#TITLE_ALTERNATIVE#

• 主要作者: Wigjokentjana (NIM 22397018), Bintang
• 格式: Theses
• 語言: Indonesia
• 在線閱讀: https://digilib.itb.ac.id/gdl/view/7441
• 機構: Institut Teknologi Bandung

ABSTRACT:<br /><br /> The transition zone between land and sea is one of the most challenging environments for the seismic method. Frequent changes in acquisition parameter and in near-surface conditions mean that considerable effort is required to obtain an image of the subsurface. One issue that often rises during the processing of such data is equalisation of the phase characteristics in different parts of the survey. Without phase equalisation it is difficult to perform integrated processing and interpretation of data from the survey as awhole.<br /><br /> When there is a substantial overlap between different portions of data they can be phase tied by conventional least squares shaping. Otherwise it must be removed by statistical wavelet processing. However standard deconvolution techniques, based on the minimum phase assumption, cannot by themselves achieve a common phase condition.<br /><br /> Statistical methods of deconvolution rely upon statistical information about reflection coefficients to separate the wavelet from the reflectivity. Most seismic deconvolution methods used in industry today are based upon second-order statistics and minimum-phase assumption. Since the mid-1980s a number of author have used higher-order statistics to solve the deconvolution problem. These methods have an additional requirement that the reflectivity is a non-Gaussian random process. These methods have their basics in the theory of cummulant and cummulant spectra.<br /><br /> The theory of cummulant in fourth-order was applied to estimate wavelet at transition zone. Observed area are zone that changed from land to sea without overlapping shotpoint when data acquisition and WVEST one of the Geovecteur Module have applied as a tool.