Bursa region earthquakes and their impacts on the settlements

The main cause of tectonic activity in Türkiye is due to the northward movement of the Arabian and African plates relative to the Anatolian Plate. As a result, the Anatolian Plate moves westward along two strike-slip fault systems called the North Anatolian Fault Zone and the East Anatolian Fault Zone. The former is a right-lateral strike-slip fault between the Eurasian and Anatolian plates and the latter is a left-lateral strike-slip fault between the Arabian and Anatolian plates. Being major fault systems and constituting the plate boundaries the North Anatolian Fault Zone and the East Anatolian Fault Zone are the tectonic elements that are the most important source of earthquake hazard in Türkiye. They produced frequent large and destructive earthquakes both in the histrorical (before 1900) and instrumental (after 1900) periods. The North Anatolian Fault Zone extends in three fault branches in the Marmara Region, called as the Northern, Middle and Southern branches. The Northern Branch passes Akyazı, Sapanca, Gölcük, Karamürsel, south of Princess Islands, offshore İstanbul, Silivri and Tekirdağ, Gelibolu Peninsula and reaches Saros Gulf in the Northern Aegean Sea. The Middle Branch passes Geyve, İznik, Gemlik, Bandırma, Karabiga and reaches to the Northern Aegean Sea. The Southern Branch passes Yenişehir, Bursa, Manyas, Gönen, Yenice, Edremit and reaches to the Northern Aegean Sea. It has been previously determined that especially the Northern Branch is much more active than the other branches, and that the Middle and Southern branches have similar activity rate. Excluding the earthquakes of the 26 November 1143 Bursa (MS = 6.4), the 12 May 1327 Karacabey (MS = 6.4) and the 11 April 1855 Bursa (MS = 6.3) earthquakes, 31 earthquakes which were considered to have magnitude MS ≥ 6.8 occurred in the Marmara Region after AD1000. The majority of these earthquakes occurred on the Northern Branch, confirming its major role in plate kinematics and the difference in the activity rates among the fault branches. In this study, the earthquake hazard and earthquake impacts on the settlements were investigated in the region (called as the Bursa Region here) that contains two important historical settlements such as Bursa and İznik and their immediate surroundings within the Marmara Region, where the earthquake risk is highest in Türkiye. The area selected as the Bursa Region is the region between 39.8o-40.6o Northern latitudes and 27.75o-30.25o Eastern longitudes. The Middle and Southern branches of the North Anatolian Fault Zone extend in the selected region, causing high seismic activity that include large and destructive earthquakes. The important earthquakes and their impacts, especially after AD1000, were compiled from the historical earthquake studies and the earthquake catalogues . In addition, the earthquake hazard was mathematically assessed from the statistical analysis of the seismic activity after 1970. It was determined that the earthquakes on the Northern Branch have relatively less impact on the Bursa Region compared to the earthquakes on the Middle and Southern fault branches. It was further determined that the earthquakes that had relatively large impacts on the Bursa Region after AD1000 were the September 1065 İznik, the 26 November 1143 Bursa, the 12 May 1327 Karacabey, the 15 March 1419 Bursa, the 10 May 1556 Southern Marmara Sea-Erdek and the 28 February-11 April 1855 Bursa earthquakes, which occurred on the Middle and Southern fault branches. The effects of the March 1855 Bursa earthquake were most documented one in the historical sources. Some of these effects were presented here to exemplify the impact of the earthquakes in the study area. Information was obtained related to the major earthquakes, which had produced consequences that affect the welfare level of the settlements in the region and even lead to enemy occupation due to the social and economic weakness they caused. Many examples of the damages and repairs to historical and cultural monuments in the settlements in the region, including the Bursa Grand Mosque and an aqueduct in Iznik, were presented. Evidence from the fault trenches along the faults for the listed earthquakes in the Bursa Region were obtained from the previous studies. Traces of the 1419 Bursa and 1857 Gemlik earthquakes were found in a trench excavated along the Middle Branch in the east of Gemlik. Traces of the 14 earthquakes in the last 2000 years were presented within the sediments under the Lake İznik, through the analysis of the taken sediments cores. A fault displacement of 0.5 meter belonging to the 1065 İznik earthquake was also documented. Traces of the fault ruptures of the 1327 Karacabey and 1964 Manyas earthquakes were obtained in a trench along the Southern Branch digged near Manyas. It was understood from this fault trench that there was an earthquake recurrence interval of 1964-1327 = 637 years between the last two earthquakes along the Southern Branch. In a trench excavated in the immediate west of Bursa along the Southern Branch, traces of the 1143 and 1855 Bursa earthquakes were documented. A time interval of 1855-1143 = 712 years between the last two earthquakes were obtained from this fault trench along the Southern Branch. Therefore, the fault trench studies suggested that the major earthquakes may have occurred along the Southern Branch at average time intervals of 650-700 years. Considering the similarity of fault activity between the Middle and Southern fault branches and that the last earthquakes that occurred along the Middle Branch were 1419 Bursa and 1556 Bandırma earthquakes the average large earthquake time interval is close to be completed. For earthquake statistics in the Bursa Region, a data consisting of MW ≥ 2.5 earthquakes covering the period between 1970 and 2020 was initially selected from the homogeneous Türkiye seismicity catalogue given in a previous study. This catalogue had been generated from the catalogues of The Kandilli Observatory and Earthquake Research Institute and International Seismology Center. Before statistical calculations, the seismicity catalogue, which include 4809 MW ≥ 2.5 earthquakes, was analysed in detail and it was seen that the catalogue became more homogeneous after 1977 and was complete for MW ≥ 2.7 earthquakes. It has been evaluated that there was probably an improvement in the earthquake monitoring and recording network after 1977. Therefore, for a more reliable statistical analysis, a new catalogue containing MW ≥ 2.8 earthquakes after 1977 was generated and a declustering process was applied. With the declustering process, foreshocks and aftershocks were mathematically eliminated as much as possible and only normal earthquake activity was strived to be obtained. Thus, the resulting earthquake catalogue, which includes a total of 1536 earthquakes, was used to obtain the Gutenberg-Richter Relationship, the basic and widely known relationship of earthquake statistics, for the study region. Gutenberg-Richter Relationship is the relation between the magnitudes and numbers of earthquakes. The Gutenberg-Richter Relationship is defined as Log N(M)=a-bM. Here, N is cumulative number of earthquakes with a magnitude exceeding a given magnitude M while a-value and b-value are constants. a-value is related to seismic activity level and b-value is inversely related to the crustal stresses. After determining a-value and b-value in Gutenberg-Richter Relationship recurrence times and the statistical probability of occurrence of the earthquakes of given magnitudes within a given future time interval can be determined. This analysis allows statistical determination of the earthquake hazard. From the calculated Gutenberg-Richter relationhip, the a-value (annual) and b-value were determined to be 4.81 and 1.17, respectively. Using the a-value (annual) and b-value in the Gutenberg-Richter relationship, earthquake recurrence intervals for the 3 targeted magnitudes (Mh =6.5, 6.8 and 7.0) in Bursa Region were calculated as 624, 1399 and 2398 years, respectively. Using the estimated a-value and b-value of Gutenberg-Richter relationship it was determined that the statistical probability of occurrence of the targeted earthquakes (e.g. Mh =6.5, 6.8 and 7.0) in the Bursa Region in the next 30 years to be 5%, 2% and 1%, respectively. The 650-700 years recurrence time determined from the fault trench studies is comparable to the earthquake recurrence times of 624, 1399 and 2398 years calculated from the earthquake statistics for the magnitudes MW =6.5, 6.8 and 7.0, respectively. Considering that 605 years have elapsed since the 1419 Bursa earthquake and 468 years since the 1556 South Marmara Sea-Erdek earthquake along the Middle Branch, it is suggested that the large earthquake recurrence interval is close to be fulfilled. The decline in the calculated b-value time variations after 2006 was interpreted as an indication of this increasing earthquake hazard, regarding its inverse relation with the crustal stresses. Taking into account that major earthquakes in the Bursa Region caused significant disasters in the history and that there is a serious earthquake hazard in the future, it has been suggested that the earthquake hazard should be continuously updated in the context of disaster management. Updated hazard-based earthquake scenarios should be prepared to have an idea about earthquake risk. The large number of historical and cultural monuments, population density and industrial development in the settlements in the region would cause high earthquake risk, which was also pointed out in the previous studies. Therfore it necessary to prioritize earthquake disaster mitigation activities in the region. All of these activities require better and integrated disaster mangement approaches to mitigate earthquake losses in the study area.