| Airborne gamma-ray spectrum survey is a method which sets the airborne gamma spectrometer in an aircraft or other vehicles to identify the radioactive elements, such as potassium, uranium and thorium etc. It is not only used to search for radioactive deposits, radioactive-elements-related potassium, rare elements and polymetallic deposits, but also mostly applied in geological mapping, solving hydrogeological problems and environmental radiation monitoring. It has been extensively utilized, because of its high efficiency, low cost, and without the restrictions from terrain. Based on data handling, the extraction of abnormal information is the foundation of interpreting regional geology characteristics, potentiality of prospecting and evaluating the quality of the airborne gamma-ray spectrum survey. Generally speaking, airborne gamma spectrum, which follows the geochemical abnormal information extraction method, is based on traditional statistics. It requires the data must follow normal distribution or lognormal distribution, which is hardly realized in practice. To ensure the accuracy of the applied model, it is necessary to eliminate the bad original data cyclically until there is no outliers. However, after eliminating the radioactive geology characteristics cannot be reflected by the incomplete data. In addition, the statistical method evaluates the abnormal information of airborne gamma spectrum only via statistical parameters, such as the average value and standard deviation etc., without the spatial frequency domain properties and geophysical field characteristics of the data themselves. Based on the points above, in this paper, several extraction methods of anomalies airborne gamma energy spectrum information are presented, which depend on spatial frequency domain filtering and geophysical field decomposition. The following studies are carried out:(1)Handle, identify and extract the airborne gamma energy spectrum anomalies information via the traditional statistical methods. Analyze and evaluate the anomalies in geoscience.(2) The one-dimensional Fourier transform method, which is frequently used in signal processing, is introduced into the data processing. Original data are divided again from the frequency domain point of view. The background and the airborne gamma energy spectrum anomalies according to the regular pattern and the difference. Meanwhile, get the spatial frequency characteristics of the original data via the spatial Fourier transform, which has the geophysical field background recovered in spatial scale. Based on above, extract the anomalies information of the airborne gamma energy spectrum.(3)The wavelet transformation, which has a characteristic of non-global time-frequency transformation, is utilized to divide the original data. And then, build an extraction from inverse wavelet transform to realize the noise reduction in high frequency.(4) The multi-fractal features of airborne gamma spectrum is studied to divide the cut off sections of the power spectrum, which is the theoretical basis of classifying the background of the airborne gamma spectrum, regional and local anomalies.(5) The average gradient of maximum entropy is applied to get rid of noise from the original data, and the mean entropy is used to extract the anomalies hiding in the measurement background caused by statistical fluctuation.The following results are obtained via the above methods in the analysis of airborne gamma spectrum, such as:(1)The area of uranium concentration anomaly is 7.11 km2, accounts for 22.97% of the whole survey area, through the general method. The range is 32.39 and contrast value is 4.35. The granite can be easily distinguished from others because of varieties of potassium, uranium and thorium in different kinds of rocks, even identifying the metallic mineral. While, the banded distribution of the false regional anomaly cannot be eliminated.(2)The one-dimensional Fourier transform can realize Low-pass and band-pass filtering, which can distinguish the to the uranium anomaly from background. The area of uranium concentration anomaly is 4.48 km2, accounts for 14.46% of the whole survey area. The range is 25.91 after interpolation. This method reduces the “Candied fruit string” effect caused by flight error. Because of the direction of Fourier transforming is along the survey lines, the false anomaly cannot be eliminated thoroughly. Spatial Fourier transform can effectively filter out the banded false uranium anomaly along the flight route. The area of anomaly declines to 1.17% and the range is 11.59. Moreover, the anomalies related to mineralization can be kept in an original degree. But, the frequency spectrum leakage may produce the sidelobe frequency, which can lead to abnormal higher-value near the central survey region.(3)The wavelet decomposition, based on the Mallat algorithm, can maintain the original total rate of airborne gamma spectrum. On the premise of no loss of data overall space characteristics,the Mallat algorithm can increase the processing efficiency via data compression. The characteristic of the compressed data is affected by the wavelet basis function adopted. The areas of anomaly, extracted by Haar, Daubechies(db N) and Bior biorthogonal wavelet basis function, are 7.27%, 6.67% and 9.47%, respectively. And the ranges calculated are 99.03, 100.92 and 105.73. All of above three functions can be applied to identify the anomalies related to mineralization and high-value lithologic anomaly. The residual error tendency of the original data, processed by wavelet decomposition, is rebuilt as a flatly changed curve, which can position the lead-zinc and iron mine in the survey region accurately. While is not very effective for distinguishing the lithology.(4)The multi-fractal method is utilized to filter the power spectrum to distinguish the anomaly from the background obviously, and even to eliminate the banded false anomalies caused by flight in high frequency. The distribution characteristic of airborne gamma energy spectrum data, in spatial domain, are summarized as: both the anomalies related to mineralization and high-value lithologic anomalies exist in all frequency domain. They account for 13.64%、6.41% and 1.55%, respectively, can be kept in good shapes. Secondly, most of the banded false anomalies exist in the low and medium frequency. There is little signal response in the high frequency space. So, in the high frequency, the band-pass filtering method can effectively eliminate the banded false anomalies.(5) The anomalies related to metal mineralization can be retained as a high-value after processing by sample entropy. The banded false anomalies, in the west of the survey region, are calibrated by two-dimensional matrix. The anomalies caused by granite, in the south of survey region, appears in the dense point distribution with the low and medium intensity, which is result of a smooth survey region background. The average entropy method has several advantages, such as high system sampling rate, extensive survey region, good continuity, etc. So the airborne gamma spectrum data reflects much more information than on the ground. The width of average entropy window, represented by “n”, is far greater than ground gamma spectrometry survey. Ultimately, the value of “n” is identified as 30 to obtain the best results by computational experiment repeatedly. |
PDF Full Text Request