DETECTION OF WEAK ANOMALIES OF GEOPHYSICAL FIELDS BASED ON THE METHODS OF STATISTICAL FILTERING

This paper discusses the problem of detection of weak anomalies and localization of anomalous objects of the north-eastern part of the Dnieper-Donets basin based on statistical methods of filtering and correlation analysis. Investigation of salt tectonogenesis factors in geological environment and their influence on the interpretation of geophysical fields are made. Possible application of adaptive filtering algorithms in order to improve useful information obtained from observed geophysical fields is proposed. The main problems of regional tectonic subdivisions and searches for hydrocarbons caused by the absence of prior i gravimetric data are studied. Basic stages of development and application of probabilistic and statistical methods for data processing of geophysical fields is characterized. The paper analyses methods and algorithms used to determine possible localization of the abnormal objects based on probabilistic and statistical methods. The authors propose basic probabilistic and statistical methods in order to identify and evaluate signals (anomalies). Feasibility of calculation of statistical and physical characteristics (attributes) to resolve a number of problems of regional tectonic subdivision and assessment of structural and tectonic elements is substantiated. Methodic of adaptive filter usage based on the evaluation of the autocorrelation function of the input signal is proposed in both one-dimensional and two-dimensional versions. Filtering analyses of gravitational field in the north-eastern part of the Dnieper-Donets basin is carried out. For the first time, the adaptive filter is applied to analyze gravity data of the north-eastern part of the Dnieper-Donets basin. It is established that at doing analysis of geophysical fields much attention should be focused on the field of extreme values. Their selection can help to solve more effectively current problem of regional tectonic subdivision into fixed area. Extreme value of statistical attribute control regions of disruption of statistical non-stationary of geofields, which in turn are confined to the boundaries of geological or tectonic zones. Regional and local components of the gravitational field are distinguished by the results of adaptive filtering. This creates a significant potential to increase the output of useful information that have observed physical fields. The results obtained shows that this type of filter is extremely effective at its application to the gravimetric data, anomalous effects of which are mainly confined to the salt structures of various types. Adaptive filter options enable to localize the abnormal effects of geological bodies and allocate less background noise power anomalies. Such a problem-solving s can be applied at regional subdivision and searches for hydrocarbons.