Study And Application Of Several Probability Statistical Filtering Methods In The Processing Of Gravity And Magnetic Data

Gravity and magnetic exploration,as two important branches of applied geophysics,studies the difference of geological body in density and magnetic and give an explanation,tostudy the crustal structure constructed and exploration and prospecting through the detectionof geological body’s gravitational and magnetic fields. However, the geological condition ofnature is more complex. The physical properties parameters of the various geological bodiesare inhomogeneities, irregular shape. Sometimes, there will be multiple geological bodyinterpenetration overlap. In addition, the field data acquisition will also be affected by variousfactors such as human error, system error and so on. So the measurement result is complicated.With the introduction of modern mathematics and computer technology, explorationgeophysics has entered a new period of development. Since the20th century, the field theoryhas become an integral part of the modern methods of mathematical physics. Sixties,mathematical statistics method is introduced to the calculation and interpretation of gravityand magnetic data. Statistical methods are a summary explanation based on the inverseproblems application of random function theory, probability theory and mathematicalstatistics. The potential field is the sum of useful abnormal and random interference. Themathematical basis of statistical methods is the randomly distributed of the field sourcesformed random distribution of potential field.The principle of probability statistics method has been well-known, but the applied ofthe interpretation and processing of gravity and magnetic data is not very extensive.In this paper, we selected three methods to study: inverse probability method, self-setting filtration, interprofile correlation. Besides, the applied of inverse probability methodand self-setting filtration method in the gravity and magnetic exploration is few. In this paper, we attempt to apply these three methods to the processing of gravity andmagnetic data.Contents and results of three methods of this paper are as follows:1、Inverse probability method is used to separate geophysical weak anomalies. Theprinciple of method applies Bayesian formula of probability theory (also known as inverseprobability formula). In probability theory, Bayesian formula is known " results "and lookedfor the "cause", that is, when observed the phenomenon F, expect to find the cause of the F.If we want to use inverse probability method, first need to give prior probability, which isdetermined by information theory of entropy. First we should make random interferenceassumption and give test, then determining whether anomaly exists through calculating theprobability of each measuring point of each measuring profile. At the same time, we can getthe distribution curve of abnormal probability. At last, we give the reliability evaluation of the results. On a number of model tests, the inverse probability method can be used to reduce theinfluence of random interference and enhance the effect of useful anomalies.2、Self-setting filtration method is a difference of a constant with the T^2-Hotelling testmethod based on the multivariate statistical theoretically, which subject to F distribution. Theadvantage of this method is that you do not have to know the form and nature of the signal inadvance, but can reduce the influence of interference and enhance the effect of usefulanomalies through adjusting its parameters. Several tests of theoretical models discussed theadvantages and disadvantages of self-setting filtration method and application effect, the lackof which has been improved At last, the conclusion is that self-setting filtration method canremove the random disturbances in the gravity and magnetic background field and play a rolein separate weak anomalies from strong background field.3、Self-autocorrelation filtration in processing of gravity and magnetic data has beendone more researches than the above two methods. In this paper, we selectedself-autocorrelation function method Qin Baohu raised in1991to study. Conclusion is thatthe local field can be separated from the strong background field, but couldn’t reflect thestrength of the local field.4、With self-setting filtration method and self-autocorrelation filtration method wereprocessed gravity and magnetic measured data in Inner Mongolia Daolang and Duge multi-metal mine. Conclusion is that self-setting filtration method enables the contour becomingsmoother and detecting the weak anomalies. Self-autocorrelation filtration method canseparate the weak anomalies from strong background, but can’t truly reflect the strength oflocal field. At the same time, the result is likely to exist part of the false anomalies.