Abstract

In the scope of this study, a modeling on the fractal dimension which is a measure of the complexity in the occurrence process of earthquakes and intermediate-term forecasting for the location of expected earthquakes for the Eastern Anatolian region (Turkey) were accomplished. For this purpose, the most suitable and reliable statistical relation was firstly tried to be determined between the seismotectonic b-value and fractal dimension Dc-value for the earthquakes in Eastern Anatolian region. Four different methods were applied for this application as; (1) Least Squares Regression (L2 norm), (2) Least Sum of Absolute Deviations Regression (L1 norm), (3) Orthogonal Regression (Total Least Squares) and, (4) Robust Regression. Also, a composite forecast map by combining the maps of relative intensity and pattern informatics is generated for the forecasting the locations of expected strong earthquakes in the Eastern Anatolian region. Earthquake catalogue used for the analyses was compiled from the Kandilli Observatory and Earthquake Research Institute. Catalogue is homogeneous for duration magnitude, MD and consists of 30462 earthquakes with magnitudes between 1.0 and 6.6 in the period between January 1, 1970 and January 1, 2014. As the primary goal, it is intended to put forward the nature of seismicity which has a fractal structure in space, time and magnitude distributions, as quantified by the fractal dimension Dcvalue and seismotectonic parameter b-value for the Eastern Anatolian region. The Eastern Anatolian region was divided into 19 different seismotectonic sub-regions in order to make a detailed assessment on a regional scale. In order to calculate more up-to-date and reliable statistical relation between two seismotectonic parameters, four different regressions were used. Thus, the relationship of Dc3.070.53*b is computed with a strong negative correlation (r = 0.95) between b-value and Dc-value for the Western Anatolia earthquake distributions. For each regression, following linear relations with their correlation coefficients were estimated: Dc 2.490.34*b, for Robust Regression (r = -0.85) Dc 2.510.35*b, for Orthogonal Regression (r = - 0.89) Although the results are very close to each other, using the Least Sum of Absolute Deviations Regression method, the relationship of Dc2.520.36*b with a strong negative correlation (r=-0.91) is obtained between Dc-value and b-value for the Eastern Anatolian region. This negative relationship is important with respect to seismotectonic and an earthquake risk can be mentioned for the Eastern Anatolian region in intermediate-term. Also, this statistical relation is in accordance with the other regional relationships existing in literature and it can be suggested as more up-to-date and reliable. As the secondary purpose, it is intended to generate a composite forecast map based on the earthquake intensities and pattern informatics for the Eastern Anatolian region. For the analyses, the earthquakes with the cut-off magnitude Mc≥3.4 and with depths shallower than 40 km in time interval between 1990 and 2014 were used. For the regional imaging of the forecasting map, a regional grid of points with a grid of 0.075 by 0.075 was used and it is tried to forecast the locations of earthquakes with MD5.0. In the forecasting time interval between January 1, 1970 and January 1, 2024, the composite forecast map was prepared in order to detect the location of expected strong earthquakes in intermediate-term in the Eastern Anatolian region. In the result of analysis, some areas in the Eastern Anatolian region were detected as hazardous regions in terms of earthquake potential in the next. These regions are in and around Askale fault, the west of Van Lake (between Suphan fault and Ercis fault), around Yuksekova-Semdinli fault zone, around and the north part of Ovacik fault, on and southwest end of the Eastern Anatolian fault and, a part of North Anatolian fault zone between Mus Thrust zone and Pulumur fault. Consequently, it is suggested that a special caution should be given to the anomalies in these regions and it must be evaluated with the other geophysical methods together by monitoring the earthquake activity

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call