9th Congress of the Balkan Geophysical Society

An evaluation based on Gutenberg-Richter b-value, fractal dimension Dc-value, completeness magnitude Mc-value, annual probability and recurrence time of earthquakes is carried out for Tunceli province of Turkey. Earthquake catalog covers the years between 21 September 1970 and 30 December 2016, and contains 5462 earthquakes with magnitude equal to and greater than 1.1. Magnitude completeness is estimated as 2.8 and b-value is found as b=0.98+/-0.01. This b-value is smaller than average value of b=1.0 given for Gutenberg-Richter relation. So, this low b-value may be an indicator of stress to build up over time at the beginning of 2017. Dc-value is obtained as 1.63+/-0.04 with 95% confidence limits. This high Dc-value indicates that seismicity in Tunceli and vicinity is more clustered at larger scales or in smaller areas. Annual probability of the earthquakes between 4.5 and 5.5 magnitude levels exhibits a value between 1 and 10. Recurrence time of the earthquakes has a value between 10 and 30 years for magnitude levels between 5.5 and 6.0, and a value between 30 and 100 years for magnitude levels between 6.0 and 6.5. Consequently, this part of Turkey may have an earthquake potential for the possibility of intense earthquake occurrences in the next future.

In seismic data processing, improving the temporal resolution of seismic data is generally provided by broadening the spectral bandwidth of seismic data and this is basis for discovering the hidden and more detailed subsurface structural and stratigraphic information. In this paper it is presented a methodology which combines the short time Fourier transform (STFT) and cepstrum analysis called logarithmic STFT to extend spectral bandwidth and so improve temporal resolution of seismic data. The performance of the proposed method has been tested and shown on a seismic wavelet, 1D synthetic thin bed model, 2D synthetic wedge model and field seismic section. Application of the method requires only a spectral decomposing window function and its length. Herein, a Gaussian window was used for the spectral decomposition because of its smooth flanks. The results show that the proposed method can considerably increase the temporal resolution of the seismic data and may be useful to interpret the seismic data from thin layered sedimentary medium.

The seismic hazard estimation involves the evaluation of the seismicity within a set of seismogenic zones, taken into consideration by using best suitable models among which the Poisson distribution is the most commonly used statistical models. Gutenberg-Richter (GR) relationship “logN = a−bM” states that the logarithm of the number of earthquakes is linearly dependent on the magnitude of the earthquakes. The constants of GR relationship are essential for hazard assessment and they can be used to predict earthquake occurrences, recurrence periods and seismicity at a given region. In this study, so-called essential parameters for the hazard assessment, a- and b-values of GR relationship for selected characteristic 7 seismogenic sources and all region in the Eastern Black Sea region near the North Anatolian Fault Zone (NAFZ) are calculated with linear least square method (LLS) and maximum likelihood method (ML) and, the probability of earthquake occurrences and recurrence periods by using Poisson model for these sub-regions are determined. According to the seismic risk estimations, the highest earthquake occurrence probability and recurrence time of MS >7.0 in the next 100 years is 69% and 85 years for sub-region 6.

This study presents recent progress “Part-2” in research of Lake Erçek Seismic Survey (LESS, 2015) Project to analyze earthquake faulting characteristics and source mechanisms of the selected aftershocks recorded at KOERI network (2011–2012) in the Lake Erçek Area (Eastern Anatolia) by using Centroid Moment Tensor (CMT) Inversion and Stress Tensor Analysis of P- and S-waves with a support of the high-resolution shallow seismic reflection data and field survey observations from Lake Erçek Basin.

The multi-beam echosounder data were collected in the southern Black Sea shelf between 2002 and 2008 by the Turkish Navy, Department of Navigation, Hydrography and Oceanography and studied under the research project of Scientific and Technological Research Council of Turkey (Project no. TUBİTAK-114Y057). Multi-beam ASCII data were exported to vector points. The data covers an area of 2.59 km2 and elevations range from −2 to −503 m. The multi-beam points were interpolated into DEM of pixel size 5 m, using Topo-to-Raster interface of ArcGIS software.

The bottom morphology offshore Cide-Sinop shelf is quite smooth plain with an approx. 25 km wide and −100 m water depth. The continental shelf edge is around −120 m isobath. Slope angles from 0 to 1 degree at the shelf plain, increases about to 40 degrees along the shelf slope. Some landslides are mapped in this zone. The shelf plain is widely characterized by sand dunes with an average height of 10 m. These hills are separated from each other by flat plains. The hills and plains are interpreted as the sand dunes and submarine plains which probably evolved during aeolian processes. These results are thought that the shelf plain was once a terrestrial landscape.

To better understand the deep structure of the Eastern Pontides orogenic belt (NE Turkey), we have just analyzed the teleseismic-records of the KTUT station as an initial study. For this purpose, P and S Receiver Functions are calculated, and then inverted jointly together with teleseismic P and S traveltime residuals. The inversion is performed using a simulated annealing technique. No high-quality signal is observed at the KTUT station for the Ps converted phase from P410s and P660s km discontinuity. An early arrival is observed at this station for P410s which is ∼0.5 s than the IASP91 global model. The crust has a thickness of about 34 km, and the high-S-velocity mantle lid is beneath the crust at a depth of about 65 km. The velocities of Vs and Vp in the crust are close to Vs= ∼3.61 km/s and Vp= ∼6.37 km/s, respectively. The S wave velocity decreases to ∼4.22 km/s at the low-S-velocity zone beneath the crust, and it reaches to ∼4.65 km/s at upper mantle. Consequently, the crust and upper mantle thicknesses derived from the receiver functions analyses are in good agreement with results from other studies.

Low Frequency-Passive Seismic Method (LF-PSM) is a new technique that uses low frequency noise to determine the reservoir’s existence in oil and gas fields. As an indicator of the hydrocarbons, such method can be used to find the existence of the reservoir. In this method, spectral characteristics of the noise induced by reservoir are investigated where the first attribute quantifies the spectral energy in the vertical wavefield component and the second attribute quantifies the maxima in the vertical-to-horizontal spectral ratios. An increase in the spectral energy, so-called hydrocarbon microtremor, generally exists in the frequencies between ∼1 and ∼3 Hz. In addition to the conventional exploration methods, it can be applied as a complementary approach.

In this study, we consider a one-dimensional (1D) earth model with variable three electromagnetic properties. There have been a number of 1D GPR forward modeling available in the literature (e.g. Annan and Chua, 1992 ; Bano, 1996 ; 2004 ; Lazaro-Mancilla and and Gomez-Trevino 1996 ; Hunziker et al., 2015 ). For instance, Lazaro-Mancilla (1996) studied a 1D GPR forward modeling based on a plane waves an n-layer earth model. Hunziker et al. (2015) developed a general EM forward modeling code including 1D GPR based on Green’s function method. In this study, we present 1D finite difference (FD) method to get a model response of n-layer earth.

Here, we also consider inversion of 1D GPR data. As an inversion type, we use simulated annealing (SA) as a global inversion method. 1D GPR inverse problem was studied by Lazaro-Mancilla and Gomez-Trevino (2000) to estimate dielectric constant, conductivity and magnetic permeability from a GPR trace. Huang and Zhang (2014) applied a spectral inversion method on 1D GPR data using a modified stochastic hill-climbing method to invert model parameters. Here, the layered responses of an earth model with three variable properties were inverted including the thickness of layers using SA optimization method to get a fullwaveform response.

Soil amplification is an important parameter to estimate the dynamic soil-structure interaction under earthquake loadings accurately. So, a good estimation of the soil amplification is required because the soil conditions of local sites have significant influences on the vulnerability of engineering structures. Although the soil amplification is directly related to impedance ratio between the rocks and soils, it is generally calculated trough empirical formulas based on Vs30 value, an average shear wave velocity (Vs) up to 30m depth. However, when the soil column includes rock units (Vs□700m/s) within the 30m depth, the empirical formulas lose their validity. In this study, we try to explain this issue and compare soil amplification values calculated by using Vs30 and impedance ratio for synthetic models and field cases which include rocks unit within the 30m depth. In synthetic models, the soil amplification for four scenarios for two-layered medium is calculated by most known empirical formulas and compared with that of impedance ratio. In field cases, we compare soil amplifications for shallow and deep bedrock cases. Synthetic and real results show that soil amplifications calculated by empirical formulas couldn’t characterize the real ground conditions, especially in the case of the hard-solid units in shallow depths.

We applied EMD for noise cancelation of potential field data, incorporating a new stopping criterion for choosing IMFs to subtract during the filtering procedure. The proposed stopping method was tested on theoretical density model.