties, or smoothness weights based on positions and orientations of the rocks. The constraints can come from mapping, sampling, analogous areas, or neighboring regions. The depth weighting function is designed to counteract the decay of the potential field response with distance from the source so that all cells have an equal likelihood of containing sources. This is necessary as there is no inherent depth information contained in the observed potential field response and a default solution to the inverse problem would result in a model with sources clustered near the surface. The depth weighting function has the form (Li and Oldenburg, 1996). In later stages of exploration and development additional information from drilling, detailed structural interpretation, trenching, and even preliminary mining will also be available. To test the accuracy of the method, we have compared the results with boreholes data. Finally, the model shows three huge anomalies in this region. The largest anomaly is located in the middle of the area of study with the average thickness of 80 m.