Institute of Geophysics, University of TehranJournal of the Earth and Space Physics2538-371X41320150923Seismic activity zoning of Zagros fold and thrust belt using fractal parametersSeismic activity zoning of Zagros fold and thrust belt using fractal parameters3633755360210.22059/jesphys.2015.53602FASomayehKalaneM.Sc., Department of Geology, Faculty of Sciences, Golestan University, Gorgan, IranMaryamAgh-AtabaiAssistant Professor, Department of Geology, Faculty of Sciences, Golestan University, Gorgan, IranJournal Article20140831Earthquakes are a particular concern because of the serious hazard they present. The existence of seismicity patterns such as the foreshock and aftershock clusters, doughnut pattern, seismic quiescence etc show that the earthquake occurrence in the active tectonic regions is not random and clustering phenomena are obvious characteristics of these events in space and time. Fractal analysis, as a statistical tool, has been applied to describe the spatial and temporal distribution of earthquakes (Bhattacharya and Kayal, 2003; Ceylan, 2006). The well-known Gutenberg- Richter (1944) relation also implies a power law relation between the energy release and the frequency of occurrence. It means size distribution of earthquakes is scale invariant and <em>b-value</em> has been suggested as a generalized fractal dimension of earthquake magnitude (Aki, 1981; Turcotte, 1997). Therefore, earthquakes have fractal structure in the distribution of size, space and time. In this paper, the spatial variations of the fractal parameters for the seismicity of Zagros fold and thrust belt, Between 25 to 37 N and 44 to 58 E, have been analyzed. For this aim, we have extracted a homogeneous catalogue with mb ≥ 4.4 from ISC and NEIC bulletins, covering a time period 1975 – 2014. In order to investigate the spatial variations of fractal parameters, the study area is covered by a 0.5◦×0.5◦ grid. Then, fractal parameters consist of the <em>b-value</em>, correlation dimensions of epicenteral (<em>D<sub>e</sub></em>) and occurrence time of earthquakes (<em>D<sub>t</sub></em>) were estimated for each sample volume, data within a fixed radii (75km) centered on every grid node. ZMAP software was used for calculating the parameters at each node. In this paper, two sets of analyses have been done for total and declustered data sets. In the final stage, the spatial variations of <em>b-value</em>, <em>D<sub>e</sub></em>and <em>D<sub>t</sub></em> parameters were drawn as maps. The <em>b-value</em> map of total data set indicates a low value of b in the Zagros-Makran transition zone, probably as a result of the increased applied stress due to an increase in depth of the seismogenic crust and/or an increase of the convergence rate between the Arabia and the Eurasian Plate at this zone. The Kazerun-Borazjan faults as transfer fault zone is characterized with low <em>b –value, </em>too. This map also show low values in the northwest Zagros, where the recent dangerous Murmuri earthquake with Mw 6.2 (NEIC) has been occurred. The interesting result of this research is a high correlation between <em>b-value</em>, <em>D<sub>e</sub></em> and <em>D<sub>t</sub></em>maps. Similar to the <em>b-value</em> map, the <em>D<sub>e</sub></em>- and <em>D<sub>t</sub></em> maps also show anomalous low values in the Zagros-Makran transition zone, Kazerun, Borazjan, Karebas, Sabz-Pushan faults and northwest Zagros, showing strong clustering of events in space and occurrence time.
In order to investigate the background seismicity pattern of the Zagros, same analysis have been done after declustering the catalogue, removing dependent events, using Reasenberg (1985) algorithm. The result of analyses of <em>b-value </em>is almost similar to the previously one, anomalous low values in the Zagros-Makran transition zone, Kazerun, Borazjan, Karebas, Sabz-Pushan faults and northwest Zagros. Since the epicentral and occurrence time distributions of earthquakes are sensitive to clustering of events, the declustering process, as expected, alters the results for two other parameters. The spatial variations map of <em>D<sub>t</sub></em> show almost uniform distribution. Despite this, the <em>D<sub>e</sub></em>- variation map still shows the lower values in the Zagros-Makran transition zone, Kazerun, Borazjan, Karebas, Sabz-Pushan faults rather than other regions. This suggests that the mainshocks occur in the clusters along these main structural trends. Hence the occurrence of smaller earthquakes, coupled with a homogeneous distribution of earthquake epicentres at the more extensive region of Zagros, show that stress released along dispersed and smaller faults. But in the Zagros- Makran transition and Kazerun-Borazjan zones, an increased clustering of epicenter distributions (decreased<em> D<sub>e</sub></em>), are associated with an increase in stress concentration on the main structural trends (such as Oman line and Qatar-Kazerun line). Temporal distributions of earthquakes in these zones are indicated a high clustering degree. This may be due to the occurrence of frequent larger earthquakes with aftershock sequences in this zone.
Generally, Spatial mapping of the fractal parameters found valuable information about the scale invariance property of the seismic activity variations in the region. These results suggest that the fractal approach can be used as a useful tool for assessing seismic energy distribution on seismotectonically active regions.Earthquakes are a particular concern because of the serious hazard they present. The existence of seismicity patterns such as the foreshock and aftershock clusters, doughnut pattern, seismic quiescence etc show that the earthquake occurrence in the active tectonic regions is not random and clustering phenomena are obvious characteristics of these events in space and time. Fractal analysis, as a statistical tool, has been applied to describe the spatial and temporal distribution of earthquakes (Bhattacharya and Kayal, 2003; Ceylan, 2006). The well-known Gutenberg- Richter (1944) relation also implies a power law relation between the energy release and the frequency of occurrence. It means size distribution of earthquakes is scale invariant and <em>b-value</em> has been suggested as a generalized fractal dimension of earthquake magnitude (Aki, 1981; Turcotte, 1997). Therefore, earthquakes have fractal structure in the distribution of size, space and time. In this paper, the spatial variations of the fractal parameters for the seismicity of Zagros fold and thrust belt, Between 25 to 37 N and 44 to 58 E, have been analyzed. For this aim, we have extracted a homogeneous catalogue with mb ≥ 4.4 from ISC and NEIC bulletins, covering a time period 1975 – 2014. In order to investigate the spatial variations of fractal parameters, the study area is covered by a 0.5◦×0.5◦ grid. Then, fractal parameters consist of the <em>b-value</em>, correlation dimensions of epicenteral (<em>D<sub>e</sub></em>) and occurrence time of earthquakes (<em>D<sub>t</sub></em>) were estimated for each sample volume, data within a fixed radii (75km) centered on every grid node. ZMAP software was used for calculating the parameters at each node. In this paper, two sets of analyses have been done for total and declustered data sets. In the final stage, the spatial variations of <em>b-value</em>, <em>D<sub>e</sub></em>and <em>D<sub>t</sub></em> parameters were drawn as maps. The <em>b-value</em> map of total data set indicates a low value of b in the Zagros-Makran transition zone, probably as a result of the increased applied stress due to an increase in depth of the seismogenic crust and/or an increase of the convergence rate between the Arabia and the Eurasian Plate at this zone. The Kazerun-Borazjan faults as transfer fault zone is characterized with low <em>b –value, </em>too. This map also show low values in the northwest Zagros, where the recent dangerous Murmuri earthquake with Mw 6.2 (NEIC) has been occurred. The interesting result of this research is a high correlation between <em>b-value</em>, <em>D<sub>e</sub></em> and <em>D<sub>t</sub></em>maps. Similar to the <em>b-value</em> map, the <em>D<sub>e</sub></em>- and <em>D<sub>t</sub></em> maps also show anomalous low values in the Zagros-Makran transition zone, Kazerun, Borazjan, Karebas, Sabz-Pushan faults and northwest Zagros, showing strong clustering of events in space and occurrence time.
In order to investigate the background seismicity pattern of the Zagros, same analysis have been done after declustering the catalogue, removing dependent events, using Reasenberg (1985) algorithm. The result of analyses of <em>b-value </em>is almost similar to the previously one, anomalous low values in the Zagros-Makran transition zone, Kazerun, Borazjan, Karebas, Sabz-Pushan faults and northwest Zagros. Since the epicentral and occurrence time distributions of earthquakes are sensitive to clustering of events, the declustering process, as expected, alters the results for two other parameters. The spatial variations map of <em>D<sub>t</sub></em> show almost uniform distribution. Despite this, the <em>D<sub>e</sub></em>- variation map still shows the lower values in the Zagros-Makran transition zone, Kazerun, Borazjan, Karebas, Sabz-Pushan faults rather than other regions. This suggests that the mainshocks occur in the clusters along these main structural trends. Hence the occurrence of smaller earthquakes, coupled with a homogeneous distribution of earthquake epicentres at the more extensive region of Zagros, show that stress released along dispersed and smaller faults. But in the Zagros- Makran transition and Kazerun-Borazjan zones, an increased clustering of epicenter distributions (decreased<em> D<sub>e</sub></em>), are associated with an increase in stress concentration on the main structural trends (such as Oman line and Qatar-Kazerun line). Temporal distributions of earthquakes in these zones are indicated a high clustering degree. This may be due to the occurrence of frequent larger earthquakes with aftershock sequences in this zone.
Generally, Spatial mapping of the fractal parameters found valuable information about the scale invariance property of the seismic activity variations in the region. These results suggest that the fractal approach can be used as a useful tool for assessing seismic energy distribution on seismotectonically active regions.https://jesphys.ut.ac.ir/article_53602_283c716d735dfe36f0e18b83b1124bc1.pdf