Elemination of window effect on time signals and its applications on microtremor data

In this study, the problems resulting from the cutting of time data obtained with microtemor method and what the period regarding the obtainment of data must be have been investigated.Data are obtained depending on time or distance in geophysics. These data must be transferred into the frequency medium so that they will be evaluated and interpreted. Because, many events that cannot be noticed in time medium can be readily analyzed in frequency medium. The gathered data, naturally, have been cut with a window in the height of measurement time. This process, in time medium, can be defined as the multiplication of infinite heighted data with a rectangular window function. According to the convolution feature of Fourier Tranform, this multiplication operation refers to convolution in frequency medium. Window function of time medium can cause differences to reduce at the stage of the resolution of data in frequency medium.As known, after the analysis of microtemor data, the dominant period of emission of ground (T), spectral period (Ta, Tb), and the growth of ground ( K) are tried to be determined. The reliability of these determined parameters is of great importance, because buildings are designed in compliance with the ground classified according to these parameters. In the analysis of data with periods of 0.1-2 seconds such as microtemor, the differentiation of traces in the frequencies that are close to each other is quite difficult. It has been observed in the frequency medium that the signals with close periods have no differences due to the convolution of sinc function, which is the spectrum of rectangular function. Therefore, it is compulsory that window effect on data be eliminated in order to make a reliable evaluation.In the study made, how the window effect reduces the difference in close frequencies has been shown and tried to be eliminated. So as to eliminate the effect, reverse convolution method has been used. At the second stage, on the other hand, the real microtemor data taken from that area have been studied and window effect has been aimed to be discharged from the data through the reverse convolution method.The results disclosed in the consequence of the studies, the window length, appiles on signal, during evaluation data should be at least six hundred.