Seismic hazard analysis of the southeastern Pamir Frontal Thrust: implications for regional seismic risk patterns

Analyzing seismic hazard risk is crucial for comprehensive risk mitigation and seismic emergency planning. A scientific evaluation of seismic hazard risk is also crucial for strengthening pre-disaster preparedness and reducing disaster losses. The Pamir Frontal Thrust (PFT) fault is the most recent deformation zone that is still active in the late Holocene, with a documented history of earthquakes of magnitude 7.0 or higher. The southeastern section of the fault passes through various densely populated townships, and there are multiple vulnerability factors in the region, which make the southeastern section of the PFT fault at high risk of seismic hazards. This study assesses the current seismic risk of the southeastern section of the PFT fault. The potential for earthquakes in the region is evaluated by analyzing characteristics such as historical seismicity and the rate of fault activity. It analyzes the distribution of ground-shaking impacts in near-fault regions by combining stochastic simulations of high-frequency ground shaking with predictive methods for near-fault effects. Building characteristics are surveyed using an integrated space-air-ground approach. Through the integration of these methods, an in-depth assessment of the seismic hazard risk was conducted. The results show that regions exposed to more intense seismic shaking demonstrate correspondingly higher damage indices, and the densely populated townships in the area with intensity of VIII and above should be paid more attention to. Moreover, under identical intensity circumstances, areas with a high proportion of earth - wood - type houses display greater anticipated damage. And a gridded hazard risk assessment was produced. Seismic risk maps for the region are also provided. It provides the basis for effective disaster prevention and preparation.

1 Introduction

Natural hazards continually threaten human life, physical infrastructure, and the global economy. Earthquakes are particularly serious and deadly natural disasters worldwide (U UNISDR, 2009). China is heavily affected by various natural disasters. Xinjiang, in particular, experiences frequent and intense seismic activity, making it one of the provinces with the highest levels of seismic activity in the country. Notably, its seismic activity zones in Xinjiang largely coincide with economic development zones, placing both lives and regional development at significant risk. Consequently, the Chinese government has prioritized reducing seismic risk. This strategic focus reflects the need to protect people’s lives and property, thereby maintaining economic and social stability (Shi and Xu, 2016). A scientific and reasonable assessment of disaster risk is crucial because it equips stakeholders and decision-makers with tools to devise mitigation strategies and safety measures for future earthquakes. This allows for the identification of weak points and the implementation of effective defensive measures prior to a disaster, thereby eliminating potential hazards considerably. This process can enhance public awareness of seismic risks prior to its occurrence (Panagiota et al., 2012; Asghari et al., 2025).

To advance seismic hazard risk reduction and transition from loss mitigation to disaster risk reduction, many countries have conducted studies for the analysis of disaster risks and implementation of scientific disaster preparedness measures. Countries such as Iran (Karimzadeh et al., 2014), Egypt (Badawy et al., 2017), Colombia (Salgado-Gálvez et al., 2017), Rome (Pavel et al., 2016), Chile (Hussain et al., 2019), U.S.A (Kircher et al., 2006), and Europe (S. Lagomarsino and Giovinazzi, 2006; Mouroux and Brun, 2006; Mouroux and Brun, 2008; Crowley et al., 2009; Crowley et al., 2018; Crowley et al., 2019; Silva et al., 2015; Silva et al., 2016) have conducted seismic hazard risk assessment studies to varying degrees. These studies range from those considering near- and far-field seismic source impacts to simulating seismic hazard scenarios, employing diverse research methodologies. Currently, mainstream research approaches include the seismic hazard risk index (Davidson, 1997), seismic impact factor analysis (Cardona et al., 2005; Cardona et al., 2007; Duzgun et al., 2011; Frolova et al., 2011), probabilistic seismic risk assessment (Marulanda et al., 2013), deterministic seismic risk assessment (Giannaraki et al., 2018), integrated phantom model (Cardona et al., 2007; Mili et al., 2018), spatial multi-criteria analysis (Sinha et al., 2014), time-varying b-value consideration (Gulia et al., 2016), GIS-based approaches (Tadjer and Bensaibi, 2017), scenario-based methods (Pavel and Vacareanu, 2016), as well as neural network and analytic hierarchy process methods (Jena et al., 2020). These methods can be refined to capture the vulnerability characteristics of regional buildings, provided detailed building inventories are acquired. However, extensive research and past earthquake experiences have revealed that securing these crucial building inventory data remains challenging, requiring significant human resource and intensive onsite survey efforts. Accurate building inventories are essential bedrock for seismic risk assessment, providing far more comprehensive and robust foundations for disaster preparedness planning and post-earthquake emergency response operations. These critical factors must be considered carefully.

In this study, we focused on the southeastern section of the Pamir Frontal Thrust (PFT) fault, a region characterized by high seismic activity, a dense distribution of seismic tectonics (Sembroni et al., 2024), and a substantial population. This area was selected as the target area to quantify the regional seismic risk. The potential for earthquakes in the region was evaluated by analyzing characteristics such as historical seismicity and fault activity rates. Seismic hazard was assessed using a deterministic method, and the ground shaking impact field in the southeastern section of the PFT fault was mapped. The exposure and vulnerability status of hazard-bearing bodies were determined through a detailed survey of the space-air-ground integration of hazard-bearing bodies. Ultimately, these elements were combined using a seismic risk analysis model to characterize the geographic distribution of seismic risk in the southeastern section of the PFT fault.

This study fills the gap in applying traditional earthquake risk assessment models to complex tectonic zones. It addresses several limitations of conventional models, including data acquisition constraints, applicability issues, and neglect of near-fault effects and complex geological structures. These innovations not only improve the scientific rigor and precision of seismic risk assessment for the southeastern PFT fault but also establish a new paradigm for tectonically complex areas, offering significant reference value for disaster prevention and mitigation in similar global regions.

The remainder of this paper is organized as follows. Section 2 comprehensively discusses seismic tectonics and activity. Section 3 presents the proposed seismic hazard risk assessment methodology. Section 4 provides a survey of disaster-prone bodies and characterizes their vulnerability. Sections 5, 6 present the study’s findings and discussions. Finally, Section 7 concludes the paper.

2 Overview of seismic tectonics and activity

The PFT fault serves as the boundary fault between the western Kunlun piedmont depression zone and Tarim block (Figure 1a), ranking among the largest thrust faults within the western Kunlun piedmont. It persists as an active deformation zone during the late Holocene. Stretching westward from Kyrgyzstan into the Chinese territory, the fault forms an arc protruding towards the NE, spanning approximately 200 km within China. Its primary tectonic manifestation involves backthrusting of Eocene-Oligocene red mudstones and sandstones beneath the Pliocene-Lower Pleistocene conglomerates, displaying significant right-lateral strike-slip characteristics. Between Uluqchati and Jiangburak, the fracture strikes nearly EW near the apex of the arc. Here, it exhibits reverse thrusting, with a south-dipping section angled between 15° and 30°. Near Wupaer, the fracture trends NNW and is dominated by dextral (right-lateral) strike-slip motion. Geomorphological evidence reveals a striking height difference of 700 m across the fault. This rupture represents a shovel-shaped structure at the leading edge of the tectonic overturning of the West Kunlun Mountains and constitutes the seismic source for the 1985 M7.1 Wuqia and 1993 M6.2 Shufu earthquakes (Figure 1b). Critically, the southeastern segment traverses multiple densely populated townships, establishing it as a high-risk zone where active seismicity coincides with dense population distribution.

Figure 1

Figure 1. Topographic and tectonic map of the Pamir Frontal Thrust (PFT) fault. (a) Shaded relief map of Central Asia, showing the location of the PFT fault. (b) Active faults and large earthquakes atop the topography of the southeastern segment of the PFT. The red dots indicate the epicenter locations of magnitude 7 earthquakes, the yellow dots represent the epicenter locations of magnitude 6 earthquakes, the purple lines denote active faults, the red lines indicate PFT faults, and the blue area marks the location of Figures 3–7, which is also the study area of this research.

Microgeomorphic measurements revealed that the PFT fracture dislocated multiphase alluvial fan, creating a fault-steepened canyon of varying heights. In general, the steepness of the southeastern section of the PFT rupture is within the range of 0.2–12 m, with the most values falling within 0.2–4 m. he scarp height increases from north to south before decreasing again, reaching its maximum in the middle section. Compared with previous research, the vertical displacement (0.2–12 m) of the faulted landforms in this section is basically comparable to that of the northwest section of the PFT fracture, with no clear signs of horizontal displacement observed.

The western section of the Southern Tien Shan and Pamir confluence region, where the PFT fault is located, is one of the most seismically active regions in Xinjiang. In a study of paleoseismicity of the PFT fault, Feng examined the 1985 Wuqia earthquake and excavated 14 paleoseismic trenches in 1986, revealing the remains of paleoseismic activity of the rupture. The results of a lateral comparative study of the stratigraphy, lithology, structure, and stratigraphic sequence of each trench suggest that at least three paleoseismic events have occurred on the fault (Feng et al., 1987). The last two ancient seismic events occurred approximately 4,900 and 700 years ago. Since 1970, 204 earthquakes of M ≥ 4.7 have occurred within 20 km of the fault. Among these, six events reached magnitude 7 or higher, 19 were between 6.0 and 6.9, and 103 were between 5.0 and 5.9. Moreover, 7,820 earthquakes of magnitude 2.5–4.7 were recorded, including 485 events of M 4.0–4.6, 3,024 events of M 3.0–3.9, and 4,311 events of M 2.5–2.9. The largest earthquake to occur on the rupture was a magnitude 7.1 event in 1985. Microearthquakes are densely distributed, particularly in clusters, in the area around the rupture. This frequent seismic activity has led to severe earthquake disasters in the region, most notably the 1985 Wuqia earthquake (magnitude: 7.1), which caused 67 deaths, more than 200 injuries, the collapse or serious damage of over 10,000 houses, and economic losses totaling more than 100 million yuan. The 1993 Shushu earthquake (magnitude: 6.2) caused two deaths, eight serious injuries, and collapse or serious damage of over 11,000 houses, along with more than 3,700 suffering varying degrees of damage. Hence, an in-depth study of the seismic hazard risk in the southeastern section of the PFT is crucial for understanding the regional disaster risk, mitigating disaster losses, and ensuring the safety of people’s lives and property. The location of the PFT and distribution of the epicenters of strong earthquakes are shown in Figure 1.

3 Seismic hazard risk assessment methodology

The geometric and kinematic characteristics of the rupture and historical occurrence of earthquakes were combined to analyze the ability of the fault to generate earthquakes (Adib and Kianoush, 2025a). A deterministic method was used to set up the earthquake scenario, and a distribution map of the ground vibration impact field in the southeastern section of the PFT rupture zone was obtained by combining it with a ground vibration attenuation model. Through a detailed investigation, the structural type and vulnerability of the buildings were determined, and a vulnerability analysis of the buildings was completed. Finally, a seismic risk assessment of the southeastern section of the PFT fault zone was conducted based on the seismic risk analysis model.

3.1 Setting the seismic method

Earthquake hazard analysis is generally classified into two types: deterministic seismic hazard analysis (DSHA) and probabilistic seismic hazard analysis (PSHA). DSHA mainly determines earthquakes based on seismic tectonics and historical seismic data in the study area and determines both the earthquake magnitude and the location of the epicenter. This method emphasizes the physical rationality of the strong ground motion characteristics of the fault near field and has proven valuable for engineering applications, particularly in the seismic design of cross-fault bridges and lifeline infrastructure. PSHA quantifies the impact of all potential earthquakes in a specific area over a future period (Li, 2001), providing a comprehensive representation of seismic ground motion effects. This method captures the spatiotemporal heterogeneity of seismic activity, offering probabilistic hazard curves that align with various design standards for critical engineering sites.

Building on the fundamental principles of seismic science, DSHA is used to correct the results of the PSHA method in the compilation of a new generation of seismic zoning maps in the United States, underscoring its irreplaceability in engineering practice (Li et al., 2022). DSHA and PSHA have different precision and processing strategies for characterizing the physical processes of fault rupture. Typical case studies show that DSHA offers distinct advantages for seismic risk assessment in specific tectonic settings and provides a more physically realistic representation of near-fault strong-motion characteristics than PSHA (Mualchin, 2011). For example, observations from the Chi-Chi earthquake indicate that the acceleration spectrum value of the hanging wall was 1.5 times higher than that of the footwall (Sheng et al., 2023). PSHA, on the other hand, focuses on the uncertainty of statistical fault parameters. Although the developed three-dimensional fault source random sampling algorithm can describe the characteristics of earthquake recurrence through Poisson distribution, it faces difficulty in accurately simulating the near-fault directional effect (Chen et al., 2023). A case study on a double-tower cable-stayed bridge showed that while PSHA performs well in predicting the ground motion attenuation relationship beyond 300 km of the fault, it underestimates the peak acceleration by about 20% within 10 km of the fault (Xu and Chen, 2025). In terms of parameter sensitivity, DSHA is more sensitive to the fault locking coefficient (e.g., the locking value of Elashan fault, 0.85), whereas PSHA is more dependent on the b-value parameter in the magnitude-frequency relationship (Jian et al., 2020). Given that this study analyzes the seismic risk of the Southeastern Pamir Frontal Thrust, the deterministic method is used.

In addition to the magnitude of the earthquake set in the study area, the location of the set earthquake was determined based on the tectonic sections where destructive earthquakes are most likely to occur and the latest activity, size, nature, and relationship with the seismicity of the southeastern section of the PFT rupture. The set earthquakes are as follows: point #1 is located in the tectonic transition part of the rupture, which is a part of the fault with strong activity; point #2 is located in the possible seismic tectonics in the northern part of the rupture, where moderate and strong earthquakes are frequent in the vicinity of the rupture; and point #3 is located in the southern part of the rupture at the basin–mountain junction, which is the convergence area of the two ruptures (Figure 2).

Figure 2

Figure 2. Geological map of the PFT, with blue dots indicating the locations of the set earthquakes.

Multidisciplinary approaches such as geostatistics and intelligent models are widely applied in seismic studies (Kianoush et al., 2023b). Based on the relationship between earthquake surface rupture length L and magnitude M, seismologists have obtained empirical formulas for different regions (Equations 1–3).

$\text{Global\hspace{0.17em}scope}:\mathrm{M}=6.24+0.619\text{\hspace{0.17em}lgl\hspace{0.17em}Error\hspace{0.17em}}0.29(1)$

$\text{United\hspace{0.17em}States\hspace{0.17em}and\hspace{0.17em}China}: \mathrm{M}=4.94+1.296\text{\hspace{0.17em}lgl\hspace{0.17em}Error\hspace{0.17em}}0.19(2)$

$\text{Western\hspace{0.17em}China}:\mathrm{M}=6.117+0.579\text{\hspace{0.17em}lgl\hspace{0.17em}Error\hspace{0.17em}}0.21(3)$

The length of the surface rupture zone of the PFT rupture is approximately 100 km. The average of the three methods is used to estimate the maximum seismicity of the PFT rupture. The weighted result is 7.42 ± 0.23; thus, the future maximum seismicity of the PFT rupture is obtained as M = 7.42 ± 0.23.

3.2 Ground shaking forecasts

With the gradual increase in the number of near-fault ground shaking records acquired, it has become evident that near-fault ground shaking differs significantly from mid- and far-field shaking. For example, the up-disk effect, specifically the up-disk effect of the reverse-strike fault, is particularly pronounced and is characterized by the ground vibration located in the up-disk of the fault being larger than that of the down-disk, and the distribution area of the strong ground vibration is larger. However, the attenuation of the ground vibration in the up-disk of the fault is slower than that in the down-disk. In addition, slip-strike effects cause severe seismic damage in near-fault regions, primarily through surface rupture and permanent deformation of the ground. Near-fault velocity pulses, characterized by pulse-like waveforms, long pulse periods, and abundant medium and long periods, contain high energy and cause large displacements in structures. Velocity pulses caused by rupture directionality occur mainly in the direction perpendicular to the fault plane, whereas those caused by permanent ground displacements occur in the component that coincides with the sliding direction of the fault. The directional effect of near-fault rupture strengthens the long-period component of ground shaking in the forward-rupture direction, producing large peaks with short holding times. Consequently, velocity time histories exhibit significant long-period pulses in the component perpendicular to the fault plane, and the ground shaking in this direction is considerably larger than the ground shaking in the direction parallel to the fault plane. A three-dimensional fault model reconstructed from geophysical data helps clarify the geometric structure of the fault (Kianoush et al., 2023). In this study, both a high-frequency ground shaking stochastic simulation method and ground shaking prediction formula method considering the near-fault effect were used to evaluate the distribution of the ground shaking impact field near the rupture fault.

3.2.1 Stochastic simulation methods for ground shaking

It is generally accepted that earthquakes occur because of faulty activity. The EXSIM program was used to simulate the high-frequency ground shaking. In 1955, Housner first proposed a stochastic model of earthquake sources. He hypothesized that result from the superposition of multiple incremental shear dislocations, and that the ground-shaking time course is formed by the superposition of numerous random acceleration pulses. Based on the assumptions of the disk rupture model of high-frequency ground shaking sources (Brune, 1970), Hanks and McQuire (Hanks, 1978; Hanks and Mcguire, 1981) developed equations for the root-mean-square value of acceleration and peak acceleration as functions of magnitude and distance, based on high-frequency direct S-wave recordings of periods smaller than the fault rupture time. In 1983, David extended the work of Hanks and Mcquire by expressing the displacement spectrum of simulated ground shaking as a simple functional relationship between the source, path, and site influences (Equation 4). This method replaces the magnitude of the empirical method with the source spectrum and modifies the shape of the spectrum with attenuation factors (Boore and Joyner, 1997) to replace the role of distance in the empirical method, while also incorporating the near-surface amplification and filtering effect of the site soil layer.

$Y\left(M_0,R,f\right)=E\left(M_0,f\right)P\left(R,f\right)G\left(f\right)(4)$

Here, $Y$ is the amplitude of ground motion, $M_0$ is the seismic moment, $R$ is the fault distance, $f$ is the frequency, $E\left(M_0,f\right)$ is the seismic source term (Equations 5, 6), $P\left(R,f\right)$ is the path attenuation term (Equation 7), and $G\left(f\right)$ is the effect of the local site (Equation 8). Obviously, the construction of models for each influential factor, such as the source, path, and site, that accurately reflect the characteristics of the study region are important for reflecting the strong regionality in the high-frequency stochastic simulation approach.

$E\left(M_0,f\right)=\frac{R_{\theta \phi }FV}{4\pi R_0{\rho }_s{\beta }_s^3}\times \frac{M_0}{{\left[1+{\left(f/f_0\right)}^a\right]}^b}(5)$

Here, $R_{\theta \phi }$ reflects the source radiation mode and the station azimuth effect, which is generally taken as 0.6; $F$ represents the amplification effect of the free surface, generally take 2.0. V is the horizontal component coefficient of seismic energy, taken $1/\sqrt{2}$ ; $R_0$ is the reference distance when ${\rho }_{\mathrm{s}}$ and ${\beta }_{\mathrm{s}}$ are selected, and it is usually 1 km. ${\rho }_{\mathrm{s}}$ and ${\beta }_{\mathrm{s}}$ are the medium density and shear wave velocity near the source, respectively. $f_0$ is the corner frequency; coefficient $a$ = 3.05–0.33 $M_W$; $b$ = 2.0/ $a$. The relationship between $f_0$ and stress drop $\Delta \sigma$ is as follows:

$f_0=4.9\times {10}^6{\beta }_s(\Delta \sigma /M_0{)}^{1/3}(6)$

Here, $P\left(R,f\right)$ is a path attenuation term, which consists of two parts: geometric attenuation and inelastic attenuation.

$P\left(R,f\right)=Z\left(R\right)\exp \left(-\frac{\pi fR}{Q\left(f\right){\beta }_s}\right)(7)$

Here, $Q\left(f\right)$ is the quality factor.

$G\left(f\right)$ Among them, the quality factor.

$G\left(f\right)$ is the influence of local site, which can be expressed as a function of amplification spectrum $A\left(f\right)$ and attenuation $D\left(f\right)$.

$G\left(f\right)=A\left(f\right)D\left(f\right)(8)$

According to a study of the fault, the seismic capacity of the fault section can reach 7.5; that is, the moment magnitude used in this simulation is 7.5, and the conversion relationship between the moment magnitude and the seismic moment (Equation 9) is

$M=\frac{2}{3}\log M_0-10.7(9)$

The specific model parameters are shown in Table 1.

Table 1

Table 1. Random simulation parameter table of seismic finite fault is set.

3.2.2 Formulaic method of ground shaking prediction

This empirical method relies directly on observational data. With a sufficiently rich dataset of strong vibrations, regression techniques are used to fit the parameters describing ground vibration intensity. Recent developments in ground vibration attenuation relationships are obtained by using parameters such as moment magnitude $M_W$, fault distance $R_{RUP}$ or fault projection distance $R_{JB}$, and converted shear wave velocity $V_{S30}$ within 30 m of the surface. Commonly used attenuation relationships include Boore attenuation relationship (Boore, 2009), Chiou and Youngs attenuation relationship, and IDRISS attenuation relationship. Based on calculations and comparisons of these three attenuation relationships, the IDRSS attenuation model was found to be in line with the seismic hazard analysis in Xinjiang; therefore, the IDRSS attenuation relationship was selected as the ground shaking attenuation relationship for this study.

The general form of the IDRISS attenuation relationship (Equation 10) obtained using strong earthquake records recorded under “quasi-linear site” conditions in the NGA-West2 database is as follows:

$\begin{array}{ll}\hfill & \hfill \ln \left(PSA\right)={\alpha }_1+{\alpha }_{2}M+{\alpha }_3{\left(8.5-M\right)}^2-\left({\beta }_1+{\beta }_{2}M\right)\ln \left(R_{rup}+10\right)\\ \hfill & + \xi \ln \left(V_{S30}\right)+\gamma R_{rup}+\phi F\hfill \end{array}(10)$

Here, PSA in g denotes the response spectral acceleration at a 5% damping ratio; $\mathbf{M}$ denotes the moment magnitude; $R_{rup}$ denotes the distance from the surface point to the rupture surface (in kilometers); $V_{S30}$ denotes the average shear wave velocity (in m/s) at a depth of 30 m from the surface; and $F$ denotes the focal mechanism, with F = 0 for strike-slip faulting and F = 1 for reverse faulting. The parameter ${\alpha }_1$, ${\alpha }_2$, ${\alpha }_3$, ${\beta }_1$, ${\beta }_2$, $R_{rup}$, $V_{S30}$, $\xi$, $\gamma$,and $\phi$ were determined from the regression results. The specific model parameters are shown in Table 2.

Table 2

Table 2. Coefficients at each cycle of the bedrock acceleration response spectrum decay relationship (site conditions: Vs30 ≥ 450 m/s and M ≥ 6.75).

3.3 Seismic hazard risk assessment of buildings

To quantitatively evaluate the seismic risk of buildings, this study establishes a seismic risk assessment index system from three dimensions: focus, adaptability, and operability. Seismic risk is determined based on four key components: seismic hazard, exposure, building vulnerability, and seismic disaster reduction ability. Among them, seismic hazard refers to the predicted seismic intensity of the grid unit; exposure represents the total building area within the unit; vulnerability corresponds to the vulnerability matrix parameters associated with different structural types; and seismic disaster reduction ability reflects the regional seismic fortification level and emergency response capability. The earthquake disaster risk indicators of buildings (Equation 11) are as follows:

$SR=\frac{KI}{PI}(11)$

Here, $SR$ is the building earthquake disaster risk coefficient; $KI$ is the index of earthquake damage degree of buildings (Equation 12); $PI$ is the disaster reduction capacity index (Equation 13). When $SR$ > 0.5, the area is classified as high risk; when 0.3 < $SR$ ≤ 0.5, it is classified as medium risk; and when 0 ≤ $SR$ ≤ 0.3, it is classified as low risk.

To quantitatively asses the seismic disaster risk of buildings, as indicator system was established based on three aspects: focus, adaptability, and operability. By analyzing and summarizing the vulnerability characteristics of various types of buildings affected by destructive earthquakes in Xinjiang, a vulnerability matrix of the main structural types of buildings in the target area was established. Based on the total building area obtained through remote sensing interpretation and the proportion of structural types obtained by the sampling survey, the degree of damage of buildings was evaluated, and the comprehensive damage degree index was calculated using a 1-km grid as basic assessment unit.

Building an earthquake disaster vulnerability matrix is the basis of earthquake damage prediction. It can be used both to assess the damage caused by a certain earthquake and to provide a scientific basis for earthquake damage prediction and prevention systems in certain regions. The vulnerability matrix comprehensively reflects the damage ratio of the main structure types in a certain area under different earthquake intensities and exhibits distinct regional characteristics. In this study, the structural characteristics of rural houses in the research area were collected and studied, vulnerability matrix information of houses from representative past earthquakes was collected, and the vulnerability matrixes for various structural types were determined by integrating historical earthquake damage results with previous research results.

3.4 Predictive analysis of building earthquake damage levels

Building area data were obtained via satellite remote sensing interpretation, supplemented by corrections derived from unmanned aerial vehicle (UAV)-based remote sensing interpretation at selected sample points. The proportions of buildings of different structural types were determined from the statistical analysis of survey data from the nearest sampled sites. Based on the calculated predicted seismic intensity for each 1-km grid, the seismic damage level of buildings within each grid was estimated based on the susceptibility matrix corresponding to each structural building type. To assess the extent of building damage within each grid, the building seismic damage index KI is defined as follows:

$KI={\displaystyle \sum _{i=1}^3}\left(S_{i1}+S_{i2}+S_{i3}/2\right)/S_{\mathrm{总}}(12)$

where $S_{i1}$, $S_{i2}$, and $S_{i3}$ are the areas of destroyed, severely damaged, and moderately damaged buildings, respectively; i = 1,2,3 corresponds to adobe, brick, and brick-concrete buildings, respectively; and $S_{\mathrm{总}}$ refers to the total area of the above three structures.

$PI=0.3\times C_1+0.3\times C_2+0.3\times C_3+0.1\times C_4(13)$

Here, $C_1$, $C_2$, $C_3$, and $C_4$ represent the engineering disaster prevention ability, disaster relief preparation ability, emergency response ability, and social support ability of each unit, respectively. All four indicators are normalized to values between 0 and 1 according to the scoring rules $C_1$ reflects the seismic fortification level of building construction and lifeline engineering within the unit and is defined as the ratio of basic intensity to estimated intensity of each unit. $C_2$ reflects the emergency rescue reserve and preparation level of each unit. $C_3$ reflects the disposal capacity of each administrative unit in the event of an earthquake. $C_4$ reflects the cognitive level, legal framework, and routine earthquake-related training and drills undertaken by all social sectors within each administrative unit.

4 Survey of disaster-bearing bodies and characterization of their vulnerability

To more comprehensively capture the area, structural type, and distribution characteristics of disaster-bearing bodies, specifically houses, a detailed survey of houses was conducted in the target area. Using the original dataset for the southeastern section of the rupture, remote sensing image interpretation, UAV image interpretation, and field sampling surveys were carried out. Combined with earthquake damage survey results from previous destructive earthquakes in Xinjiang, we analyzed the overall vulnerability characteristics of disaster-bearing bodies in the region.

4.1 Remote sensing image interpretation

Machine learning methods are widely used for specialized information extraction and model parameter estimation (Kianoush et al., 2023c). High-resolution remote sensing data were obtained under cloud-free and fog-free conditions, ensuring good image quality. After ortho-correction and geometric correction, an object-oriented machine learning method was used to extract building information from the remote sensing images. Object-oriented classification begins with multi-scale segmentation, in which pixels are merged bottom-up into larger objects as long as their heterogeneity remains below a set threshold, resulting in consistent and homogeneous image units. Next, feature values based on spectral, size, texture, shape, context, and other information are then selected and calculated for each cell. Finally, each object is categorized to identify the spatial distribution of the building.

Accuracy evaluation (Yu et al., 2012) is to analyze the accuracy of building information extraction results and determine the advantages and disadvantages of the extraction results and if they are available. Manually selected building regions of interest were used as ground samples, and the accuracy of building information extraction results was evaluated. The total accuracy of building information extraction reached 83.42%, with a Kappa coefficient of 0.77, indicating a high level of accuracy, and only a small amount of missing extraction in the study area. However, some mis-extraction occurred, primarily in vegetation areas where vegetation textures are similar to building textures. In addition, due to pixel heterogeneity, the boundaries of the houses appeared irregularly distributed, and a small amount of mis-extraction occurred around the surrounding area after regularization. Overall, the algorithm demonstrated high accuracy, and the mis-extracted areas were corrected manually. Post-processing evaluation of remote sensing interpretation results was performed through manual visual interpretation of Google images, comparison with extracted features from UAV images, and field sampling surveys to further refine the remote sensing extraction results, ensuring the accuracy of data extraction. The remote sensing interpretation results are presented in Figure 3e.

Figure 3

Figure 3. Spatial distribution map of buildings obtained from remote sensing image interpretation, unmanned aerial vehicles (UAVs), and field surveys. (a–d) Show the UAV image processing and interpretation results of the (a) digital surface model (DSM)-, (b) digital elevation model (DEM)-, (c) normalized DSM (nDSM)-, (d) UAV-extracted house building distribution. (e) Shows the spatial distribution of buildings, where the blue grid signifies areas covered by unmanned aerial photography, and the yellow zones indicate the spatial distribution of buildings interpreted through remote sensing. (f–h) Shows pictures of typical house structures for the field survey: (f) earth and wood-framed houses; (g) brick and wood-framed houses; (h) brick-concrete houses. (i–K) Shows the histogram of housing vulnerability matrix in Xinjiang obtained from earthquake field investigation results: (i) earth and wood-framed houses; (j) brick and wood-framed houses; (k) brick-concrete houses.

4.2 UAV aerial survey and disaster bearing body identification

To compare the results of remote sensing image interpretation, high-resolution aerial images were acquired using a DJI Inspire 1 micro UAV equipped with a Zenmuse X3 gimbal. These images were used to extract high-precision building information. Several sampling areas, each covering approximately 4 km² within the study area, were selected for aerial surveying. Flight routes were established along the main streets and housing areas, with a minimum of two flights per route. The acquired UAV images were finally processed through relative orientation, absolute orientation, area network leveling, and mosaicing to produce digital orthophoto models (DOMs), digital elevation models (DEMs), and digital surface models (DSMs). Based on the UAV images obtained via aerial surveying, the DEM image was subtracted from the DSM image, and the UAV image building information was automatically extracted from the resulting normalized DSM (nDSM) (Figures 3a–d).

4.3 Vulnerability analysis of housing construction

For comparison with the data obtained through remote sensing and UAV image interpretations, a field sampling survey of house structure and area information was conducted within the UAV aerial survey area. The combined results from image interpretation, UAV photography, and field survey indicate that the types of housing structures in the study area mainly include earth-wood, brick-wood, brick-concrete, and reinforced concrete frame.

Earth-wood structures were widely distributed in rural areas of Xinjiang before 2003. Following the launch of the earthquake-resistant housing project in 2004, these structures have been gradually phased out; however, a certain number still remain in rural areas. These houses have low story heights, small openings, shallow depths, and adobe masonry walls. Because of the extremely poor mechanical properties of the materials, earth–wood houses have suffered the most severe damage in destructive earthquakes in Xinjiang, accounting for more than 90% of all severely damaged and collapsed houses, with noticeable damage has been even in areas experiencing intensity VI shaking. Currently, the proportion of earth–wood houses is less than 20%. With the continued implementation of rich people’s housing projects and the demolition of the old and the construction of new houses, this proportion will be further reduced, and very few of them will be used for living (Figures 3f,i).

4.3.1 Characteristics of brick- and wood-framed houses and earthquake damage

This type of houses can be divided into three categories according to the construction age and presence of anti-seismic measures. The first category consists of older, self-built brick-wood houses, which generally have poorer overall anti-seismic performance and exhibit high damage ratios during destructive earthquakes. The second category includes newer brick-wood houses constructed by residents in recent years that incorporate basic anti-seismic measures. The third category comprises houses built under rural housing projects. These structures are gradually increasing in number and spatial distribution and are constructed following formal design and construction standards. They are equipped with ring beams and structural columns, and feature herringbone ridged roofs with wood-framed beams, and use lightweight steel roofing panels that provide both waterproofing and aesthetic advantages. This type of housing is highly resistant to earthquakes, and houses built since 2004 have performed well during earthquakes events (Figures 3g,j).

4.3.2 Characteristics of brick- and concrete-framed houses and earthquake damage

Brick and concrete-framed houses are mainly concentrated in towns and cities. Self-built residential houses of this type often lack anti-seismic measures, have poor construction quality, and exhibit low seismic performance. In contrast, public housing is generally earthquake-resistant; although different versions of the earthquake-resistant design code are used, their seismic performance is relatively good (Figures 3h,k).

4.3.3 Characteristics of reinforced concrete frame structure houses and earthquake damage

Reinforced concrete frame buildings are mainly located in the urban areas of cities and counties. Some of these are assembled monolithic frame structures, in which the beams and columns are cast-in-place using reinforced concrete frame structures, while floor panels are prefabricated. This type of structure is used primarily in schools, hospitals, and offices. In general, the seismic performance is good and no serious damage or destruction has been reported during any of the historical destructive earthquakes.

5 Results

5.1 Ground vibration prediction results

5.1.1 Stochastic simulation prediction results:

The peak acceleration contour maps generated from the three earthquake settings indicate that extreme value zones exhibit significant zoning characteristics and are concentrated in the setting of earthquakes and their surrounding areas. The peak acceleration on the upper plate of the rupture (SW side of the fault) was higher than that on the lower plate (NE side of the rupture). A horizontal comparison of the results of the peak acceleration calculations and zoning of the three setting earthquakes shows that the high value area of setting earthquake 1 is the largest, with an area of 748.2 km² experiencing peak accelerations greater than 1,000 gal, accounting for 50.1% of the total area of the study area (Figure 4a). Conversely, for setting earthquake 2, the area with peak acceleration exceeding 1,000 gal was 715.5 km², accounting for 47.9% of the total study area (Figure 4b), whereas setting earthquake 3 has an area of 708.7 km², with peak acceleration exceeding than 1,000 gal, accounting for 47.4% of the study area (Figure 4c).

Currently, seismic intensity characterizes the intensity of seismic impacts for seismic hazard prediction; therefore, the results of the above acceleration calculations are converted into seismic intensity. The computational relationship between the single-component peak acceleration and intensity is

$I_{PGA}=3.23{\log }_{10}\left(PGA\right)+6.82(14)$

The contour distribution is plotted using the converted seismic intensity values. From the distribution of the seismic intensity contours, the peak acceleration calculated from the setup earthquake corresponds to seismic intensities above 8°, reaching a maximum of 12°. The dominant distribution is in the range of 9°–10°, and the areas above 10°, specifically 11°, are mainly distributed along the active rupture and clustered at the location of the setup earthquake (Figure 4d).

Figure 4

Figure 4. Contour map of distribution of ground shaking impact field for set earthquakes ((a). Set earthquake 1; (b) Set earthquake 2; (c) Set earthquake 3 (d). Composite intensity contour map).

5.1.2 IDRSS attenuation relation ground shaking prediction results

The calculated ground shaking impact field of the setting earthquake shows considerable elliptical distribution characteristics centered on the epicenter of the setting earthquake, and the long-axis direction is extended along the active tectonic distribution direction. The peak acceleration is in the range of 100–2,200 gal, and it decays rapidly with increasing fault distance.

To predict the distribution of the ground shaking impact field along the PFT rupture and its neighboring areas under the occurrence of a 7.5-magnitude earthquake, the peak ground shaking acceleration at the grid points was calculated using the high-frequency stochastic simulation and ground shaking prediction equation methods. The peak acceleration of ground shaking calculated using the former method is generally higher than that of the attenuation relation, but it declines more rapidly. The later method fully applies available near-fault ground vibration records. To maximize the advantages of the two methods, the maximum value of the results of the two computational methods in space was selected as the prediction of the ground vibration at the point and plotted on a contour map, as shown in Figure 5. The figure shows that the predicted intensities of the target areas of the PFT rupture all exceed 9°, and approximately 75% of the areas have predicted intensities greater than or equal to 10°; of this, approximately 0.9% of the very high intensity zones have predicted intensities of 12°, approximately 16.8% of the zones have predicted intensities of 11°, and approximately 57.0% of the zones have predicted intensities of 10°. The predicted results represent the combined results of the two methods for the three setting earthquakes and can be considered the maximum predicted value of seismic intensity in the target study area for the three setting earthquakes (Figure 6).

Figure 5

Figure 5. Acceleration contour pictures calculated by IDRSS attenuation relation ((a). Setting earthquake 1; (b) Setting earthquake 2; (c) Setting earthquake 3 (d). Composite acceleration contour pictures).

Figure 6

Figure 6. Composite contour and grid map of seismic intensity predictions for the study area. (a) Composite contour map of seismic intensity prediction. (b) Grid map of seismic intensity prediction, where red lines indicate PFT faults and blue dots represent the epicenter locations of setting earthquakes.

5.2 Results of the seismic hazard risk assessment

In the study area, the seismicity index of houses is higher in the northern and eastern parts of Oyitag Township in Aktau County, parts of Baren Township, northern part of Tashmirik Township in Shushu County, central part of Mogale Forest Farm, and northern part of Aigusi Township in Injisha County (Figure 7). Grids with high-intensity impacts have correspondingly higher housing damage indices. Under the same intensity, the housing damage index is higher in subzones with a high proportion of earth-wood houses. Most of the study area was classified as general disaster area; Tashmilik Township was classified as moderate seismic disaster area; and Wupaer and Aigusi Townships were classified as severe seismic disaster areas.

Figure 7

Figure 7. Distribution of housing damage levels.

The risk distribution shown in Figure 7 clearly reveals the coupling relationship among the three factors. The intensity of ground motion is the fundamental driver of risk hotspot formation, and the high intensity area distributed along the PFT fault zone directly determines the potential upper limit of building damage. Building vulnerability is a key variable of risk amplification. Risk hotspots are the spatial embodiment of the superposition of the two. The risk distribution map therefore not only reflects the spatial superposition of ground motion and building vulnerability but also provides a scientific basis for the precise delivery of regional disaster reduction resources.

The towns in the study area belong to Shufu County, Akto County, Yengisar County, and Wuqia County. Overall, the towns with the highest population concentration are primarily located in Shufu County and Akto County. The disaster reduction capacity of each county was investigated, and the results were scored to obtain the county disaster reduction capacity index. Shufu County places strong emphasis on disaster-relief preparedness. It maintains dedicated disaster-relief materials, has established earthquake-response teams with adequate personnel and equipment, and provides a township-level emergency shelter centrally located in the township. The shelter covers an area of 3,000 m² and can accommodate up to 1,500 people. The buildings and infrastructure of new, modified, and expanded buildings are designed and constructed in accordance with seismic fortification requirements. Shufu County is about 10 km away from Kashgar City, offering convenient transportation and good disaster reduction capacity. In Akto County, most new residential buildings in towns and villages are uniformly designed and constructed. The county has an independent earthquake authority and possesses the ability to quickly obtain on-site earthquake and disaster information. Multiple earthquake emergency drills are organized every year, and a comprehensive disaster rescue team equipped with complete personnel and equipment has been established. In 2016, a 6.7-magnitude earthquake occurred in Akto County, providing a real-world test of its emergency preparedness and response capabilities. The event demonstrated that Aketao County has a strong disaster-reduction capacity. The calculated disaster reduction capacity indexes were 0.57, 0.64, 0.73, and 0.87 for Yengisar, Shufu, Akto, and Wuqia, respectively. Based on these values, the calculated risk coefficient of earthquake disaster foe the study area exceeds 0.5, classifying it as a high-risk region. Therefore, the southeastern PFT zone stands out as a region with dense housing and population distribution but relatively low resistance to earthquake damage. This zone should therefore be prioritized for strengthening earthquake prevention and disaster reduction capabilities in the future.

6 Discussion

Deep-seated structures influence shallow deformation and seismicity (Ghalandari et al., 2023; Tavakoli and Barfizadeh, 2024). This study examines the intensity of seismic activity and the spatial distribution of earthquake occurrences to assess the seismogenic potential of major faults. It forecasts the spatial distribution of potential ground motion intensity from earthquakes, investigates the extent and distribution of building damage, and evaluates the possible damage levels that various types of building structures might endure during earthquakes within a specified future period.

The PFT fault extends from northwest to southeast, encompassing 22 townships and farms within a 20-km near-field range across Wuqia, Aketao, Shufu, and Yingjisha counties, with a total population of approximately 380,000. This study utilizes an integrated space-air-ground detailed survey of housing construction to ascertain their exposure and vulnerability conditions. Remote sensing interpretation is used to rapidly identify the spatial distribution of housing, completing the exposure analysis. The structural types of buildings were determined using UAV and field surveys, thereby completing the vulnerability analysis. Because UAV and field surveys are conducted using sampling methods rather than building-by-building inspections, building data within grids are inferred and may contain certain errors. Nevertheless, the data obtained through the integrated space-air-ground detailed survey method for housing construction saves time and effort compared to relying solely on field surveys; additionally, it enables the rapid acquisition of housing data over large areas, demonstrating strong potential for wider application.

Based on remote sensing image interpretation, UAV photography, and field survey results, the study area incorporates structures with varying degrees of vulnerability, including large-scale earth-wood, brick-wood, brick-concrete, and reinforced concrete houses. Recent historical seismic damage investigations in Xinjiang indicate that reinforced concrete houses experience relatively minor damage during earthquakes, whereas other types of structures experienced varying degrees of damage. The seismic damage matrix for buildings in this study is derived from historical seismic damage surveys in Xinjiang; however, additional data are required to supplement the information on the damage to reinforced concrete houses in high-intensity epicentral areas. With the implementation of the Prosperous Housing Project, the proportion of earth-wood houses in Xinjiang has significantly declined. The remaining earth-wood houses are now primarily used as utility buildings and are no longer inhabited, resulting in a low probability of human casualties owing to structural damage. Consequently, such houses may receive less attention in terms of damage assessment in future studies. Additionally, finite element analysis for stability assessment under complex coupled processes provides context for the simpler vulnerability indices used for buildings and could serve as a starting point for future work (Pirhadi et al., 2023).

By simulating earthquake scenarios, it is possible to rapidly assess the seismic intensity and analyze the levels of regional building damage caused by seismic events. This study employs an equidistant scenario setting method by establishing three earthquake epicenters. However, earthquakes of varying magnitudes can occur at any point along the entire fault. This study analyzed regional building damage using the maximum possible magnitude that could occur along the fault, elucidating potential worst-case damage scenarios to provide a basis for local disaster prevention and mitigation planning.

Based on the comprehensive seismic intensity prediction results, special attention should be paid to disaster risks, such as building damage in towns and villages within zones of intensity VIII and above. Therefore, effective response measures should be promptly implemented. Seismic calculations and a comprehensive analysis of seismic intensity and earthquake disasters indicate that the northern and eastern parts of Aoyitake Town in Akto County, certain areas of Baren Town, the northern part of Tashimilike Town in Shufu County, and the central and northern parts of Aigusi Town in Yingjisha County have suffered relatively severe damage. Priority should be given to densely populated villages and towns, such as Wupaer, Tierimu, Tashimilike, and Aigusi, which serve as key areas for disaster relief measures. In urban areas, various population-dense locations, including schools, hospitals, large public venues, and education transformation centers, should prioritize evacuation and relocation arrangements. Older urban districts, urban villages, and non-standardized, poorly fortified buildings in urban-rural transitional zones remain vulnerable. Although the earthquake resistance and disaster prevention capabilities of rural residences have been significantly improved through the Safe Housing Project, older self-built houses in urban villages, urban-rural transitional areas, and surrounding towns still present considerable seismic safety hazards. Considering the identified high-risk townships, the following targeted measures are proposed:

• Strengthen building seismic reinforcement, prioritizing the implementation of seismic reconstruction of rural self-built houses, and carrying out special safety assessment and reinforcement projects for public buildings such as schools and hospitals

• Establish earthquake emergency shelters in high-risk areas, equipped with emergency material reserve points.

• Conduct regularly organize community-level earthquake emergency drills to enhance residents’ self-rescue and mutual rescue capabilities.

• Strictly limit new high-risk industrial projects in high-intensity areas.

These measures can effectively reduce regional seismic risk and enhance the overall disaster reduction capacity.

In summary, although this study uses scientific research methods to evaluate the seismic risk of the frontal thrust belt in the southeast of Pamir and provides a basis for local implementation of scientific disaster prevention measures, several limitations remain. For example, the building attribute data is inferred from the sampling samples, which may introduce some errors. The selected fault scenario is inherently simplified; although the set earthquake reveals the potential worst damage scenario, they do not encompass all possible scenarios. The building earthquake damage matrix is derived from the survey data of historical earthquake damage in Xinjiang; hence, the applicable accuracy in specific areas under current construction practices needs to be verified. Additionally, the study did not carry out detailed site effect analysis and directly assumed the possible errors of bedrock conditions and the lack of research on secondary disasters. Only the seismic disaster risk analysis of buildings is carried out, and the spatial distribution of population and possible impact of secondary disasters such as casualties, economic losses, geological disasters, and fires have not been analyzed. Future research should address these aspects in greater depth. By incorporating comprehensive disaster risk factors and strengthening assessments of disaster reduction capacity, the evaluation results will become more comprehensive.

This study employs deterministic methods to investigate structural damage caused by earthquakes. Future research could utilize actual seismic inversion to construct precise seismic wave velocity models for ground motion simulation (Kianoush et al., 2023a). Fractal geometry is introduced to more accurately model the complexity and multi-scale characteristics of fault systems (Kianoush et al., 2024b). The concept of time-dependent deformation from rock and soil mechanics is applied to explore long-term seismic cycles, stress accumulation on PFTs, and other aspects of seismic hazard assessment (Khoshmagham et al., 2025; Adib et al., 2025b). Advanced numerical modeling techniques are employed to evaluate structural damage by considering the impact of complex physical interactions in thermal-porosity-elastic coupling on engineering stability (Pirhadi et al., 2025). The study also highlights that such events often trigger multiple secondary disasters, emphasizing the need for future research to focus on potential cascading risks in PFT regions (Rouhi et al., 2022).

7 Conclusion

In this study, we analyzed the seismic hazard risk of the southeastern section of the PFT rupture using the deterministic assessment method. The main conclusions are as follows:

1. The results of the peak acceleration obtained from the calculation of the setting earthquake indicate that the distribution of the extreme value zone has significant zoning characteristics and is concentrated in the setting earthquake region and its surrounding areas. The peak acceleration of the upper disk of the rupture (southwest side of the fault) is higher than that of the lower disk (northeast side of the rupture). The predicted intensities are all greater than 9°.

2. The results of remote sensing image interpretation, UAV photography, and field investigation indicate the presence of houses with civil, brick, brick-concrete, and steel-concrete structures with large differences in structural vulnerability. Within the near-field region, several lifeline infrastructures—including national, provincial, county, and township roads, along with water conservancy, electric power facilities, and other lifeline projects—are at risk. The Kashi–Hotan section of the Southern Railway also requires particular attention due to its seismic hazard risk.

3. The calculated housing damage indices (degree of damage) show that the northern and eastern parts of Aoyitake Town in Akto County, parts of Baren Town, northern part of Tashimilike Town in Shufu County, and northern part of Aigusi Town are expected to experience higher damage levels. Areas subjected to higher intensity shaking exhibit correspondingly higher damage indices, and under the same intensity conditions, areas with a high proportion of earth-wood-type houses show greater expected damage.

Overall, by combining the ground shaking prediction method with the vulnerability assessment of the disaster-bearing body, this study provides a scientifically grounded evaluation of the seismic hazard risk along the southeast section of the PFT rupture. This method is crucial for the scientific and reasonable evaluation of seismic hazard risk, effective pre-disaster preparedness, and reducing disaster losses.

Data availability statement

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

Author contributions

JL: Funding acquisition, Writing – original draft, Writing – review and editing. YY: Data curation, Formal Analysis, Funding acquisition, Writing – review and editing. LT: Formal Analysis, Investigation, Methodology, Writing – review and editing. WH: Formal Analysis, Investigation, Writing – review and editing. HW: Formal Analysis, Investigation, Writing – review and editing. XC: Formal Analysis, Methodology, Writing – review and editing.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This research was sponsored by Natural Science Foundation of Xinjiang Uygur Autonomous Region (Grant No. 2023D01A103), the Key Research and Development Program of Xinjiang Production and Construction Corps (Grant No. 2024AB077),Tianshan Talent Training Program (2023TSYCCX0097), and Xinjiang Earthquake Science and Technology mission special project (Grant No. 2025RWL02).

Acknowledgements

We sincerely thank the reviewers and associated editors for their valuable and constructive comments and suggestions, which helped to improve this paper.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Adib, A., and Kianoush, P. (2025a). Enhanced seismic hazard assessment and risk zoning in the kashan Region, central Iran: insights from historical data and advanced modeling techniques. Results Earth Sci. 3, 100098. doi:10.1016/j.rines.2025.100098

CrossRef Full Text | Google Scholar

Adib, A., and Kianoush, P. (2025b). Geotechnical and geological characterization of the Meskani mine Complex, Yazd block, Central Iran: a multidisciplinary study. Results Earth Sci. 3, 100072. doi:10.1016/j.rines.2025.100072

CrossRef Full Text | Google Scholar

Asghari, M., Maleki, Z., Solgi, A., Ganjavian, M. A., and Kianoush, P. (2025). Geohazard impact and gas reservoir pressure dynamics in the Zagros fold-thrust belt: an environmental perspective. Geosystems Geoenvironment 4 (2), 100362. doi:10.1016/j.geogeo.2025.100362

CrossRef Full Text | Google Scholar

Badawy, A., Korrat, I., El-Hadidy, M., and Gaber, H. (2017). Update earthquake risk assessment in Cairo, Egypt. J. Seismol. 21, 571–589. doi:10.1007/s10950-016-9621-5

CrossRef Full Text | Google Scholar

Boore, D. M. (2009). Comparing stochastic point-source and finite-source ground-motion simulations: SMSIM and EXSIM. Bull. Seismol. Soc. Am. 99 (6), 3202–3216. doi:10.1785/0120090056

CrossRef Full Text | Google Scholar

Boore, D. M., and Joyner, W. B. (1997). Site amplifications for generic rock sites. Bull. Seismol. Soc. Am. 87 (2), 327–341. doi:10.1785/bssa0870020327

CrossRef Full Text | Google Scholar

Boore, D. M., and Thompson, E. M. (2014). Thompson E M. Path durations for use in the stochastic-method simulation of ground motions. Bull. Seismol. Soc. Am. 104 (5), 2541–2552. doi:10.1785/0120140058

CrossRef Full Text | Google Scholar

Boore, D. M., and Thompson, E. M. (2015). Thompson E M. Revisions to some parameters used in stochastic-method simulations of ground motion. Bull. Seismol. Soc. Am. 105 (2A), 1029–1041. doi:10.1785/0120140281

CrossRef Full Text | Google Scholar

Brune, J. N. (1970). Tectonic stress and the spectra of seismic shear waves from earthquakes. J. Geophys. Res. 75 (26), 4997–5009. doi:10.1029/jb075i026p04997

CrossRef Full Text | Google Scholar

Cardona, O. D., Hurtado, J. E., Duque, G., Moreno, A., Chardon, A. C., Elasquez, L. S., et al. (2005). System of indicators for disaster risk management: main technical report. IDB/IDEA Programme of indicators for disaster risk management, Universidad Nacional de Colombia, Instituto de Estudios Ambientales, IDEA, Manizales.

Google Scholar

Cardona, O. D., Carreno, M., and Barbat, A. H. (2007). Urban seismic risk evaluation: a holistic approach. Nat. Hazards 40, 137–172. doi:10.1007/s11069-006-0008-8

CrossRef Full Text | Google Scholar

Chen, K., Gao, M. T., Yu, Y. X., Xu, W. J., Du, Y., Li, X. J., et al. (2023). Probabilistic seismic hazard analysis algorithm integrating three-dimensional fault sources and potential seismic source zone using random sampling. Seismol. Geol. 45 (02), 435–454. doi:10.3969/j.issn.0253-4967.2023.02.008

CrossRef Full Text | Google Scholar

Crowley, H., Colombi, M., Borzi, B., Faravelli, M., Onida, M., Lopez, M., et al. (2009). A comparison of seismic risk maps for Italy. Bull. Earthq. Eng. 7, 149–180. doi:10.1007/s10518-008-9100-7

CrossRef Full Text | Google Scholar

Crowley, H., Rodrigues, D., Silva, V., Despotaki, V., Romao, X., Castro, M., et al. (2018). “Towards a uniform earthquake risk model for Europe,” in In proceedings of the 16th European conference on earthquake engineering, Thessaloniki, Greece, 18–21.

Google Scholar

Crowley, H., Rodrigues, D., Silva, V., Despotaki, V., Martins, L., Romão, X., et al. (2019). “The European seismic risk model 2020 (ESRM 2020),” in Proceedings of the 2nd international conference on natural hazards & infrastructure, Chania, Crete Island (Greece), 23–26.

Google Scholar

Davidson, R. (1997). A multidisciplinary urban earthquake disaster risk index. Earthq. Spectra 13 (2), 211–223. doi:10.1193/1.1585942

CrossRef Full Text | Google Scholar

Duzgun, H. S. B., Yucemen, M. S., Kalaycioglu, H. S., Celik, K., Kemec, S., Ertugay, K., et al. (2011). An integrated earthquake vulnerability assessment framework for urban areas. Nat. Hazards 59, 917–947. doi:10.1007/s11069-011-9808-6

CrossRef Full Text | Google Scholar

Feng, X. Y., Nuan, C. Q., Li, J., and Zhang, Y. (1987). Study on the palaeo earthquakes along the KAZIKEARTE fault. Inland Earthquakes 1 (3), 232–239.

Google Scholar

Frolova, N. I., Larionov, V. I., Suschev, S. P., and Bonnin, J. (2011). Estimation of loss caused by earthquakes and secondary technological hazards. Nat Hazards Earth Syst Sci 11 (11), 3025–3033. doi:10.5194/nhess-11-3025-2011

CrossRef Full Text | Google Scholar

Ghalandari, S., Maleki, Z., Arian, M., Solgi, A., and Aleali, M. (2023). Basement faults effect on the folding style: a case study from Hendurabi fault, Zagros, Iran. Geotectonics 57 (4), 513–523. doi:10.1134/s0016852123040064

CrossRef Full Text | Google Scholar

Giannaraki, G., Kassaras, I., Roumelioti, Z., Kazantzidou-Firtinidou, D., and Ganas, A. (2018). Deterministic seismic risk assessment in the city of Aigion (W. Corinth Gulf, Greece) and juxtaposition with real damage due to the 1995 Mw6. 4 earthquake. Bull. Earthq. Eng. 17, 603–634. doi:10.1007/s10518-018-0464-z

CrossRef Full Text | Google Scholar

Gulia, L., Tormann, T., Wiemer, S., Herrmann, M., and Seif, S. (2016). Short-term probabilistic earthquake risk assessment considering time-dependent b values. Geophys. Res. Lett. 43, 1100–1108. doi:10.1002/2015gl066686

CrossRef Full Text | Google Scholar

Hanks, T. C. (1978). Characterization of high-frequency strong ground motion. (1978). Tectonophysics 49 (3-4), 263. doi:10.1016/0040-1951(78)90187-7

CrossRef Full Text | Google Scholar

Hanks, T. C., and Mcguire, R. K. (1981). The character of high-frequency strong ground motion. Bull. Seismol. Soc. Am. 71 (3), 1897–1919. doi:10.1785/bssa0710062071

CrossRef Full Text | Google Scholar

Hussain, E., Elliott, J. R., Silva, V., Vilar-Vega, M., and Kane, D. (2019). Contrasting seismic risk for Santiago, Chile, from near-field and distant earthquake sources. Nat. Hazards Earth Syst. Sci. 20, 1533–1555. doi:10.5194/nhess-20-1533-2020

CrossRef Full Text | Google Scholar

Jena, R., Pradhan, B., Beydoun, G., Nizamuddin, A., Sofyan, H., and Affan, M. (2020). Integrated model for earthquake risk assessment using neural network and analytic hierarchy process: aceh province, Indonesia. Geosci. Front. 11, (2) 613–634. doi:10.1016/j.gsf.2019.07.006

CrossRef Full Text | Google Scholar

Jian, H. Z., Wang, L. F., Ren, Z. K., Gong, W. Y., Li, Y. C., and Liu, J. R. (2020). Present-day slip rate and interseismic fault coupling along the Elashan fault using GPS. Chin. J. Geophys. 63 (03), 1127–1142. doi:10.6038/cjg2020N0229

CrossRef Full Text | Google Scholar

Karimzadeh, S., Miyajima, M., Hassanzadeh, R., Amiraslanzadeh, R., and Kamel, B. (2014). A GIS-based seismic hazard, building vulnerability and human loss assessment for the earthquake scenario in Tabriz. Soil Dyn. Earthq. Eng. 66, 263–280. doi:10.1016/j.soildyn.2014.06.026

CrossRef Full Text | Google Scholar

Khoshmagham, A., Hosseini Alaee, N., Shirinabadi, R., Bangian Tabrizi, A. H., Gholinejad, M., and Kianoush, P. (2025). Exploring the time-dependent behaviors of the rock mass roof of tunnels in open underground spaces of coal mines through numerical investigation. Sci. Rep. 15 (1), 8703. doi:10.1038/s41598-025-93017-1

PubMed Abstract | CrossRef Full Text | Google Scholar

Kianoush, P., Keshavarz Faraj Khah, N., Hosseini, S. A., Jamshidi, E., Afzal, P., and Ebrahimabadi, A. (2023). Geobody estimation by Bhattacharyya method utilizing nonlinear inverse modeling of magnetic data in Baba-ali iron deposit, NW Iran. Heliyon 9 (11), e21115. doi:10.1016/j.heliyon.2023.e21115

PubMed Abstract | CrossRef Full Text | Google Scholar

Kianoush, P., Mohammadi, G., Hosseini, S. A., Keshavarz Faraj Khah, N., and Afzal, P. (2023a). Inversion of seismic data to modeling the interval velocity in an oilfield of SW Iran. Results Geophys. Sci. 13, 100051. doi:10.1016/j.ringps.2023.100051

CrossRef Full Text | Google Scholar

Kianoush, P., Mohammadi, G., Hosseini, S. A., Keshavarz Faraj Khah, N., and Afzal, P. (2023b). Determining the drilling mud window by integration of geostatistics, intelligent, and conditional programming models in an oilfield of SW Iran. J. Petroleum Explor. Prod. Technol. 13 (6), 1391–1418. doi:10.1007/s13202-023-01613-6

CrossRef Full Text | Google Scholar

Kianoush, P., Mohammadi, G., Hosseini, S. A., Keshavarz Faraj Khah, N., and Afzal, P. (2023c). ANN-based estimation of pore pressure of hydrocarbon reservoirs—a case study. Arabian J. Geosciences 16 (5), 302. doi:10.1007/s12517-023-11373-6

CrossRef Full Text | Google Scholar

Kianoush, P., Keshavarz Faraj Khah, N., Afzal, P., Jamshidi, E., Bangian Tabrizi, A. H., and Kadkhodaie, A. (2024b). Formation pressures determination utilizing the integration of fractal and geostatistical modelling in a hydrocarbon formation of SW Iran. J. Anal. Numer. Methods Min. Eng. 14 (40), 11–34.

Google Scholar

Kircher, C. A., Whitman, R. V., and Holmes, W. T. (2006). HAZUS earthquake loss estimation methods. Nat. Hazards Rev. 7, 45–59. doi:10.1061/(asce)1527-6988(2006)7:2(45)

CrossRef Full Text | Google Scholar

Lagomarsino, S., and Giovinazzi, S. (2006). Macroseismic and mechanical models for the vulnerability and damage assessment of current buildings. Bull. Earthq. Eng. 4, 415–443. doi:10.1007/s10518-006-9024-z

CrossRef Full Text | Google Scholar

Li, X. J. (2001). Comments on phenomena of great earthquake damages and engineering seismological researches. Recent Dev. World Seismol. (8), 26–32.

Google Scholar

Li, Q., Chen, X. L., Gao, M. T., and Li, Z. C. (2022). Overview of research progress on numerical simulation in physics-based seismic hazard analysis. World Earthq. Eng. 38 (02), 239–250.

Google Scholar

Marulanda, M. C., Carreno, M. L., Cardona, O. D., Ordaz, M. G., and Barbat, A. H. (2013). Probabilistic earthquake risk assessment using CAPRA: application to the city of Barcelona. Spain. Nat. Hazards 69, 59–84. doi:10.1007/s11069-013-0685-z

CrossRef Full Text | Google Scholar

Mili, R. R., Hosseini, K. A., and Izadkhah, Y. O. (2018). Developing a holistic model for earthquake risk assessment and disaster management interventions in urban fabrics. Int. J. Disaster Risk Reduct. 27, 355–365. doi:10.1016/j.ijdrr.2017.10.022

CrossRef Full Text | Google Scholar

Mouroux, P., and Brun, B. L. (2006). Risk-Ue project: an advanced approach to earthquake risk scenarios with application to different European towns. Netherlands: Springer.

Google Scholar

Mouroux, P., and Brun, B. L. (2008). “Risk-UE project: an advanced approach to earthquake risk scenarios with application to different European towns,” in Assessing and managing earthquake risk (Springer), 479–508.

Google Scholar

Mualchin, L. (2011). History of modern earthquake hazard mapping and assessment in California using a deterministic or scenario approach. Pure Appl. Geophys. 168 (3-4), 383–407. doi:10.1007/s00024-010-0121-1

CrossRef Full Text | Google Scholar

Panagiota, M., Jocelyn, C., Erwan, P., and Philippe, G. A. (2012). A support vector regression approach for building seismic vulnerability assessment and evaluation from remote sensing and in-situ data. 2012 IEEE International Geoscience and Remote Sensing Symposium, 7533–7536. doi:10.1109/IGARSS.2012.6351888

CrossRef Full Text | Google Scholar

Pavel, F., and Vacareanu, R. (2016). Scenario-based earthquake risk assessment for Bucharest, Romania. Int. J. Disaster Risk Reduct. 20, 138–144. doi:10.1016/j.ijdrr.2016.11.006

CrossRef Full Text | Google Scholar

Pirhadi, A., Kianoush, P., Ebrahimabadi, A., and Shirinabadi, R. (2023). Wellbore stability in a depleted reservoir by finite element analysis of coupled thermo-poro-elastic units in an oilfield, SW Iran. Results Earth Sci. 1, 100005. doi:10.1016/j.rines.2023.100005

CrossRef Full Text | Google Scholar

Pirhadi, A., Kianoush, P., Varkouhi, S., Shirinabadi, R., Shirazy, A., Shirazi, A., et al. (2025). Thermo-poroelastic analysis of drilling fluid pressure and temperature on wellbore stresses in the Mansouri oilfield, SW Iran. Results Earth Sci. 3, 100061. doi:10.1016/j.rines.2025.100061

CrossRef Full Text | Google Scholar

Rouhi, J., Delchiaro, M., Della Seta, M., and Martino, S. (2022). New insights on the emplacement kinematics of the Seymareh landslide (Zagros Mts., Iran) through a novel spatial statistical approach. Front. Earth Sci. 10, 869391. doi:10.3389/feart.2022.869391

CrossRef Full Text | Google Scholar

Salgado-Gálvez, M. A., Bernal, G. A., Zuloaga, D., Marulanda, M. C., Cardona, O.-D., and Henao, S. (2017). Probabilistic seismic risk assessment in Manizales, Colombia: quantifying losses for insurance purposes. Int. J. Disaster Risk Sci. 8, 296–307. doi:10.1007/s13753-017-0137-6

CrossRef Full Text | Google Scholar

Sembroni, A., Reitano, R., Faccenna, C., and Callieri, P. (2024). The geologic configuration of the Zagros fold and thrust belt: an overview. Mediterr. Geosci. Rev. 6 (2), 61–86. doi:10.1007/s42990-024-00118-6

CrossRef Full Text | Google Scholar

Sheng, Z. Q., Hu, J. J., and Xie, L. L. (2023). Review on the simulation of input ground motion for fault-crossing structure and its application. Earthq. Eng. Eng. Dyn. 43 (01), 1–13. doi:10.13197/j.eeed.2023.0101

CrossRef Full Text | Google Scholar

Shi, P., and Xu, W. (2016). “Natural disaster system in China,” in Natural disasters in China (Springer), 1–36.

Google Scholar

Silva, V., Crowley, H., Varum, H., and Pinho, R. (2015). Seismic risk assessment for mainland Portugal. Bull. Earthq. Eng. 13, 429–457. doi:10.1007/s10518-014-9630-0

CrossRef Full Text | Google Scholar

Silva, V., Crowley, H., and Bazzurro, P. (2016). Exploring risk-targeted hazard maps for Europe. Earthq. Spectra 32, 1165–1186. doi:10.1193/112514eqs198m

CrossRef Full Text | Google Scholar

Sinha, N., Priyanka, N., and Joshi, P. (2014). Using spatial multi-criteria analysis and ranking tool (SMART) in earthquake risk assessment: a case study of Delhi region, India. Geomat. Nat. Hazards Risk 7, 680–701. doi:10.1080/19475705.2014.945100

CrossRef Full Text | Google Scholar

Tadjer, K., and Bensaibi, M. (2017). Earthquake risk assessment of Blida (Algeria) using GIS. Energy Procedia 139, 645–650. doi:10.1016/j.egypro.2017.11.266

CrossRef Full Text | Google Scholar

Tavakoli, V., and Barfizadeh, H. (2024). The role of plate movements on reservoir development of the Iranian carbonate formations: a review of the interplay between tectonics, paleoclimate, and diagenesis. Results Earth Sci. 2, 100037. doi:10.1016/j.rines.2024.100037

CrossRef Full Text | Google Scholar

U UNISDR (2009). Terminology on disaster risk reduction. Geneva, Switzerland.

Google Scholar

Xu, Y., and Chen, M. W. (2025). Multi-source input based risk analysis of typical twin-tower cable-stayed bridges in coastal areas under seismic-wind combined actions. J. Southeast Univ. Sci. Ed. 55 (05), 1301–1310.

Google Scholar

Yu, X. C., An, W. J., and He, H. (2012). A method of auto classification based on object oriented unsupervised classification. Prog. Geophys. 27 (2), 744–749. doi:10.6038/j.issn.1004-2903.2012.02.042

CrossRef Full Text | Google Scholar