Document Type : Original Article
Extended Abstract
1. Introduction
Earthquake, as a natural phenomenon, poses a significant challenge that humanity has grappled with for centuries. The rapid increase in urban population in recent years has amplified the potential for damage, worsened by the fact that half of these cities are situated near earthquake belts. The location of 90 percent of Iranian cities within the hazard zone of fault lines (20 kilometers) makes them vulnerable to even a 5-Richter scale earthquake. Like other cities in the country, earthquake is the most significant natural disaster for Neyshabur, and its historical background over the past 200 years demonstrates its seismicity. There are three active compressional faults near the city of Neyshabur, which are of great importance. This is because compressional faults contain more energy than strike-slip and normal faults. Two faults are located in the northern part of the city, south of the Binalud Mountain, and another fault is located in the eastern area, less than 20 kilometers from the city. Based on a comparison of population census statistics from 1956 to 2016, Neyshabur has experienced a significant increase in urbanization. The migration of villagers to the city, followed by an increase in the urban population, has led to urban population density. Also, in recent years, high-rise construction has caused population density in urban neighborhoods. Given the conditions of Neyshabur in terms of human characteristics and the possibility of earthquakes, based on the prevailing geographical conditions in the region, this city may be vulnerable to earthquakes. Considering the lack of accurate knowledge of the earthquake return period and the lack of sufficient studies in this field, if urban sprawl is improperly planned, an earthquake could cause severe financial and human losses. It could also seriously damage urban infrastructure.
2. Materials and Methods
In the present study, 5 influential criteria in the intensity of earthquakes at the city level and the 30-year occurrence of earthquakes in Neyshabur city are used. Human factors influencing the intensity of earthquakes are population density, gross residential density, net urban density, average number of building stories, and the quality of buildings in Neyshabur city, each of which plays a role in increasing the city's vulnerability to earthquakes. In this research, Weka data mining software was used to implement various algorithms in assessing the city's vulnerability to earthquakes. Also, GIS software was used to output the final maps of the research. The models used in this research are from the Weka software models. The Bagging model with the RepTree algorithm, the ClassificationViaRegression model with the MP5 algorithm, the Random Committee model with the RandomTree algorithm, and the Trees model with the Random Forest algorithm are among the models and algorithms used in this research.
3. Results and Discussion
Research findings indicate that supervised tree-based machine learning algorithms possess the capability to delineate and monitor areas at risk of earthquakes with sufficient accuracy. The four aforementioned algorithms have demonstrated an assessment accuracy exceeding 85 percent. Furthermore, following the integration of human factors exacerbating earthquake damage within the urban area with the results obtained from the algorithms, the vulnerability assessment output of the urban area of Neyshabur was obtained. The final output indicates that Neyshabur is earthquake-prone and its urban area is relatively vulnerable to it.
4. Conclusions
In this research, after assessing the areas at risk in the city and integrating factors exacerbating earthquake vulnerability at the city level, and based on the findings, we concluded that, by integrating the predictions of four algorithms, approximately 42% of the area of the city of Neyshabur is located within areas with very high and high vulnerability. According to the input factors to the software, the findings show that the central areas of Neyshabur are more vulnerable than its other areas, and vulnerability decreases as you move from the city center towards the periphery. Also, the north of the city is relatively safe from earthquake vulnerability. The high population density in the city center, the small area of various urban land uses resulting in high urban fabric density, the small area of residential land use resulting in high gross residential density, the concentration of commercial and administrative land use in this part of the city resulting in a high density of both resident and non-resident population, and the presence of the city's core in this area are the main reasons for identifying the city center as the most vulnerable point in the current assessment. Therefore, comparing the predictions made in the research with the reality of the current situation in the city of Neyshabur, the overall conclusion is that the research findings are logical. Furthermore, the implemented algorithms were evaluated using accuracy assessment techniques, and the results show the high accuracy of the algorithms in classifying earthquake-prone areas. Therefore, in general, half of the city's area is located within a very vulnerable and high-risk zone for earthquakes.
5. Acknowledgment & Funding
We sincerely appreciate the Nishabur Municipality Organization for their assistance in advancing this research.
6. Conflict of Interest
The authors declare that they have no conflicts of interest.
