Study on the seismic lethal level of buildings and seismic disaster risk in Guangzhou, China

Abstract

Seismic disaster risk assessment results are an important way to reduce disaster losses and casualties. In this article, a seismic disaster risk assessment method is constructed based on the lethal level, and a quantitative assessment of the earthquake disaster risk in Guangzhou is carried out. There are four typical building types in this study area: brick-wood structures, brick-concrete structures, reinforced concrete structures, and civil structures. Typical building structural components (vertical load-bearing component, ring beam, mortar of infilled wall, roof or floor structure, foundation, height of building, nonstructural component, anti-seismic joint or other anti-seismic measures) are selected to build the lethal level index system of buildings. On the basis of the field survey, a lethal level of 431 survey points is obtained. The average lethal level in urban areas is 0.3476, and in rural areas is 0.4198, the overall lethal level of each township is 0.389, and the errors between the urban and rural areas are 0.0414 and 0.0308, and the fitting results coefficient of determination (R²) values are 0.6803 and 0.6714, respectively, the results is good. Furthermore, considering the seismic hazard (peak ground acceleration, PGA), population exposure and lethal level, the seismic disaster risk distribution of the Guangzhou area is determined. Under different intensities, the risk of death has shown obvious characteristics, as the intensity increases, the number of possible deaths increases rapidly, ranging from a few to several thousand, the higher risk areas are relatively fixed, generally mainly located in the central city, the quantitative assessment results of disaster risk also have obvious directivity. The areas with the highest risk are located in most areas of Tianhe District, Liwan District and Haizhu District, the highest grid death result is more than 1; the lethal level of Baiyun District and Huangpu District is low, but the death risk is also higher, generally between 0.5 and 1, in rural areas in the southern and northern regions of Guangzhou, the risk is relatively low, generally between 0 and 0.5. The distribution map of seismic disaster risk can provide solid scientific and technological support for the government to investigate potential risk points.

Keywords:

• Seismic disaster risk
• lethal level
• fatality
• quantification

Introduction

Earthquakes are sudden and unpredictable, easily causing damage to buildings; additionally, secondary disasters, such as sand liquefaction, collapse, landslides, and mudslides, which often cause tremendous casualties and economic losses (Smith 2013). Seismic disaster risk assessment can provide an important scientific and technical basis for regional development planning, earthquake relief work, and effective reduction of casualties and property losses caused by earthquakes (Liu and Chen 2019). The seismic disaster risk is measured by its potential damage to buildings, as well as the number of people expected to be injured or killed, the process includes 3 steps: first, the earthquake hazard is estimated (earthquakes with probable intensity in a given geographical area); second, the damage of buildings based on different vulnerability assessment methods is evaluated; and third, based on the available data from the exposure model, the risk of losses, such as casualties and economic losses during the earthquake, is evaluated (Alexandru et al. 2017; Chieffo and Formisano 2019). At present, there are many different earthquake disaster risk research projects (Li et al. 2015), for example, global natural disaster hotspot projects, using the population death and economic loss recorded by historical earthquakes, the population death loss rate and economic loss rate caused by earthquake disasters as the vulnerability coefficient can be calculated, and then the coefficient with the earthquake frequency can be modified to estimate the world's earthquake disaster population death and economic loss risk level (Badal et al. 2005). Furthermore, for the Global Risk and Vulnerability Index Trends Program, using the data of the hazard factor intensity and the number of deaths recorded in historical earthquakes, the vulnerability of the earthquake can be calculated, and a loss function is constructed in combination with other economic indicators to estimate the average annual expected loss (UNDP. 2010). Apart from this, the research that addresses the indirect impact of earthquakes on communities (Miles and Chang 2011), which are focused on earthquake casualty estimation, such as the Interdependent Networked Community Resilience Modeling Environment (IN-CORE) which is under development as part of collaborative research work supported by the National Institute of Standards and Technology (NIST) (Milad 2020; Milad et al. 2021), the Prompt Assessment of Global Earthquakes for Response (PAGER) which is rapidly assesses earthquake impacts by comparing the population exposed to each level of shaking intensity with models of economic and fatality losses based on past earthquakes in each country or region of the world supported by the United States Geological Survey(USGS) (Jaiswal et al. 2011; Jaiswal and Wald 2008). The ultimate goal of seismic disaster risk assessment is to create a safer environment. This assessment can provide scientific support for earthquake prevention and disaster reduction planning, reduction of life and property losses, and emergency rescue after earthquakes.

Many scholars have carried out various studies on the cause mechanism of earthquakes, ground motions, building vulnerability, rapid assessment methods and disaster risk assessment. Based on different research purposes, different scholars have proposed various evaluation methods, including seismic hazard analysis (Ademovi et al. 2020; Ongkowijoyo et al. 2020; Jena et al. 2020a,b; Marshall 2020; Tsang et al. 2020; Zheng et al. 2020), building vulnerability analysis (Bessason et al. 2020, 2022; Maqsood et al. 2016; Gautam et al. 2020, 2021) and risk analysis (Chettri et al. 2021; Verderame et al. 2014), among them, building vulnerability assessment is a central component that includes hazard identification along with an accounting of social, cultural, economic, and built systems impacted (Bessason et al. 2012). Generally, the method is mainly divided into two types, qualitative methods and quantitative methods, the qualitative risk assessments are mainly based on expert knowledge, whereas quantitative risk assessments use mathematical methods (Bühlmann 1996; Li et al. 2018). Whether it is a qualitative method or a quantitative method, there is a certain degree of uncertainty; therefore, the expression and handling of fuzzy and random information is a key issue in modern risk assessment modeling (Yu 2017).

Although several different approaches have been developed in the world in recent years to estimate the seismic risk for whole countries or larger cities, all of these approaches are examples of very detailed studies that require a large amount of data (in particular, a detailed building database) (Bessason and Bjarnason 2016), which are not available in all countries. Although a greater focus has been placed on seismic hazards and vulnerability assessment in the scientific articles researched thus far, very little attention has been given to seismic exposure and the relationship to assessing seismic vulnerability. It can be said that risk assessment is a quantitative alternative to earthquake preparedness and response. The seismic disaster risk assessment process combines the results of the analyses of seismic hazards, exposed physical and social values, and seismic damage (Pavic et al. 2020).

Therefore, how to establish the relationship between buildings, etc. and the probability of death, and conduct risk assessment based on this, is the key to quantitative earthquake disaster risk assessment. This article is based on the lethal level obtained from field investigations, and through the improvement of the traditional earthquake disaster risk assessment model, the risk assessment method of earthquake casualties is carried out. This article is based on the lethal level of various types of buildings obtained by field investigation to obtain the overall lethal level of the area. The lethal level of area actually reflects when the area under different intensity, the possible probability of death is actually converting the probability of building damage, this reducing the uncertainty of traditional process of directly converting the proportion of building damage to the proportion of death, such as the "destroyed but not collapsed" situation of buildings after the earthquake, that is, serious damage to the building, but it does not necessarily cause death. Based on this, we can divide an area into different grid units, and each grid unit corresponds to a group of intensity personnel mortality. In this way, as long as the intensity range of each grid unit in the earthquake area is determined, it can be the rapid assessment and calculation based on the population, based on the mortality rate of each intensity represented by the lethal level, combines high-precision population data, and improves the accuracy and applicability of the results, it also provides a new idea and scientific basis for the pre-earthquake risk assessment and rapid post-earthquake assessment of earthquake casualties.

Data and method

Data

The data used in this paper include peak ground acceleration (PGA) data, historical loss data from earthquake disasters (mortality rate by intensity), grid population data, and the regional lethal level of field survey points, as shown in Table 1.

Table 1. Data resource and description.

Name	Time	Resolution	Source	Data description
Peak ground acceleration	China Earthquake Parameter Zoning Map GB18306-2015	Township level	China Earthquake Parameter Zoning Map	The 50a exceedance probability, 10% exceedance probability (that is, the recurrence period of the earthquake is 475 years).
Loss of data	1967-2014	Single earthquake	China Earthquake Disaster Loss Assessment Report	Earthquake occurrence time, magnitude, latitude and longitude, number of casualties, population density, mortality rate, etc.
population	2020	0.1 km*0.1 km	World pop project	The community standard of the world population distribution is the data with the highest resolution of the world population distribution.
Regional lethality level	2020	Township grid level	Field investigation	The lethality level results are obtained based on field investigations.
Mortality rate	2020	Lethal level and intensity	Xia et al. 2020	Based on the lethality matrix, obtained mortality data of different intensity under different lethal levels.

The population grid data are provided by the World Pop Project (Bondarenko et al. 2020), and the accuracy of the data is 0.1 km*0.1 km. The data have high accuracy and can better reflect the distribution characteristics of the population in the study area. The specific distribution results are shown in Figure 1.

Figure 1. Results of the population grid distribution in the study area (0.1 km*0.1 km).

Method

The introduction highlights the need to create a reliable estimating model of earthquake losses that will cover all possible scenarios that arise during and after an earthquake. And many research groups develop a better framework, such as the principles and procedures behind seismic hazard and risk analysis (Baker et al. 2021), the new relationships between collapse and fatality rate, and proposing guidelines to rate functions in several quality dimensions (Porter et al. 2012), selected ground motion records by employing the information provided from the seismic hazard analysis (Elefante et al. 2010), based on different focuses, the evaluation methods and frameworks adopted are also different, based on these research results, we propose the framework of this study as shown in Figure 2. The basic concept of the algorithm consists of seismic hazard evaluation, lethal level assessment, and earthquake risk assessment.

Figure 2. Technical process of the earthquake disaster risk assessment method based on the lethal level.

The basic process of the earthquake disaster risk assessment method is as follows: first, based on the PGA data and other parameters and the empirical attenuation model, obtain the possible intensity distribution location results in the area; second, experience or analysis methods are used to obtain the vulnerability analysis results of the area; and third, disaster risk assessment is performed based on the calculation results of the first two steps. This assessment method often only considers the possibility of different degrees of damage to the buildings in an area after the earthquake; that is, this method only considers the different degrees of damage corresponding to different mortality rates. There is no analysis of the ability of different buildings to cause casualties, which can easily cause errors.

In fact, after an earthquake, after a building is destroyed, different buildings have different possibilities of causing deaths. This difference is actually directly reflected in the mortality rate of different intensities. The lethal level and intensity in an area are different, so even at the same level of vulnerability and damage to buildings, the possibility of death is different. That is, the lethal level of an area and the intensity may together determine the level of earthquake risk. In this paper, the traditional method is changed. On the basis of obtaining the intensity distribution range in the first step, based on the lethal level in each region (districts, counties, towns), obtaining the corresponding mortality rate of the area under a certain intensity, combined with the results of high-precision population distribution, a quantitative assessment of earthquake disaster risk is carried out. Therefore, the acquisition of the lethality matrix and the regional lethal level is the key to it.

Matrix of the lethal level

The lethal level represents the comprehensive level of casualties after the earthquake. This indicates the mortality rate at different intensities. The higher the lethal level is, the higher the corresponding mortality rate at each intensity (Nie et al. 2020; Xia 2020, Xia et al. 2021). After an earthquake, the factors leading to death are very complex, including building destruction, wall collapse, gate tower collapse, falling wall hangings, rolling rocks, landslides, fires, and inadequate treatment. To comprehensively describe the possible level of deaths caused by earthquakes in various disaster areas, a new concept needs to be introduced, namely, the regional lethal level. The so-called lethal level is the comprehensive possibility of a variety of factors leading to death after a building is destroyed. It is expressed by the possible mortality rate of each intensity under a certain lethal level and under the same magnitude and intensity. Since the lethal level corresponds to the mortality rate of a component of intensity, the higher the lethal level of an area, the higher the corresponding mortality rate of the individual intensities. Therefore, this article is based on previous research results and uses the lethal level matrix as the basis for calculating the vulnerability data of personnel, as shown in Table 2 (Xia 2020).

Table 2. Matrix based on the lethal level and intensity.

No	Lethal level	Mortality rate of each intensity
		VI	VII	VIII	IX	X	XI
1	0-5%	0.0000005	0.0000054	0.0000538	0.0005382	0.0053817	0.0538175
2	5-15%	0.0000011	0.0000102	0.0000542	0.0019236	0.0160393	0.1148306
3	15-25%	0.0000022	0.0000140	0.0000802	0.0024175	0.0208464	0.1476940
4	25-35%	0.0000028	0.0000193	0.0001187	0.0030381	0.0270941	0.1658035
5	35-45%	0.0000036	0.0000265	0.0001756	0.0038181	0.0352142	0.1826483
6	45-55%	0.0000047	0.0000365	0.0002599	0.0047983	0.0457680	0.2012044
7	55-65%	0.0000060	0.0000502	0.0003845	0.0060301	0.0594847	0.2216456
8	65-75%	0.0000078	0.0000690	0.0005689	0.0075782	0.0773124	0.2441637
9	75-85%	0.0000101	0.0000949	0.0008418	0.0095237	0.1004830	0.2585556
10	85-95%	0.0000131	0.0001305	0.0012456	0.0119687	0.1305980	0.2828812
11	95-100%	0.0000169	0.0001794	0.0018431	0.0150414	0.1697384	0.3094955

The lethality matrix in this article is mainly based on previous related research results, in the relevant literature (Xia 2020), the lethality matrix has been elaborated. The model construction based on historical earthquake data, such as building damage data, mortality data, etc., the damage ratio or damage probability of different types of buildings under different intensity ranges has been considered. This probability has been converted into the mortality rate of different intensities in the matrix. In fact, the traditional mechanics-based vulnerability functions focus on the damage probability of various types of buildings in different intensity areas, and then obtain the damage ratio based on the damage probability, and then calculate the number of deaths. We use the lethal level of each type to express the possibility of death caused by the destruction of buildings, this probability can be linked to intensity and mortality in terms of lethal levels. Therefore, this is also a new way of thinking and method different from traditional methods.

The seismic hazard evaluation

This article mainly adopts probabilistic seismic risk assessment (PSHA) methods, which are used to estimate the probabilities of experiencing a certain value of the selected ground motion parameter (Pavic et al. 2020), with the seismic hazard now being represented by PGA values. Numerous studies have also shown that there is actually a clear correlation between seismic intensity and seismic parameters. There have been many studies on the correlation between seismic intensity and ground motion parameters, such as the study of the relationship between intensity and PGA (Hershberger 1956; Sweny 2012; Tian et al. 2021) and research on the relationship between intensity and peak ground velocity (Ahmadzadeh et al. 2020; Wu et al. 2003). Based on different research methods, the quantitative relationship between different seismic intensities and different ground motion parameters is obtained (Baorong et al. 2017). Therefore, the key to earthquake disaster risk assessment is to determine the intensity level and range that the area may suffer and realize the conversion of ground motion parameters to seismic intensity.

Since PGA has been widely used in related work and research, such as earthquake disaster risk assessment and mapping, we have obtained the relationship model between PGA and seismic intensity based on related research results (Baorong et al. 2017), as shown in Equation (1): (1) $$I=3.73\hspace{0.17em}\text{log}\hspace{0.17em}\mathit{PGA}-1.23$$(1) where I is the intensity and PGA is the peak ground acceleration.

On this basis, the PGA of each township is converted into the corresponding seismic intensity. Based on this result, the hazard grid map with only seismic intensity information is vectorized, and a grid result of the intensity distribution is obtained. To determine the mortality rate of each grid in the study area, another major factor is the lethal level in the grid. Therefore, determining the distribution of the lethal level of the grid is the key.

Calculation method of the regional lethal level

Different building types have different probability levels to cause casualties after an earthquake. Even for the same types of buildings, due to a series of factors such as construction methods, construction materials, and construction ages, their respective lethal levels are different. Based on the range of lethal levels of different buildings, through field investigations, the lethal levels of different types of buildings are obtained. On this basis, combined with the proportion data of various types of buildings, the lethal level of an area can be obtained.

The lethal level range of different types of buildings is mainly based on the author’s previous research results, as shown in Table 3. Based on the analysis and calculation of historical earthquakes, the lethal levels of various types of buildings are obtained, and it is found that the lethal level of each type of building is a relatively fixed interval range, therefore, even for the same type of building, due to the influence of different factors such as building materials, building quality, and construction age, the actual lethal level is not completely the same, but no matter what the change, the actual lethal level is within the range of the lethal level of this type of building. At the same time, the lethal level range of various types of buildings is a non-equal interval range, and there are overlapping areas between each other, this is also consistent with the actual situation, the lethal level of a certain type of building is not necessarily lower than that of another type of building, for example, the lethal level of a brick-wood structure with a relatively good construction quality may be lower than that of a brick-concrete structure with a poor construction quality and a longer construction age (Xia 2020).

Table 3. The interval and influencing factors of the lethal level of buildings.

Building structure type			Lethal level range
Building Type	Secondary classification	Influencing factors
Steel structure		Construction measures, foundation type, construction time, use type	0.05-0.15
Reinforced concrete structure	RCa	Structural measures, foundation type, construction time, height	0.1-0.3
	RCb
Wood structure	Wa	Structural measures, foundation type, construction time, structural style	0.2-0.4
	Wb
Brick-concrete structure	Ba (Fortified)	Structural measures, foundation type, construction time, use type, height,	0.25-0.7
	Bb (Nonfortified)
Brick-wood structure		Structural measures, foundation type, construction time, structural style	0.6-0.9
Civil structure		Construction measures, foundation type, construction time, building materials	0.7-0.95
Stone and wood structure		Wall type, foundation type, construction time, building materials	0.55-0.9
Adobe structure		Wall type, foundation type, construction time, building materials	0.85-1

Instruction: The construction materials of civil structures mainly include bamboo, wood, rammed earth, straw, hay, adobe bricks and tiles. The structural form commonly used in rural residential houses is that bamboo strips are used instead of steel bars, and clay (some of which will be mixed with straw) instead of concrete is tamped to make the walls of houses. Various types of houses in various architectural forms (https://baike.baidu.com/item/%E5%9C%9F%E6%9C%A8%E7%BB%93%E6%9E%84/405111).

Based on the interval range of each type of building and the influence weight ratio coefficient of each influencing factor, a calculation method for the actual lethal level of each type of building is constructed, as shown in Equation (2)(2) $$C{L}_{\mathit{actual}}=C{L}_{\max }-\sum _{i=1}^n\left(C{L}_{\max }-C{L}_{\min }\right)\cdot w_i\cdot w_{ij}$$(2) : (2) $$C{L}_{\mathit{actual}}=C{L}_{\max }-\sum _{i=1}^n\left(C{L}_{\max }-C{L}_{\min }\right)\cdot w_i\cdot w_{ij}$$(2)

Among them, CL_actual is the actual lethal level of a certain type of building, CL_max is the upper limit of the lethal level interval for a certain type of building, CL_min is the lower limit of the lethal level interval of a certain type of building, and n is the type of factor affecting the lethal level of a certain type of building, which is selected according to the actual conditions of different types of buildings. w_i is the weight ratio coefficient of the i-th type of influencing factor in the building, and the selection is based on the actual situation of the building. w_ij is the weight ratio coefficient of the j-th building structure among the i-th type of influencing factor in the building, and it is selected based on the actual situation of the building.

Based on the lethal level of various types of buildings and the proportional data of each type of building, a calculation model for the overall lethal level of the survey area is established, as shown in Equation (3)(3) $$C{L}_{\mathit{all}}=\sum _{i=1}^{n}C{L}_{\mathit{actual}-i}\cdot P_i$$(3) : (3) $$C{L}_{\mathit{all}}=\sum _{i=1}^{n}C{L}_{\mathit{actual}-i}\cdot P_i$$(3) where CL_all is the lethal level of the region based on buildings, CL_actual-i is the actual lethal level of the i-th type of building in the region, and P_i is the proportion of the i-th type of building.

On the basis of the lethal level of each survey site, a lethal level of the town-level calculation method is constructed based on the population proportion coefficient, as shown in Equation (4)(4) $$C{L}_{\mathit{town}}=\alpha \cdot C{L}_{\mathit{township}}+\beta \cdot C{L}_{\mathit{countryside}}$$(4) : (4) $$C{L}_{\mathit{town}}=\alpha \cdot C{L}_{\mathit{township}}+\beta \cdot C{L}_{\mathit{countryside}}$$(4) where CL_town is the overall lethal level of the township administrative unit, CL_township is the lethal level in the township area, CL_countryside is the lethal level of the village in the administrative unit of the township, α is the ratio coefficient of the population of the town to the population of the whole township, β is the ratio coefficient of the population of the administrative village to the population of the whole town, and α + β = 1.

Earthquake disaster risk model

Due to the influence of many factors such as buildings, geographical environment, and traffic roads, the lethal levels of various regions are different. Two adjacent administrative regions are in the same intensity range, but due to the different lethal levels, the corresponding mortality rates are also different. Similarly, for the same area, the lethal level is the same, but because the area is in a different intensity range, the corresponding mortality rate is also different. Based on this, each intensity can be divided into different ranges based on the lethal level, and if different regional ranges are grid-processed, then different grids correspond to different population numbers and mortality rates. Based on the mortality rate and population, an earthquake disaster risk assessment method can be carried out, as shown in Equation (5)(5) $$D=\sum f\left(L/I\right)\cdot P$$(5) : (5) $$D=\sum f\left(L/I\right)\cdot P$$(5) where f(L/I) is the mortality rate of a certain grid at a certain intensity level at a certain lethal level, P is the number of people, that is, the population grid data, and D is the number of possible deaths number.

Results

The field survey method adopts the "town to town" method; that is, each administrative unit at the town level conducts a field survey, and the streets in the urban area are determined according to the town level. Each town unit selects 3–4 administrative villages as survey points. The survey points include 1 downtown area (central area) and 2–3 representative administrative villages. The contents of the survey include structural characteristics, construction methods, building materials, adhesive types, foundation types, roof types, construction years, connection methods and foundation forms of the samples of various building structures in the survey sites, as well as the geographical and geomorphic environment of the area, traffic road conditions, building ancillary structure conditions, etc.

However, because the building materials and methods in different cities are not completely the same, the types of buildings are uniformly classified according to structural materials, including the steel structure, reinforced concrete structure, brick-concrete structure (fortified brick-concrete, nonfortified brick-concrete), brick and wood structure, wood structure, civil structure and traditional structure types unique to each region (determined according to the actual situation of the survey area).

As shown in Figure 3, Guangzhou city involves 11 districts and counties, including a total of 171 townships. The survey points are 431, the distribution of survey points achieves full coverage at the township level. The coverage is uniform, and the lethal level obtained by integrating the survey points of each township can represent the average situation of each town.

Figure 3. Distribution map of field survey points in Guangzhou city.

Based on the results of the field survey, the wall types of buildings in Guangzhou mainly include reinforced concrete structures, brick-concrete structures (fortified brick-concrete and nonfortified brick-concrete), brick-wood structures, and civil structures. The areas with good earthquake resistance are concentrated in the Baiyun, Huangpu and Tianhe districts, however, in the central areas of cities such as the Yuexiu, Liwan and Haizhu districts, the earthquake resistance of buildings is poor, the main reason is that although there are a large number of buildings with good construction quality, such as reinforced concrete structure in the central area of the city, there are also a large number of old brick-wood and brick-concrete structures, most of the construction time was from the 1970s to the 1990s, the building height is generally approximately 1–3 floors, the walls are mainly empty bucket walls, the wall thickness is generally 18 cm, and the roof types are mostly cast-in-roofs, with thicknesses of approximately 10 cm. In addition, most brick-concrete buildings do not have ring beams and structural columns, and the overall building quality is poor, with weak earthquake resistance and a high lethal level.

Based on the field investigation results, another obvious feature of buildings in Guangzhou is that the architectural form and quality of buildings are greatly affected by economic development. For self-built brick-concrete structures, there are a large number of buildings without structural columns, the first-floor roof is mainly a prefabricated slab roof, and with the improvement of economic level, the second floor added in the later stage is mainly cast-in-place roofs, which is also one of the reasons for the higher lethal level.

There are a large number of urban villages in the urban area of Guangzhou, and the construction age of the buildings in the urban area is generally in the 1970s and 1990s. The structural measures of the buildings are poor, there are no ring beams and no structural columns, and the building height is generally above 2 stories, another main feature is that the building spacing is small, especially in the central area where there are a large number of buildings above 5 floors, the layout of the buildings is messy and the building spacing is generally within 2 m. Another important aspect is that there are many hanging objects outside the walls of buildings, and there are a large number of nonstructural and accessory facilities such as air-conditioning external hanging objects and billboards.

The buildings in the rural areas of Guangzhou show different characteristics, the buildings are mainly brick-concrete structures, and also include civil and brick-wood structures, the walls are mostly 18 walls, and the wall bonding material is generally lime mortar or loess mud, the foundation is shallow, generally a gravel foundation of approximately 80–110 cm, the quality of the building is poor, the overall earthquake resistance of the building is not strong, and the lethal level is high, as shown in Table 4.

Table 4. Types and specific characteristics of buildings in Guangzhou.

Building type		Field research pictures		Specific characteristics
Brick-wood structure	a			A 24 cm wall, the bonding material is cement mortar or lime mortar, and the gravel foundation is approximately 1 m.
	b			An 18 cm wall, the bonding material is lime mortar or loess mud, and the gravel foundation is approximately 1 m.
Brick-concrete structures	a			There are ring beams, structural columns, a cast-in-place roof, a 24 cm wall, and a 1.5 m reinforced concrete foundation.
	b			There are ring beams, structural columns, a prefabricated slab roof, a 24 cm wall, and a 1.5 m brick foundation.
	c			With ring beams, no structural columns, a cast-in-place roof, a 24 cm wall, and an approximately 1.5 m concrete foundation or brick foundation.
	d			There are ring beams, no structural columns, a prefabricated slab roof, 24 cm wall, and 1.5 m brick foundation.
	e			No ring beams, no structural columns, a prefabricated slab roof, a 24 cm wall, and a 1.5 m brick foundation.
Reinforced concrete structure				Standard frame structure, concrete frame, complete beam and column system, and concrete foundation.
Civil structure				Without wooden pillar support, the roof is a sloping roof, the walls are mud bricks or adobe walls, and the soil is bonded or unbonded.
				Small space between buildings, many external hanging objects and ancillary structures.

As shown in Figure 4, based on the lethal level and the proportion data of each type of building in 431 survey sites, the actual lethal level of each survey site is obtained. The lethal level distribution shows obvious regional characteristics. The lethal level is relatively high in the rural area of the new city and in the urban center of the old city, and the lethal level in several other districts and counties is relatively low, this also shows a clear difference from the distribution characteristics of cities in other regions.

Figure 4. Distribution of lethality levels based on field survey results.

Based on the survey sites of each township and the population coefficients of different survey sites, the lethal level of each town is obtained, as shown in Table 5. Based on the logarithmic cumulative normal distribution function of the lethal levels in urban and rural areas, the lethal level in urban areas is much lower than that in rural areas. Among them, the average lethal level in urban areas is 0.3476, and the standard deviation is 0.1242. The average lethality level in rural areas is 0.4198, and the standard deviation is 0.117, which is also consistent with the actual situation, as shown in Figure 5. Generally, the economic level of urban areas is higher than that of rural areas, and the quality of buildings and earthquake resistance are also better than those of rural areas.

Figure 5. The log cumulative normal distribution of the lethal level in urban and rural areas.

Table 5. Calculation results of the lethal level in various townships in Guangzhou.

No.	Location	Urban level	Urban coefficient	Rural level	Rural coefficient	Lethal level
1	Timian Town, Huadu District	0.5155	0.2667	0.577	0.7333	0.560598
2	Shiling Town, Huadu District	0.365	0.3	0.475	0.7	0.442
3	Chini Town, Huadu District	0.5925	0.058	0.6425	0.942	0.6396
4	Xinya Street, Huadu District	0.48	0.4324	0.435	0.5676	0.454458
5	Xinhua Street, Huadu District	0.49125	0.4915	0.485	0.5085	0.488072
6	Xiuquan Street, Huadu District	0.2	0.1855	0.4925	0.8145	0.438241
7	Tanbu Town, Huadu District	0.57	0.306	0.6775	0.694	0.644605
8	Huadong Town, Huadu District	0.28	0.2	0.455	0.8	0.42
9	Huashan Town, Huadu District	0.28	0.2	0.414	0.8	0.3872
10	Huacheng Street, Huadu District	0.26	0.4	0.405	0.6	0.347
11	Chengjiao Street, Conghua District	0.358	0.2	0.4825	0.8	0.4576
12	Aotou Town, Conghua District	0.266	0.2	0.505	0.8	0.4572
13	Taiping Town, Conghua District	0.2415	0.2	0.4375	0.8	0.3983
14	Liangkou Town, Conghua District	0.32	0.2	0.64	0.8	0.576
15	Lvtian Town, Conghua District	0.48	0.1	0.49	0.9	0.489
16	Wenquan Town, Conghua District	0.37	0.06	0.51	0.94	0.5016
17	Jiangpu Street, Conghua District	0.32	0.8	0.31	0.2	0.318
18	Jiekou Street, Conghua District	0.44	0.65	0.39	0.35	0.4225
19	Shijing Street, Baiyun District	0.285	0.2	0.336	0.8	0.3258
20	Shimen Street, Baiyun District	0.268	0.2	0.268	0.8	0.268
21	Baiyunhu Street, Baiyun District	0.268	0.4	0.285	0.6	0.2782
22	Jiahe Street, Baiyun District	0.2655	0.2	0.271625	0.8	0.2704
23	Junhe Street, Baiyun District	0.25775	0.2	0.272	0.8	0.26915
24	Yongping Street, Baiyun District	0.253	0.2	0.2675	0.8	0.2646
25	Longgui Street, Baiyun District	0.53	0.25	0.455	0.75	0.47375
26	Jianggao Town, Baiyun District	0.3875	0.11	0.3675	0.89	0.3697
27	Renhe Town, Baiyun District	0.37	0.12	0.4245	0.88	0.41796
28	Dayuan Street, Baiyun District	0.25	0.4	0.36	0.6	0.316
29	Taihe Town, Baiyun District	0.24	0.3	0.49	0.7	0.415
30	Zhongluotan Town, Baiyun District	0.57	0.5	0.57	0.5	0.57
31	Yuncheng Street, Baiyun District	0.36	0.5	0.36	0.5	0.36
32	Huangshi Street, Baiyun District	0.37	0.5	0.37	0.5	0.37
33	Helong Street, Baiyun District	0.25	0.5	0.25	0.5	0.25
34	Sanyuanli Street, Baiyun District	0.277	0.2	0.2565	0.8	0.2606
35	Tangjing Street, Baiyun District	0.25675	0.2	0.25675	0.8	0.25675
36	Tongde Street, Baiyun District	0.2675	0.2	0.2515	0.8	0.2547
37	Jinsha Street, Baiyun District	0.2675	0.2	0.333	0.8	0.3199
38	Songzhou Street, Baiyun District	0.2675	0.2	0.4425	0.8	0.4075
39	Jingxi Street, Baiyun District	0.2	0.0625	0.317	0.9375	0.309688
40	Tonghe Street, Baiyun District	0.2	0.67	0.24	0.33	0.2132
41	Jingtai Street, Baiyun District	0.25	0.11	0.3592	0.89	0.347188
42	Longhu Street, Huangpu District	0.2	0.7283	0.5175	0.2717	0.286265
43	Xinlong Town, Huangpu District	0.2	0.9189	0.5645	0.0811	0.229561
44	Jiufo Street, Huangpu District	0.2	1			0.2
45	Lianhe Street, Huangpu District	0.2	0.2417	0.26125	0.7583	0.246446
46	Changling Street, Huangpu District	0.2	0.82	0.445	0.18	0.2441
47	Yuzhu Street, Huangpu District	0.24	0.2	0.25	0.8	0.248
48	Huangpu Street, Huangpu District	0.25	0.2	0.487	0.8	0.4396
49	Changzhou Street, Huangpu District	0.285	0.2	0.404	0.8	0.3802
50	Hongshan Street, Huangpu District	0.25	0.9	0.474375	0.1	0.272438
51	Wenchong Street, Huangpu District	0.25	0.9	0.35	0.1	0.26
52	Dasha Street, Huangpu District	0.25	0.9	0.37375	0.1	0.262375
53	Suidong Street, Huangpu District	0.284	1			0.284
54	Xiagang Street, Huangpu District	0.2775	1			0.2775
55	Nangang Street, Huangpu District	0.35875	1			0.35875
56	Yunpu Street, Huangpu District	0.15	0.714286	0.3	0.285714	0.192857
57	Yonghe Street, Huangpu District	0.1	0.714286	0.39	0.29	0.184529
58	Luogang Street, Huangpu District	0.15	0.5	0.25	0.5	0.2
59	Xiaolou Town, Zengcheng District	0.32	0.2	0.3875	0.8	0.374
60	Paitan Town, Zengcheng District	0.38	0.2	0.44	0.8	0.428
61	Zhengguo Town, Zengcheng District	0.425	0.2	0.416	0.8	0.4178
62	Xintang Town, Zengcheng District	0.152	0.65	0.37	0.35	0.2283
63	Xiancun Town, Zengcheng District	0.41	0.074	0.545	0.926	0.53501
64	Shitan Town, Zengcheng District	0.2525	0.25	0.6713	0.75	0.5666
65	Yongning Street, Zengcheng District			0.3575	1	0.3575
66	Ningxi Street, Zengcheng District			0.455	1	0.455
67	Zhongxin Town, Zengcheng District			0.44	1	0.44
68	Lihu Street, Zengcheng District	0.15	0.888889	0.51	0.111111	0.19
69	Zengjiang Street, Zengcheng District	0.31	0.446809	0.69	0.55	0.518011
70	Zhucun Street, Zengcheng District	0.299	0.23196	0.42025	0.76804	0.392125
71	Licheng Street, Zengcheng District	0.3	0.2	0.63	0.8	0.564
72	Hualong Town, Panyu District	0.4225	0.2222	0.59625	0.7778	0.557643
73	Shilou Town, Panyu District	0.26	0.4267	0.3931	0.5733	0.336306
74	Shiji Town, Panyu District	0.3	0.08	0.405	0.92	0.3966
75	Nancun Town, Panyu District	0.46	0.2758	0.3725	0.7242	0.396633
76	Xinzao Town, Panyu District	0.495	0.12	0.37	0.88	0.385
77	Xiaoguwei Street, Panyu District	0.395	0.1	0.295	0.9	0.305
78	Zhongcun Street, Panyu District			0.285	1	0.285
79	Shatou Street, Panyu District			0.315	1	0.315
80	Shawan Town, Panyu District			0.3975	1	0.3975
81	Donghuan Street, Panyu District	0.4735	0.8	0.3939	0.2	0.45758
82	Shiqiao Street, Panyu District	0.4825	0.8	0.48	0.2	0.482
83	Dalong Street, Panyu District	0.412	0.2	0.34975	0.8	0.3622
84	Qiaonan Street, Panyu District	0.4645	0.2	0.46945	0.8	0.46846
85	Dashi Street, Panyu District	0.307	0.7314	0.5106	0.2686	0.361687
86	Shibi Street, Panyu District	0.2856	0.1	0.2856	0.9	0.2856
87	Luopu Street, Panyu District	0.15	0.6	0.29125	0.4	0.2065
88	Dongyong Town, Nansha District	0.3	0.1737	0.4075	0.8263	0.388827
89	Huangge Town, Nansha District	0.24	0.2273	0.3975	0.7727	0.3617
90	Nansha Street, Nansha District	0.22	0.046	0.36	0.954	0.35356
91	Wanqingsha Town, Nansha District	0.5837	0.1	0.5837	0.9	0.5837
92	Zhujiang Street, Nansha District	0.6072	0.9	0.6072	0.1	0.6072
93	Longxue Street, Nansha District	0.15	0.7	0.4538	0.3	0.24114
94	Lanhe Town, Nansha District	0.4175	0.2179	0.475	0.7821	0.462471
95	Dagang Town, Nansha District	0.25	0.2308	0.5012	0.7692	0.443223
96	Hengli Town, Nansha District	0.295	0.175	0.52	0.825	0.480625
97	Shayuan Street, Nansha District	0.45	0.1175	0.4375	0.8825	0.438969
98	Huazhou Street, Haizhu District			0.319	1	0.319
99	Guanzhou Street, Haizhu District			0.32875	1	0.32875
100	Pazhou Street, Haizhu District	0.15	0.9	0.59	0.1	0.194
101	Chigang Street, Haizhu District	0.48	0.8	0.49	0.2	0.482
102	Nanzhou Street, Haizhu District	0.46	0.8	0.50725	0.2	0.46945
103	Jianghai Street, Haizhu District	0.415	0.8	0.41275	0.2	0.41455
104	Ruibao Street, Haizhu District	0.2	0.0238	0.36	0.9762	0.356192
105	Nanshitou Street, Haizhu District	0.21	0.039	0.555	0.961	0.541545
106	Nanhuaxi Street, Haizhu District	0.74	0.5	0.68	0.5	0.71
107	Haizhuang Street, Haizhu District	0.66	0.5	0.63	0.5	0.645
108	Fengyang Street, Haizhu District	0.3295	0.9	0.3295	0.1	0.3295
109	Changgang Street, Haizhu District	0.2375	0.7945	0.4285	0.2055	0.276751
110	Jiangnanzhong Street, Haizhu District	0.45775	0.9	0.7111	0.1	0.483085
111	Xingang Street, Haizhu District	0.59	0.2	0.525	0.8	0.538
112	Sushe Street, Haizhu District	0.44	0.2	0.4875	0.8	0.478
113	Binjiang Street, Haizhu District	0.45	0.2	0.565	0.8	0.542
114	Longfeng Street, Haizhu District	0.55	0.0769	0.675	0.9231	0.665388
115	Zhuguang Street, Yuexiu District	0.4026	0.9	0.4026	0.1	0.4026
116	Renmin Street, Yuexiu District	0.47125	0.9	0.47125	0.1	0.47125
117	Guangta Street, Yuexiu District	0.39385	0.9	0.39385	0.1	0.39385
118	Huanghuagang Street, Yuexiu District	0.2584	1			0.2584
119	Huale Street, Yuexiu District	0.375	1			0.375
120	Dengfeng Street, Yuexiu District	0.2633	1			0.2633
121	Liurong Street, Yuexiu District	0.265	0.2	0.26	0.8	0.261
122	Liuhua Street, Yuexiu District	0.45	0.6	0.25	0.4	0.37
123	Kuangquan Street, Yuexiu District	0.31	0.2	0.347	0.8	0.3396
124	Baiyun Street, Yuexiu District	0.27	0.7	0.28	0.3	0.273
125	Dongshan Street, Yuexiu District	0.55	0.3	0.72	0.7	0.669
126	Dadong Street, Yuexiu District	0.41725	1			0.41725
127	Nonglin Street, Yuexiu District	0.2605	0.2	0.31	0.8	0.3001
128	Hongqiao Street, Yuexiu District	0.4488	0.8	0.378	0.2	0.43464
129	Beijing Street, Yuexiu District	0.44875	0.8	0.4165	0.2	0.4423
130	Jianshe Street, Yuexiu District	0.505	0.8	0.421	0.2	0.4882
131	Huangcun Street, Tianhe District	0.253	0.0727	0.3625	0.9273	0.354539
132	Zhuji Street, Tianhe District	0.399	0.0468	0.35	0.9532	0.352293
133	Qianjin Street, Tianhe District	0.3745	0.0615	0.4264	0.9385	0.423208
134	Yuancun Street, Tianhe District	0.46	0.2	0.4675	0.8	0.466
135	Tianyuan Street, Tianhe District	0.265	0.2	0.25	0.8	0.253
136	Chebei Street, Tianhe District	0.33625	0.2	0.3385	0.8	0.33805
137	Xiancun Street, Tianhe District	0.2	0.073	0.2325	0.927	0.230128
138	Liede Street, Tianhe District	0.2	0.052	0.25	0.948	0.2474
139	Tianhenan Street, Tianhe District	0.2025	0.9107	0.433	0.0893	0.223084
140	Shipai Street, Tianhe District	0.33625	0.2	0.33175	0.8	0.33265
141	Wushan Street, Tianhe District	0.204	0.2	0.325	0.8	0.3008
142	Tangxia Street, Tianhe District	0.27	0.2	0.3295	0.8	0.3176
143	Shadong Street, Tianhe District	0.31	0.2748	0.3	0.7252	0.302748
144	Shahe Street, Tianhe District	0.43	0.052	0.471	0.948	0.468868
145	Linhe Street, Tianhe District	0.25	0.9819	0.41725	0.0181	0.253027
146	Xintang Street, Tianhe District	0.2	0.0805	0.3237	0.9195	0.313742
147	Longdong Street, Tianhe District	0.3	0.1325	0.3	0.8675	0.3
148	Fenghuang Street, Tianhe District	0.371	0.1333	0.395	0.8667	0.391801
149	Meihuacun Street, Liwan District	0.3425	0.065	0.48	0.935	0.471063
150	Shiweitang Street, Liwan District	0.31	0.5	0.42	0.5	0.365
151	Huadi Street, Liwan District	0.33	0.5	0.63	0.5	0.48
152	Changhua Street, Liwan District	0.5888	0.9	0.5888	0.1	0.5888
153	Duobao Street, Liwan District	0.64281	0.9	0.64281	0.1	0.64281
154	Qiaozhong Street, Liwan District	0.155	0.501	0.3554	0.499	0.255
155	Shamian Street, Liwan District	0.2975	1			0.2975
156	Lingnan Street, Liwan District	0.605	0.2	0.575	0.8	0.581
157	Longjin Street, Liwan District	0.2925	1			0.2925
158	Hualin Street, Liwan District	0.525	0.2	0.545	0.8	0.541
159	Fengyuan Street, Liwan District	0.4848	0.8	0.569	0.2	0.50164
160	Jinhua Street, Liwan District	0.39	0.8	0.4225	0.2	0.3965
161	Caihong Street, Liwan District	0.4875	0.8	0.5158	0.2	0.49316
162	Zhanqian Street, Liwan District	0.4083	1			0.4083
163	Xicun Street, Liwan District	0.5125	1			0.5125
164	Nanyuan Street, Liwan District	0.485	1			0.485
165	Dongnao Street, Liwan District	0.2	0.1	0.3225	0.9	0.31025
166	Zhongnan Street, Liwan District	0.425	0.3	0.39	0.7	0.4005
167	Hailong Street, Liwan District	0.41	0.4	0.46	0.6	0.44
168	Baihedong Street, Liwan District	0.492	0.8	0.554	0.2	0.5044
169	Dongsha Street, Liwan District	0.37	0.8	0.36225	0.2	0.36845
170	Chongkou Street, Liwan District	0.437	0.8	0.479	0.2	0.4454
171	Chajiao Street, Liwan District	0.53	0.3	0.507	0.7	0.5139

The lethal level in urban areas is lower than that in rural areas, which does not mean that the lethal level distribution in urban areas is necessarily lower than that in rural areas. In fact, the main distribution interval of lethal level in urban areas is between 0.2–0.5, while that in rural areas is between 0.2–0.6. The reason for the difference is the influence of the population coefficient. At the same time, there are areas with relatively high lethal levels in the urban areas above 0.6. The main reason is that there are a large number of urban villages in Guangzhou, with poor construction quality and a relatively dense population.

As shown in Figure 6, the lethal level in the urban area is generally lower than the lethal level of each township, the combined result is good, the R² value is 0.6803; the lethal level in rural areas is much higher than the lethal level of each township, and the R² value is 0.6714. The distribution range of the overall lethal level of each township is more inclined to the level of the urban area. The lethal level of each township obtained through the population coefficient and the respective lethal level of the township and rural areas can also better reflect the overall level of each township. In this way, in the process of earthquake lethal risk assessment, the lethal level can also be used as a more accurate parameter to evaluate and calculate.

Figure 6. Fitting results of the overall lethal level of urban areas, rural areas and each town.

Based on the lethality matrix, it is possible to calculate the death risk in the study area under different intensity ranges. As shown in Figure 7, we have calculated the number of possible deaths in the study area under different intensities based on the lethal level and the corresponding mortality data of different intensities. In general, when the study area suffered from different intensities, the overall distribution characteristics of death risk are relatively similar. Generally, they were mainly concentrated in the old urban areas located in the central area of the city, such as Liwan district, Tianhe district, Haizhu district etc. the regional distribution of death risks is relatively fixed.

Figure 7. Grid assessment results of earthquake lethal risk in Guangzhou with different intensity (a: the risk result when the study area suffers from an intensity of VI; b: the risk result when the study area suffers from an intensity of VI; c: the risk result when the study area suffers from an intensity of VI; d: the risk result when the study area suffers from an intensity of VI; e: the risk result when the study area suffers from an intensity of VI; f: the risk result when the study area suffers from an intensity of VI).

However, when the same area is subjected to different intensities, there is a big difference in the number of deaths, generally speaking, when the intensity of the study area is VI (Figure 7a), the highest risk of death of each grid is less than 1 person, which means that the risk of death is low or will not cause casualties; when the intensity of the study area is VII (Figure 7b), highest risk of death of each grid is 1.7 people, which means it may cause death, but it is usually between several people; when the intensity of the study area is VIII (Figure 7c), there is a significant difference in the risk of death, the highest is 14 people, which means that the risk of death is generally between dozens of people; when the intensity of the study area is IX (Figure 7d), the number of areas with a lower risk of death is also between 0 and 2, and should reach around 187 people, which means that the risk of death is between dozens and hundreds of people; when the intensity of the study area is X (Figure 7e), the highest risk of death may reach about 1750 people, which means that the risk of death is between dozens and thousands of people; when the intensity of the study area is XI (Figure 7f), the highest risk of death may reach about 5900 people, and the lowest risk of death is between 0 and dozens of people, which means that the risk of death is huge, based on the results of the possible death risks of different intensities in the study area, as the intensity increases, the risk of death is increases, and the location is relatively obvious, however, the calculation results based on each intensity can only reflects an overall trend in the risk of death, in fact, different regions may suffer from different levels of intensity, how to carry out targeted calculations according to different intensity ranges is the key to carrying out quantitative evaluation of the risk.

Therefore, in order to obtain a more targeted assessment result, based on the PGA of the study area, combined with the formula 1, we calculated the intensity distribution range of each area in the study area, as shown in Figure 8, the distribution characteristics of PGA are more obvious, and the commission is centered at the boundary, the PGA of the southern area is 0.1, including Nansha District, Panyu District, Haizhu District, Liwan District and Tianhe District, which have a relatively high level, while the northern area, including Huangpu District, Baiyun District, Huadu District, Zengcheng District, and Conghua District, has a lower level of 0.05, there is a significant difference between the southern and northern regions. Since this article is mainly aimed at Guangzhou, the distribution map of PGA in the study area is mainly calculated based on the actual value of each township unit, fatalities can be better correlated with local scale PGA (site-specific) analysis since site characteristics are very much dominant when collapse or severe damage to structures is concerned, but because the final risk analysis results in this article are also for the entire city, Therefore, we did not perform local scale PGA analysis, at the same time, for a city, the PGA parameters at the township level are more in line with the actual scale.

Figure 8. PGA distribution results in Guangzhou.

Based on the PGA distribution, combined with Formula 1, the intensity distribution range of the study area is obtained, and on this basis, the mortality rate of each area is obtained based on the lethality matrix. As shown in Figure 9, based on the grid distribution result of the lethal level and intensity of each township, the mortality rate corresponding to each grid is obtained. Under different intensity levels, the mortality rate is different, including 9 levels. Areas with relatively high mortality rates are mainly concentrated in 3 urban areas, namely, Liwan District, Tianhe District, and Haizhu District, as well as some townships in Panyu District and Nansha District. The mortality rate is generally more than 0.000013. The mortality rate in counties is relatively low, generally below 0.000009. In general, the mortality rate in the southern region was higher than that in the northern region.

Figure 9. The grid distribution results of the mortality rate of various townships in Guangzhou.

Based on the mortality rate and population data of each grid, the results of the death risk of an earthquake are obtained. Since the population distribution accuracy is 0.1 km*0.1 km, the data accuracy is relatively high. However, for earthquake disaster risk assessment, such data accuracy has been able to meet the requirements. At the same time, because the lethal level obtained in this paper is based on the field survey results of the survey points at the administrative village level, relative to the population data accuracy, the granularity is relatively coarse. In fact, the lethal level according to the grid of 1 km*1 km can also reflect the distribution characteristics. Therefore, we will also use the population grid data according to the data accuracy of 1 km*1 km.

As shown in Figure 10, it is found that the areas with a higher risk of death are mainly concentrated in urban areas such as Liwan District, Yuexiu District, Haizhu District, Changhua Street in Liwan District, Longfeng Street in Haizhu District, and Dongshan Street in Yuexiu District; the maximum number of deaths in the grid population is more than 2 people. However, the results in Huadu District, Conghua District and Zengcheng District are relatively small compared to the old city, and the average value is basically 0-1. At the same time, there are also a large number of low disaster risk areas in this area, and there may be no casualties after the earthquake.

Figure 10. Grid assessment results of earthquake lethal risk in Guangzhou.

Discussion

Traditional earthquake disaster risk assessment methods are mostly based on the vulnerability level of buildings and are combined with other relevant data to carry out risk assessments (Del Gaudio et al. 2017). Due to the limitation of the selected parameters, most of the disaster risk results are expressed in terms of risk levels, and it is difficult to obtain quantitative assessment results. At the same time, due to the relatively large number of parameters selected by traditional methods, the results are more redundant, especially for the assessment of the risk of casualties, most of the calculations are based on the vulnerability level of the building, and the lack of the ability of different building types to cause casualties under the same damaged state will also increase the error.

The basis of this method is the lethal level of the area, different regions are affected by factors such as buildings and the geographical environment and have different regional lethal levels, which represents mortality rates of different intensities. That is, the lethal level is the same in two different areas; if it is in a different intensity range after the earthquake, then the mortality rate is also different. Therefore, based on the PGA and the lethal level, the possible mortality rate for each grid is determined, and the result is combined with high-precision population grid data to evaluate the number of possible casualties.

This paper selected 431 survey points to conduct a field survey in Guangzhou, and obtained the lethality level of 171 township units based on calculations, integrated with additional data about the location and dimension of buildings collected during independent field surveys, the lethal level of the area is more accurate, and it is in line with the actual situation. The accuracy of the data can provide support for the quantitative assessment of earthquake disaster risk.

The overall quality of buildings in Guangzhou is good, and the earthquake resistance is strong. The areas with low lethal levels are concentrated in the districts and counties in the middle of the city, especially in the newly developed districts and counties that have been built in recent years, such as Baiyun District and Huangpu District. In contrast, the old urban areas of Guangzhou, such as Tianhe, Liwan, and Haizhu District, with poor economic development, have a relatively high lethal level, one of the main reasons is that there are more urban villages in these districts and counties, additionally, the construction is old, the structural measures are incomplete, the construction quality is poor, and the earthquake resistance is low.

The average level in urban areas is 0.3476, and the average level in rural areas is 0.4198. That is, the levels of urban areas and rural areas are actually not much different. Combining the population proportion coefficients in urban and rural areas, the overall lethal level of each township is obtained. The results indicate that the average lethal level of townships is 0.389, and the errors between the urban and rural areas are 0.0414 and 0.0308, respectively, which are relatively small. On the other hand, through the fitting analysis of the lethal level in urban areas and rural areas and townships, it is found that the fitting results coefficient of determination (R²) values are 0.6803 and 0.6714, respectively; these results are also relatively good.

Based on the mortality rate of each grid, it was found that the high mortality rate in Guangzhou was mainly distributed in the southern region, including some townships in Tianhe District, Liwan District, Huangpu District and Baiyun District, Panyu District and Nansha District. The mortality rate is generally between 0.000009–0.00009, while the areas with relatively low mortality rates are concentrated in parts of Huadu District, Conghua District, Zengcheng District and Baiyun District, which is generally between 0.0000004–0.000009. The mortality rate of some towns in the central area of Guangzhou is much higher than that of other regions. The main reason is that the intensity is relatively high, and the other is that the lethal level in the old city areas is relatively high.

In the study area under different intensities, the characteristics of death risk are relatively obvious, and the distribution locations are relatively concentrated, under different intensity situations, there are obvious differences in the quantitative results of death risk, generally speaking, when the intensity is VI, the death risk is relatively small, and it can be regarded as no casualties under normal circumstances; with the increase in intensity, the death risk is increasing, the maximum risk is as follows, degree VII, between a few people; degree VIII, between dozens of people; degree IX, around a hundred people; degree X, around a thousand people; degree XI, around thousands, different intensities have different death risks, but the location distribution of high-risk areas is relatively concentrated, mainly concentrated in the old urban areas located in the central area of the city, such as Liwan district, Tianhe district, Haizhu district etc.

However, based on the intensity distribution of each area obtained based on the PGA, the death risk of earthquake in the study area are significantly different, the risks of different regions are also quite different. First, for Tianhe District, Liwan District and Haizhu District in the center of the city, the result of the risk of death is still relatively high, generally above 1. Second, in Panyu District and Nansha District, where the mortality rate is relatively high, there is a large difference in the risk assessment results; the risk is at a medium level, generally between 0.1 and 1. Third, the risk of Huadu District, Conghua District and Zengcheng District showed obvious differences; the results of most areas is 0, and the risk is at a medium level, generally between 0-0.1.

Conclusions

In this work, we improved the traditional method and proposed a quantitative assessment method based on the level of regional lethal. Compared with traditional evaluation methods, the method transforms the vulnerability level into a lethal level.

The methodology is based on the structural characteristics of building data from the field survey to define the lethal level and based on the PGA to evaluate the seismic capacity, Infill elements are included in the model.

The determination of mortality is mainly based on the lethality matrix constructed by the previous research. Based on the intensity range and lethal level corresponding to each grid in the study area, the mortality rate of each grid is determined. This mortality rate actually refers to the probability of death for the grid to experience different intensities.

Analytical results showed good agreement with the field survey, the distribution of lethal levels in Guangzhou showed obvious regional characteristics, the lethal level of newly developed and constructed districts and counties in central Guangzhou are generally below 0.3. The lethal level of the districts and counties in the center of cities with more developed economies are generally above 0.6. The lethal level in rural areas with an average value above 0.4.

The distribution characteristics of PGA are more obvious, showing obvious north-south differences. With Tianhe District and Huangpu District as the boundary, the southern area is higher than the northern area. The mortality rate of each grid determined by the lethal level and the PGA is significantly different, the areas with higher mortality are mainly concentrated in the central city and parts of the south, while the mortality rate in the northern region is generally lower.

Under different intensities in the study area, the risk of death has shown obvious characteristics. As the intensity increases, the number of possible deaths increases rapidly, ranging from a few to several thousand. At the same time, the higher risk areas are relatively fixed, generally mainly located in the central city, combine high-precision population data, the quantitative assessment results of disaster risk also have obvious directivity. The areas with the highest risk are located in most areas of Tianhe District, Liwan District and Haizhu District. The highest grid death result is more than 1. The lethal level of Baiyun District and Huangpu District is low, but the death risk is also higher, generally between 0.5 and 1. In rural areas in the southern and northern regions of Guangzhou, the risk is relatively low, generally between 0 and 0.5.

The lethal level provides a feasible way to quantitatively assess earthquake risk. The application of the lethal level can usefully support the validation of its reliability in view of further applications to large scale seismic risk analyses, which may reduce the errors caused by qualitative expression of various factors, and more accurate quantitative evaluation results can be obtained.

Disclosure statement

The authors have no conflicts of interest to declare that are relevant to the content of this article.

Data availability statement

The data that support the findings of this study are available from the corresponding author, Chaoxu Xia, upon reasonable request.

Additional information

Funding

This work was jointly supported by the National Nonprofit Fundamental Research Grant of China, Institute of Geology, China Earthquake Administration (Grant No. IGCEA2106 and Grant No. IGCEA2133) and the National Key R&D Program of China (No. 2018YFC1504503 and No. 2018YFC1504403).