Manyetotelürik tepkilerdeki topoğrafik etkilerin düzeltilmesi ve Batı Anadolu verilerine uygulanması

Abstract

important. For the TM mode, the apparent resistivity has been calculated less than its real value on the hills however, it has become greater in the valleys. For the TE mode this effect has been reversed. Magnetotelluric resistivity and phase values can be represented versus frequency; on account of this the terrain effect has been also investigated versus frequency. It has been observed that, at the lower frequencies, the effect for TM mode has increased and for TE mode it has decreased. Furthermore, in TM mode the effect is frequency dependent at the higher frequencies although it becomes static at the lower frequencies. All our findings agree with the results which the other researchers obtained by different methods and models. The terrain effect is removed by the distortion stripping technique. The method has been proposed previously by Larsen (1977), to correct the effect due to near surface inhomogeneities and then it has been applied to remove the topographic distortions by Mozley (1988), Jiracek et al (1989) and Chouteau and Bouchard (1988). The authors realized the correction on the model data, however they used homogeneous subsurface resistivity to calculate the distortion coefficients. In this study, our goal is to correct the MT response data measured from West Anatolia and to discuss how the topographic structure of Anatolia distorts the MT measurements in Turkey. It has been recognized that the homogeneous subsurface resistivity approach is insufficient to correct the real (measured) data from the regions having complex geology, that is why the ID subsurface model approach has been proposed. This new ID model approach has provided more reliable results and did not distort the real subsurface structures like homogeneous subsurface approach has done. There is a MT measurement site in Western Taurides at the same location with a drill hole (Demre-1) site (6200m). This supplied an opportunity to check the ID model approach. Comparison between the best fit models for the uncorrected and corrected data indicates a clear difference for the layer thicknesses of the models. For example, the thickness of the second layer which is found 1594 m before correction has become 310 m. after the correction. The real thickness (from litho logical log) for this layer 123is 250 m. The correlation between the lithological log obtained from the Demre-1 bore hole which is drilled at the same point and the best fit model for the corrected data shows a very good agreement. This result show the success of the new correction technique further it puts forward the importance of terrain correction for the data measured in Anatolia on very complex structures. The correction applied to another data set from Northwest Anatolia (D-Line) which is measured by İlkışık (1987). After the correction a good agreement with the known structures of the surface geology has been obtained. 2D modelling of terrain corrected data has revealed a very conductive zone (1-10 ohm.m) at 26 km depth at south of Pamukova. This result may indicate an upwelling of the asthonosphere or a partial melting in the crust. According to the model, four layer geoelectrical structure of the region from Ucgaziler to Black Sea shows an agreement with the opinion which states that the Istanbul sequence is not the real substratum of the region (Okay et. al 1994). As a result the features of the terrain effect has been described, the topographic correction has been discussed, and it has been shown that ID approximation is more efficient and necessary in the case which pertains to the subsurface having a layered structural component. The correction technique has been applied to a real data set measured in West Anatolia and hence a more reliable 2D model has been suggested which fits the corrected data. 124important. For the TM mode, the apparent resistivity has been calculated less than its real value on the hills however, it has become greater in the valleys. For the TE mode this effect has been reversed. Magnetotelluric resistivity and phase values can be represented versus frequency; on account of this the terrain effect has been also investigated versus frequency. It has been observed that, at the lower frequencies, the effect for TM mode has increased and for TE mode it has decreased. Furthermore, in TM mode the effect is frequency dependent at the higher frequencies although it becomes static at the lower frequencies. All our findings agree with the results which the other researchers obtained by different methods and models. The terrain effect is removed by the distortion stripping technique. The method has been proposed previously by Larsen (1977), to correct the effect due to near surface inhomogeneities and then it has been applied to remove the topographic distortions by Mozley (1988), Jiracek et al (1989) and Chouteau and Bouchard (1988). The authors realized the correction on the model data, however they used homogeneous subsurface resistivity to calculate the distortion coefficients. In this study, our goal is to correct the MT response data measured from West Anatolia and to discuss how the topographic structure of Anatolia distorts the MT measurements in Turkey. It has been recognized that the homogeneous subsurface resistivity approach is insufficient to correct the real (measured) data from the regions having complex geology, that is why the ID subsurface model approach has been proposed. This new ID model approach has provided more reliable results and did not distort the real subsurface structures like homogeneous subsurface approach has done. There is a MT measurement site in Western Taurides at the same location with a drill hole (Demre-1) site (6200m). This supplied an opportunity to check the ID model approach. Comparison between the best fit models for the uncorrected and corrected data indicates a clear difference for the layer thicknesses of the models. For example, the thickness of the second layer which is found 1594 m before correction has become 310 m. after the correction. The real thickness (from litho logical log) for this layer 123

SUMMARY THE TOPOGRAPHIC CORRECTION FOR THE EFFECTS ON THE MAGNETOTELLURIC RESPONSE FUNCTIONS AND AN APPLICATION TO THE HEST ANATOLIAN DATA. The measured magnetotelluric field components on the earth surface are affected by the surface undulations (topography). That is why before the modelling (and interpretation) the terrain effect on the data should be removed; otherwise, it is not possible to reach a correct subsurface model. The aims of this study are to study the topographic effect, to calculate the terrain correction using numerical methods tool, to compare our results with the former researchers' results, to apply the correction method to the real field data and to remove the effect f rom West Anatolian data, and hence to fit a model to the corrected data. Description of the terrain effect and the cases whereby the distortion is more effective are useful in defining the correction technique and in applying it properly. in this study, first, the terrain effect will be described and as a second step the data will be corrected using the distortion stripping techniques. The height and the slope of the topographic feature, the distance between the measurement sites and skin depth of MT signals strongly affect the quantity of the distortion. For several slope and height values the distortions on MT data is investigated as compared with the ratio between the distance from topographic breaks and skin depth. The terrain effect increases by the increasing slope and height and it decreases by increasing distance from the topographic discontinuities. Another result of this investigation is that the quantity of the distortion in the valleys is less than that on the hills. in the case of the electric field which is measured perpendicular to the terrain discontinuity (TM mode), the measurements are strongly affected by the surface undulations. On the contrary, for the electric field measured parallel to the topography (TE mode), the distortion is less 122important. For the TM mode, the apparent resistivity has been calculated less than its real value on the hills however, it has become greater in the valleys. For the TE mode this effect has been reversed. Magnetotelluric resistivity and phase values can be represented versus frequency; on account of this the terrain effect has been also investigated versus frequency. It has been observed that, at the lower frequencies, the effect for TM mode has increased and for TE mode it has decreased. Furthermore, in TM mode the effect is frequency dependent at the higher frequencies although it becomes static at the lower frequencies. All our findings agree with the results which the other researchers obtained by different methods and models. The terrain effect is removed by the distortion stripping technique. The method has been proposed previously by Larsen (1977), to correct the effect due to near surface inhomogeneities and then it has been applied to remove the topographic distortions by Mozley (1988), Jiracek et al (1989) and Chouteau and Bouchard (1988). The authors realized the correction on the model data, however they used homogeneous subsurface resistivity to calculate the distortion coefficients. In this study, our goal is to correct the MT response data measured from West Anatolia and to discuss how the topographic structure of Anatolia distorts the MT measurements in Turkey. It has been recognized that the homogeneous subsurface resistivity approach is insufficient to correct the real (measured) data from the regions having complex geology, that is why the ID subsurface model approach has been proposed. This new ID model approach has provided more reliable results and did not distort the real subsurface structures like homogeneous subsurface approach has done. There is a MT measurement site in Western Taurides at the same location with a drill hole (Demre-1) site (6200m). This supplied an opportunity to check the ID model approach. Comparison between the best fit models for the uncorrected and corrected data indicates a clear difference for the layer thicknesses of the models. For example, the thickness of the second layer which is found 1594 m before correction has become 310 m. after the correction. The real thickness (from litho logical log) for this layer 123is 250 m. The correlation between the lithological log obtained from the Demre-1 bore hole which is drilled at the same point and the best fit model for the corrected data shows a very good agreement. This result show the success of the new correction technique further it puts forward the importance of terrain correction for the data measured in Anatolia on very complex structures. The correction applied to another data set from Northwest Anatolia (D-Line) which is measured by İlkışık (1987). After the correction a good agreement with the known structures of the surface geology has been obtained. 2D modelling of terrain corrected data has revealed a very conductive zone (1-10 ohm.m) at 26 km depth at south of Pamukova. This result may indicate an upwelling of the asthonosphere or a partial melting in the crust. According to the model, four layer geoelectrical structure of the region from Ucgaziler to Black Sea shows an agreement with the opinion which states that the Istanbul sequence is not the real substratum of the region (Okay et. al 1994). As a result the features of the terrain effect has been described, the topographic correction has been discussed, and it has been shown that ID approximation is more efficient and necessary in the case which pertains to the subsurface having a layered structural component. The correction technique has been applied to a real data set measured in West Anatolia and hence a more reliable 2D model has been suggested which fits the corrected data. 124