Short Period Spectral Acceleration Zonation of Tehran a Comparison between Slip and Activity Rates Data’s

Corresponding Author: Hadi Jarahi Department of Geosciences, North Tehran Branch, Islamic Azad University (IAU), Tehran, Iran Email: hadijarahi@gmail.com Abstract: This study presents seismic hazard analysis and provides spectral iso acceleration maps based on slip rate (SR) and also activity rate (AR) of the faults. Nowadays modern seismic design of structures is complicated so special attention to the nature of the seismicity in active seismic zones like Tehran is required. Many studies, based on seismic hazard analysis have been done in Tehran. However, none of these studies provided comparison between SR, AR and spectral seismic zonation. In order to reduce the loss of life and property in metropolitan Tehran such studies are undeniable. The process of this study is identification of seismic sources, estimation of seismic parameters and interpretation the results of paleoseismology. Seismic hazard assessment for grid points has been done based on seismic sources and the determination of shear wave velocity. Based on these studies, the shear wave velocity for the upper 30 m has been prepared as a basic parameter for seismic hazard analysis. In addition, iso acceleration spectral maps were prepared based on the SR and AR, for 475 years return period. Referring to paleoseismology studies, it is apparent that some faults in South and South East of Tehran should be considered as an ancient coast line; therefore, they were excluded from the seismic studies. As a result, acceleration in South East Tehran has dramatically declined. In a general comparison, it can be stated that the acceleration based on data from the SR is higher than the AR for the period of 0.4 seconds. The results are equal, in period of 0.4 s and after that the acceleration based on AR is higher than SR.


Introduction
Tehran area with a population of over 15 million is one of the most active seismic zones in the Middle East. Iran plateau is situated between the interaction of Arabian plate and Eurasian plate (Copley and Jackson, 2006) which is waiting for a huge catastrophe this collision area stretches from west of Turkey to east of Iran and Alborz Mountain forms the most part of it. Iran is situated between two old continents Eurasia; in north and Africa-Arabia in south and behaves like a brittle platform and is known for its tectonics. Awareness of the dangers of such earthquakes in this area needs study seismic hazard analysis. Some seismic hazard studies have been done based on activity rate in this area, but none of them considered paleosesimology or slip rate (Gholipur et al., 2006;Mansouri et al., 2010;JICA, 2000;Mansouri and Ghafory-Ashtiany, 2009;Jafari, 2003a;2003b;Jafari et al., 2005;Ghodrati et al., 2003;2010;Majd Jabari and Zare, 2007). The list of events used in this study included from 30000 BC to 2014 (Ambraseys and Melville, 1982;Berberian, 1994;Engdahl et al., 2006;Moinfar et al., 1994;Berberian, 2014; ICS center) that according to the long recurrence periods earthquake in this region no longer properly reflects the seismicity and seismotectonics nature of seismic sources (Gholipur et al., 2006;Ghodrati et al., 2010).
In other words, considering only paleoseismology for Tehran faults may hide the effect of the recent seismicity of the sources. Therefore, in this metropolis, studies based on both types of data, based on seismic slip rates and activity rate seem to be useful. This research is about a comparison between the results of activity rate and slip rate, by using the spectral seismic zonation of acceleration in Tehran (Pitilakis et al., 2006;Borcherdt, 1994;Zaslavsky et al., 2009). In this study, the seismic sources have been defined (Nazari, 2005;Hessami et al., 2003;Abbasi and Shabanian, 1999;Berberian, 2014;Berberian et al., 1985). Earthquake catalogue of Tehran and adjacent areas was updated and seismic parameters have been estimated (Gutenberg and Richter, 1956;Kijko, 2010;2012) then slip rate was determined for each seismic source, based on paleoseismology. To determine the distribution of shear wave velocity, in the top 30 m of soil, (Vs 30 ) as the key variables in the calculation, Vs 30 distribution map was prepared. After study the tectonics (Shoja-Taheri et al., 2010) and attenuation relations (Douglas, 2011, five New Generation Attenuation relations (NGA) have been selected and considered for this study (Chiou and Youngs, 2008;Campbell and Bozorgnia, 2008;Idriss, 2008;Abrahamson and Silva, 2008;Boore and Atkinson, 2008). Finally, regular grid points were selected and spectral acceleration seismic zonation, based on the slip rate and activity rate, were considered.

Geographical Location and General Tectonic
The considered region covers a quadrangle, limited by 35° 33' 58" N to 35 50'6"N and 51 03' 34" E to 51 35' 48"E including North of Iran in Tehran province. Active tectonics plays an important role in shaping this part of the Iranian plateau, between Eurasia and Arabian plate (Berberian, 1983;Berberian and King, 1981;Berberian et al., 1982) and shows Convergence of about 25 mm per year in direction of north, north eastsouth and south west (Sella et al., 2002). Current physiographic and active morphotectonic in this region, including high mountains, fertile plains, springs of water and active faults in the vicinity of the mountain passages, are caused by this compression. Measured current changes by GPS, in crust of Iran, in period of two years (Vernant et al., 2004) and six years (Masson et al., 2007) reported historical earthquakes. Till 1755 morphotectonic evidences show that the faults in this area are active, hence knowledge of seismic design variables for retrofitting this metropolis has a great importance (Fig. 1).  (Nazari et al., 2010); Meizoseismal area of earthquakes and active faults are modified from (Berberian et al., 1985;Berberian, 2014)

Seismicity Variables
These variables are generally divided into two parts; slip rate and activity rate. The estimation of these variables will be discussed (Golabatunchi, 2013).

Slip Rate
Many researchers in recent decades such as Nazari (2006;Nazari et al., 2009;2010;Ritz et al., 2006;2012) and others have been studied about the paleoseismology and historical events in Tehran region ( Table 1). The review of existing research results together, with aleoseismology studies, have been considered in this study. The results of such studies can be divided into two main parts in the following. The first part is about the date and magnitude of the pre historical events and the second part shows slip rate of fault, based on millimeter per year (during the active time). Normally, geochronology data are associated with errors and in some cases; geochronology is not useful for some events. Also magnitude of the each seismic event could not be exactly determined. Therefore, these earthquakes cannot be used in statistical analysis. However, since the amount of displacement along the fault is known, the total slip rate can be calculated with good accuracy. Hence the study of slip rate could be useful and important.

Activity Rate
A uniform catalog of earthquakes containing prehistoric, historical and instrumental events covering the period from 30000 BC to 2014 is used. The earthquake database is mainly compiled from ISC and USGS/NEIC for the modern instrumental time period , (also Engdahl et al., 2006;Moinfar et al., 1994;Zare et al., 2014) and the catalog of earthquakes provided by Ambraseys and Melville (1982) and Berberian (1994) is the basic source of parameters for the historical (before 1900) and early instrumental  time periods. Prehistoric earthquakes are given from Ritz et al. (2012) paleoseismological studies ( Table 2).
The catalog of earthquakes has been made uniform using the relationships between M s and m b defined by Mirzaei et al. (1998) for Alborz-Azerbaijan seismotectonic province. Activity rate shows the number of earthquakes of a certain magnitude over a year and will be calculated based on statistical analysis of the seismic data, in a specific region. All aftershocks and foreshocks were detected and eliminated from the catalogue (Powell and Duda, 1975;Keilis-Borok et al., 1972;Gardener and Knopoff, 1974). Since we encountered an incomplete earthquake catalog in the study region, the procedures introduced by Kijko and Sellevoll (1992), which permit incorporation of magnitude uncertainty to estimate seismicity parameters from incomplete data files, are applied to the uniform catalog of earthquakes for estimating the seismicity parameters. The seismicity parameters a and b and activity rate were calculated for each seismic source (Kijko and Sellevoll, 1992;Kijko, 2010;2012). Results of the estimation of seismicity parameters in the study area are presented in Table 3. Seismic sources have been delimited mainly based on the fault extent, seismogenic crust (a part of the earth crust in which large earthquakes usually originate) and mechanism of earthquake faulting or a type of active faults. The estimation of maximum magnitude in potential seismic sources is usually according to the features of seismic activity and tectonic analogy.

Seismic Source
To identify the seismic sources, satellite images, seismotectonic maps and geomorphology studies were considered (Nazari, 2005;Hessami et al., 2003;Abbasi and Shabanian, 1999). Rupture length of each fault has been estimated (Ambraseys and Melville, 1982;Zare, 1995) based on empirical relations that have a good connection with seismicity and also seismicity of each source has been calculated (Ambraseys and Jackson, 1998;Wells and Coppersmith, 1994;Nowroozi, 1985;Zare, 1995). Other variables such as magnitude of completeness (Willemann, 1999), maximum magnitude (Kijko, 2004) and rupture variables (Wells and Coppersmith, 1994) are estimated and have been shown in Table 3.

Attenuation Relations
Different attenuation relations were studied (Douglas, 2011), based of geology, seismotectonic and tectonics characters of the region (Shoji-Taheri, 2010) and finally five New Generation Attenuation relations (NGA) have been selected and considered for seismic hazard assessment in this area (Chiou and Youngs, 2008;Campbell and Bozorgnia, 2008;Idriss, 2008;Abrahamson and Silva, 2008;Boore and Atkinson, 2008).

Shear Wave Velocity Distribution (Vs 30 ) in Tehran
Geology and shear wave velocity in Rock and Alluvium, especially in upper 30 meters, play an important role in seismic hazard studies. To determine the distribution of shear wave velocity in the Greater Tehran, a down hole shear wave measurements data from last studies: Jafari et al. (2002) and Tehran Municipality (Appendix 1), were collected. All points were located by GIS and the Vs 30 contours were drawn.

Spectral Acceleration Zoning
Normal zoning and spectrum zoning is largely similar to each other, except that in the spectral method, acceleration is calculated for periods higher than zero (Pitilakis et al., 2006;Borcherdt, 1994;Zaslavsky et al., 2009). Based on this method and other studies (Jarahi et al., 2013;Jarahi, 2011), spectral acceleration zoning in short periods of 0, 0.2, 0.5 and 1 s, based on probability of exceedance in 100 years (McGuire, 1995), were calculated for Tehran Region. For this purpose, 400 of points with the distance of 1km have been considered (Fig. 2). Horizontal acceleration has been calculated according to attenuation relations and Vs 30 for each point. Calculation has been done in two parts, based on activity rate and slip rate (Fig. 3). At a glance, it can be clearly seen that the greatest danger is on the north east of Tehran, where there is a sharp bend in the region of North Tehran fault (Madadi, 2012). Most of metropolitan Tehran residential areas are located in the southern and center part. The acceleration is low to moderate in this regions and this is against of previous studies (Gholipur et al., 2006;Mansouri et al., 2010;JICA, 2000;Mansouri and Ghafory-Ashtiany, 2009;Jafari, 2003a;2003b;Jafari et al., 2005;Ghodrati et al., 2003;2010). As mentioned, active faults in the south and southeast of the region (North Rey, South Rey and Kahrizak Fault) were the beach line and never considered as the seismic source, consequently the southern and south-east and the vast majority of residential metropolitan Tehran has been placed in low-risk areas. Therefore seismic design standards such as the 2800 Code should be reformed. Seismicity rate, after removal of mentioned faults in this area, remains constant in sequence most probably the attributed events to these faults are related to the other seismic sources, like North Tehran, Firoozkooh and Mosha faults. Actually, the seismic share of other faults has been increased dramatically and the acceleration related to other sources has been increased. This results in the more residential areas of northwest of Tehran, in the area, with high relative risk, is justified. Comparing results based on activity rate and slip rate shows that for the period of 0 to 0.4 s maximum and minimum spectrum acceleration is higher for the data based on slip rate comparing with the data of activity rate. For the period of 0.4 s spectrum acceleration is equal for both data and after this period Spectral acceleration values of the slip rate is less than the rate of activity rate (Fig. 4 and Table 4).

Conclusion
As mentioned before, spectral acceleration zoning has not been done in Tehran before, so this research has a considerable importance in seismic design of structures. The main results of this study could be summarized as followed: • Based on Paleoseismology North Rey, South Rey and Kahrizak are considered as ancient lake shore not an active fault • Based on Vs 30 most dense, urban areas are situated on alluvial materials with shear wave velocity less than 360 m/s • Acceleration is increased in period of 0 to 0.2 s and in this period, it reaches to its maximum value and then shows downward trend • Comparing results, based on activity rate and slip rate, shows that for the period of 0 to 0.4 s, maximum and minimum spectrum acceleration, is higher for the data, based on slip rate comparing with the data of activity rate. For the period of 0.4 s spectrum acceleration is equal, for both data and after this period Spectral acceleration values of the slip rate is less than the rate of activity rate so combination of both method based on logic tree, is recommended to estimate spectrum acceleration for seismic designs