Nhaz D 23 02247

Natural Hazards
Rockfall hazard assessment and governing strategy for a cliff feature zone based on
the high-resolution online images and terrain analysis
--Manuscript Draft--
Manuscript Number: NHAZ-D-23-02247
Full Title: Rockfall hazard assessment and governing strategy for a cliff feature zone based on
the high-resolution online images and terrain analysis
Article Type: Manuscript
Keywords: rockfall hazard assessment; governing strategy; landslide; terrain analysis;

Geographic Information System
Abstract: The geological disaster caused by the slope landslide rockfall has some significant
properties, such as suddenness, unpredictability, higher harmfulness. In this paper, a
cliff feature zone in Guizhou Province of China is selected as a case, the topographic
data in the study area are obtained by the field survey, and the slop angle distribution
(SAD) method with the topographic map is adopted to identify the potential rockfall
source zone. The Rockfall Analyst software is applied to simulate the trajectory,
frequency, height, and energy of the slope landslide rockfalls, and the accuracy of the
rockfall model is verified by comparing the simulation results with the actual accident.
Then, the rockfall hazard assessment is conducted for the whole Jinfeng street, and
relevant governing strategies are proposed. The results show that the rockfall source
zone area in the Jinfeng Street validated by using the SAD method with the rock
outcrops in the topographic map is identified as 522,316 m2, which is 0.8% of the total
area. The building area in the hazard zone are 81,105 m2, and the percentage of
building area in the high-hazard and ultra-high-hazard zone reaches 25.6%. To reduce
the hazards caused by the slope landslide rockfalls, according to the Code for Design
of Rock Retaining Wall Engineering in Geological Hazards (T/CAGHP060 -2019), a
stone retaining wall with the height of 5 m is built, and its stability is verified. The results
indicate that after adding the stone retaining wall, the rockfall hazard zone area
decreases to 2,580,371 m2, there is no ultra-high hazard zone, and the building area in
the rockfall hazard zone reduces to 50,432 m2. After adding the stone retaining wall,
for the residents who are still in the high hazard zone, a unified relocation strategy is
recommended to ensure the safety of residents’ lives and property
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Manuscript in DOC format Click here to view linked References
1 Rockfall hazard assessment and governing strategy for a cliff feature zone based
2 on the high-resolution online images and terrain analysis
3 Rongguo Zhao* 1,2, Jinchen Yang1,3, Suhua Zhou4, Minghua Huang4, Minghui Yang5
4 1. Key Laboratory of Dynamics and Reliability of Engineering Structures of College of Hunan Province,
5 Xiangtan University, Xiangtan 411105, China;
6 2. School of Mechanical Engineering and Mechanics, Xiangtan University, Xiangtan 411105, China;
7 3. College of Civil Engineering, Xiangtan University, Xiangtan 411105, China;
8 4. College of Civil Engineering, Hunan University, Changsha 410082, China;
9 5. School of Architecture and Civil Engineering, Xiamen University, Xiamen 361005, China.
10 Corresponding authors: zhaorongguo@xtu.edu.cn
11
12 Abstract
13 The geological disaster caused by the slope landslide rockfall has some significant properties, such as
14 suddenness, unpredictability, higher harmfulness. In this paper, a cliff feature zone in Guizhou Province of
15 China is selected as a case, the topographic data in the study area are obtained by the field survey, and the
16 slop angle distribution (SAD) method with the topographic map is adopted to identify the potential rockfall
17 source zone. The Rockfall Analyst software is applied to simulate the trajectory, frequency, height, and
18 energy of the slope landslide rockfalls, and the accuracy of the rockfall model is verified by comparing the
19 simulation results with the actual accident. Then, the rockfall hazard assessment is conducted for the whole
20 Jinfeng street, and relevant governing strategies are proposed. The results show that the rockfall source
21 zone area in the Jinfeng Street validated by using the SAD method with the rock outcrops in the
22 topographic map is identified as 522,316 m2, which is 0.8% of the total area. The building area in the
1 hazard zone are 81,105 m2, and the percentage of building area in the high-hazard and ultra-high-hazard
2 zone reaches 25.6%. To reduce the hazards caused by the slope landslide rockfalls, according to the Code
3 for Design of Rock Retaining Wall Engineering in Geological Hazards (T/CAGHP060 -2019), a stone
4 retaining wall with the height of 5 m is built, and its stability is verified. The results indicate that after
5 adding the stone retaining wall, the rockfall hazard zone area decreases to 2,580,371 m2, there is no
6 ultra-high hazard zone, and the building area in the rockfall hazard zone reduces to 50,432 m2. After adding
7 the stone retaining wall, for the residents who are still in the high hazard zone, a unified relocation strategy
8 is recommended to ensure the safety of residents’ lives and property.
9 Key words
10 rockfall hazard assessment; governing strategy; landslide; terrain analysis; Geographic Information System
11
12 1. Introduction
13 Rockfall means the collapse of rocks on the surface of steep cliffs and slopes under the gravity force
14 [1-4]. Usually, the rockfalls are sudden, unpredictable, fast, high energy and hazardous, and occur
15 frequently around the world [5-7]. Nowadays, many scholars have focused their eyes on the geological
16 hazard caused by the slope landslide rockfalls. Therefore, to reduce the hazards of rockfalls, how to identify
17 the potential rockfalls quickly and accurately, and how to evaluate the impact of avalanche rockfall hazards,
18 have become the continuous concerned questions in the academic community.
19 A rockfall hazard assessment requires the identification of potential rockfall sources [8-10]. Since the
20 potential rockfalls are basically distributed on the cliffs or steep slopes in the mountainous areas, the
21 traditional manual investigation methods are greatly limited. Then, some instruments and techniques, such
22 as the Unmanned Aerial Vehicle (UAV) photography and the three-dimension (3D) scanning, have been
1 utilizing for the rockfall source identification. However, these instruments and techniques have some strict
2 requirements for the application environment, which limits their utilization in the complex mountainous
3 terrain [11-13]. Subsequently, the methods of determining the source of landslide rockfalls by using the
4 aerial photographs, the remote sensing images, and the high-precision Digital Elevation Models (DEMs),
5 which have the characteristics of high precision, high accuracy, and wide applicability, are widely used in
6 the investigation of the area of the potential landslide rockfall source [14-16]. Based on the statistical data
7 on the location of historical avalanche rockfalls, Guzzetti [17] suggested that the cliffs and slopes became
8 the potential sources of the avalanche rockfall when the critical slope angle was greater than 60°. Frattini [6]
9 constructed a 3D rockfall model and mapped the rockfall sources by using the Guzzetti’s model, and
10 empirically identified the potential avalanche rockfall sources when the critical slope angle is greater than
11 37°. Such studies are based on the statistical analysis of the historical data in the specific areas combined
12 with the theoretical assumptions, and some further research works are needed to determine the critical slope
13 angles in the other areas. Thereafter, based on the Slop Angle Distribution (SAD), a method of dividing the
14 homogeneous morphometric areas by using the geomorphic analysis and then determining the critical slope
15 angle of the avalanche rockfalls is proposed, which has an excellent applicability in the area with different
16 geological patterns [14].
17 According to the parameters of rockfalls, such as the trajectory and energy, the rockfall hazard
18 assessment is conducted on estimating the extent of damage [18-19]. The earliest method to assess the
19 rockfall hazard was to analyse the impact factors of rockfalls and established a corresponding evaluation
20 system, but this method only gave a qualitative evaluation of rockfall hazard, which brought some highly
21 subjective factors [20-21]. With the development of computer and calculation software, pursuing the
22 method based on the physical-mechanical models to quantitatively simulate the trajectories of the
23 avalanche rockfalls and assess the rockfall hazard have gradually become a mainstream [22]. Among them,
1 the commonly used two-dimension (2D) simulation software is the Rockfall, while the popularly adopted
2 3D ones are the Rockyfor3D, the Rockfall Analyst, etc. The Rockfall software, which is based on the
3 statistics, is a two-dimensional simulation software that allows almost all of its inputs to be varied
4 randomly [23-24]. However, the profile of rockfall’s motion is artificially specified, so there is a certain
5 degree of human control. Simultaneously, because it is two-dimensional, it cannot consider the lateral
6 displacement of the collapse fallout. The Rockyfor3D software has been widely used to evaluate the
7 rockfall hazard in the forested areas [25-26], which takes the shapes and sizes of individual rock into
8 account in the simulation, and the more accurate the rockfall parameters are, the more accurate the
9 simulated results will be. Yet, it is difficult to obtain the accurate data of the rockfalls, so its accuracy will
10 be affected when the question with a wide range of rockfalls is faced. The Rockfall Analyst software adopts
11 the centralized quality method to simulate the trajectory of rockfall [27-28], which, as an extension of
12 ArcGIS, presents the simulated trajectory, bouncing height, velocity and other distributions of rockfall
13 simulation in the ArcGIS platform by using the vector and raster data, and it has a wide range of
14 applications in the prediction and assessment of the rockfall hazards. Comparing with the other software,
15 the Rockfall Analyst software does not need to consider the size and shape of the rockfall either, nor does it
16 need to consider the profile direction of the rockfall movement. Due to the benefits that there are few
17 human influence factors, it is more suitable for the simulation of the avalanche rockfalls that occurs
18 suddenly and without warning.
19 In this paper, the background of May 8, 2022, Baiyan Village, Chengguan Town, Zijin County, Bijie
20 City, Guizhou Province is taken as the case of the rockfall event. Firstly, the topographic data for the
21 Baiyan Village is obtained via the field survey. Then, according to the topographic data, the potential
22 sources of rockfalls are predicted by using the method of slop angle distribution (SAD) combined with the
23 topographic maps. Following that, according to the potential sources of rockfalls and the high-precision
1 DEM, the avalanche rockfalls are simulated by using the Rockfall Analyst software, the trajectory,
2 frequency, bouncing height and other kinetic characteristics of rockfalls are calculated, and then the
3 rockfall hazard is evaluated by assigning various weights to the frequency, bouncing height, and energy of
4 rockfalls. Finally, to validate the accuracy of the estimated model, the simulation results of the Baiyan
5 Village rockfall event are compared with the actual situation, and then the rockfall hazard in the Jinfeng
6 Street is simulated and evaluated, and the corresponding reasonable strategies for preventing and
7 controlling the rockfall hazard are proposed.
8 2. Geographical location and geological environment
9 2.1 Geographical location
10 Guizhou Province is in the southwestern China, with steep terrain and mountainous areas accounting
11 for 90% of the province's area. Every year, Guizhou is affected by a variety of geologic hazards. According
12 to the Bureau of Natural Resources of Guizhou Province, there were a total of 28,221 potential geologic
13 hazards with various types as of the end of 2021, including 4,812 landslides, 3,091 avalanches and rockfalls,
14 114 mudslides, and 1,377 unstable slopes. At 12:00 a.m. on May 8, 2022, a mountain collapse occurred on
15 the northwest side of the Baiyan Village, Jinfeng Street, Chengguan Town, Zhijin County, Bijie City,
16 Guizhou Province. The map and geographical location of the Jinfeng Street are shown in Fig. 1. The volume of
17 debris from the avalanche was about 25,000 m3, and the rockfalls extended as far as about 950 m, burying
18 three farmers working in the surrounding area and destroying more than 150 m of the local road and several
19 houses. The avalanche occurred on a steep slope with an elevation of about 1,350 m and a gradient of 64.3°.
20 The collapsed rock mass was mainly sandstone, chert, and shale. No rainfall, earthquakes, engineering
21 excavations, or blasting activities were recorded or reported in the vicinity prior to the avalanche. Therefore,
22 it is initially speculated that the avalanche may be the collapse of a broken rock body caused by long-term
1 weathering. In addition, according to the Zijin County Emergency Response Department, there is still a
2 large possibility of collapse destabilization in the mountains around the avalanche site, and the potential
3 volume of rockfall from the avalanche is about 10,000 m3. According to the relevant information from the
4 local government of Zhijin County, the area where the avalanche occurred was identified as a
5 collapse-prone site as early as ten years ago, but its potential area of influence was not delineated.
6 The Google Earth history images of the Baiyan Village in January 2014, September 2014, December
7 2019, and March 2022 were individually shown in Fig. 2(a), 2(b), 2(c) and 2(d). In can be found from Fig.
8 2(a) that in January 2014, the whole mountainous area in the Baiyan Village was covered by the
9 surrounding lush vegetation. However, in September 2014, some exposed rocks began to appear on the
10 slopes in the Baiyan Village, as shown by the yellow block diagram in Fig. 2(b). As time passed, in
11 December 2019, the exposed rocks on the slops in the Baiyan Village developed into the collapsed rocks,
12 and some new exposed rocks appeared. The collapsed rocks and new exposed rocks are shown by the red
13 block diagrams and the yellow one in Fig. 2(c), respectively. In March 2022, the rockfall area in the Baiyan
14 Village was recorded by the Google Earth, and the avalanche rockfall and the exposed rocks are
15 individually shown by the red block diagram and the yellow one in Fig. 2(d).
1
2 Fig. 1 Map and geographical location: (a-c) Geographical location of study area; (d) Avalanche location in the Jinfeng Street.
3 (a) (b)
4 (c) (d)
5 Fig. 2 Google Earth history images of Baiyan Village: (a) Image in January 2014; (b) Image in September 2014; (c) Image in
6 December 2019; (d) Image in March 2022.

1 2.2 Geological environment
2 The Jinfeng Street, at which the avalanche occurred, is dominated by the highland mountainous
3 terrain, and the elevation of the study area ranges from 1,341 m to 2,129 m as shown in Fig. 1(d), with a
4 mean slope of 20.3°, while the steepest area reaches 80°, as shown in Fig. 2(a). In the study area, there are
5 two obvious "northeast-southwest" trending steep slope zones, with a lot of exposed weathered rock bodies
6 distributed on the high steeps or even near-erect slopes, which have a high hazard of collapse. As shown in
7 Fig. 3(b), the main stratigraphy of the study area includes the following three types. The first type is the
8 Upper Permian (P↓2-3): intermediate chert and flint limestone with sandstone shale interbedded with coal
9 at the bottom. The second one is the Shallow Marine Lower Triassic (T↓1↑MS): grayish-green shale at the
10 bottom of stone. The third one is the Carboniferous (CP): dolomite and chert at the bottom of stone.
11
12 Fig. 3 Slope map and geological map of Jinfeng Street: (a) Slope map; (b) Geological map.
13 3. Methodology for identification and assessment
14 The assessment of the impact area of the rockfall hazard is divided into two main steps, as shown in
15 Fig. 4. Firstly, the rockfall source area needs to be identified. Secondly, after the avalanche occurs, the
16 trajectory of the potential destructive point needs to be analysed.
17 In this paper, based on the terrain slope distribution theory proposed by Loye [14], the areas with the
1 same characteristics of lithology, historical tectonic morphology are divided into homogeneous
2 morphometric areas (HMAs), and then the Gaussian distribution is applied to the terrain morphometric
3 elements in the HMAs to find out the critical value of the slope angle. Combining with the vegetation
4 distribution in the HMAs, the exposed rocks are determined, and the potential sources of avalanche
5 rockfalls are obtained. According to the potential rockfall sources and the site investigation, the Rockfall
6 Analyst software is used to simulate the rockfall, and the trajectory, frequency, speed, height, and energy
7 maps of rockfall are derived. To predict the rockfall hazard zone, the different weights are assigned to the
8 parameters of frequency, height, and energy. The technical flow chart used in this paper is shown in Fig. 4.
10 Fig. 4 Technique flow chart for identifying rockfall sources and estimating rockfall hazard.
11 3.1 Identification of potential rockfall sources
12 Different lithologies, tectonic patterns, and stratigraphic geologies have some effects on rockfall.
13 Therefore, areas with the same lithology, historical tectonic pattern, and geology are classified as the
14 homogeneous morphometric areas [29]. By using ArcGIS, the lithology, historical tectonic morphology,
15 and geology of the study area are superimposed to delineate the homogeneous morphometric areas, and
16 then the SAD method is used to identify the potential avalanche rockfalls in the homogeneous
17 morphometric areas. According to the SAD characteristics, the homogeneous morphometric areas is
18 decomposed into the Gaussian distributions with the specific terrain patterns, and the distributions of the
1 cliffs, steep slopes, foot slopes, and plains in the topographical feature map are shown in Fig. 5.
3 Fig. 5 The distributions of cliffs, steep slopes, foot slopes, and plains in the topographical feature map of Jinfeng Street.
4 Since the effective area of each raster in the DEM depends on the size of the slope angle, the actual
5 terrain area is weighted according to the slope angle of the raster cells [14], and the expression for the
6 frequency of slope angle distribution is written as
Ah
7 W  (1)
AHMA cos β
8 where, Ah is the total area of the DEM with a slope angle of β , and AHMA is the sum of the areas of
9 the HMA.
10 Using the slope distribution method, the terrain is divided into several major morphological units, and
11 then the Gaussian distribution is applied to these units. Fig. 6 shows the Gaussian distributions of terrain
12 slope for various geological patterns. The terrain studied in this paper can be divided into the following
13 four geological patterns.
14 (1) Plains: the low-slope areas characterized by the rivers and glacial deposits.
15 (2) Foot slopes: an area of gentle slope angle in the lower portion of a slope characterized by the
16 alluvial fans, debris flows, and landslide deposits.

1 (3) Steep slopes: an area of steeper slopes characterized by the slope deposits and plant-covered rock
2 outcrops.
3 (4) Cliff: an area of steep slopes characterized by the exposed rock outcrops.
Frequency
0.06 Slope Angle Distribution(SAD)
Sum of SAD R2=0.984
Plains
0.05
Foot slopes
B Steep slopes
0.04 Cliff
0.03
0.02
0.01
A
0
0 10 20 30 40 50 60 70 80 90
Slope (°)
4
5 Fig. 6 Gaussian distributions of slope angle for various geological patterns.
6 In the above morphological elements, the steep slope element and cliff one may contain the rockfall
7 sources. Therefore, for each homogeneous zone's slope angle, the Gaussian distribution is used separately
8 to determine the critical slope angle. The specific method for calculating the critical slope is as follows.
9 (1) The critical slope angle is defined as the point of intersection between the two steepest
10 morphological elements, as shown the point A in Fig. 6, which is the surface of lightly vegetated cliff
11 element and steep slope one, and is recognized as a potential source of avalanche rockfall when the slope
12 angle is greater than the critical slope angle, regardless of lithology and land cover.
13 (2) In some special cases, there may be no cliff element. At this time, the steep slope elements can be
14 considered as the steepest geological morphological elements with the property of the Gaussian
15 distribution.
16 (3) In the steep regions, based on the slope angle analysis, two cliff form elements with the property of
17 the Gaussian distribution are divided. In this case, the cliff element with a lower slope is selected as the
18 research object, and then the critical slope angle is calculated according to (1).
1 (4) If the rockfalls occur in the cliff and steep slope elements, then considering the entire steep slope
2 as a rockfall source will overestimate the region of the potential rockfall source. In this case, only if the
3 angle is higher than the mean of steep slope, as shown the point B in Fig. 6, the rockfall source is identified
4 as a potential rockfall source.
5 Combining with the terrain slope distribution method and the high-precision DEM maps, the various
6 HMAs are investigated by using the Geographic Information System (GIS), and the potential avalanche
7 rockfall resources are identified.
8 3.2 Assessment of rockfall hazards
9 The rockfall simulation is conducted by using the Rockfall Analyst software, which is an extensive
10 software of GIS. The rockfall simulation is divided into two main steps.
11 (1) Constructing a 3D model for the avalanche rockfall;
12 (2) Constructing a grid model for the spatial distribution of the collapse rockfalls.
13 For the Rockfall Analyst software, the centralized mass method is adopted, which means that when
14 using this software to simulate the avalanche of rockfalls, the sizes and shapes of the rockfalls do not need
15 to be considered, but only needs to input the parameters, such as the area, number, and the initial velocity
16 of the avalanche rockfall. The interaction between the avalanche rockfall and the slope is controlled by
17 three coefficients known as the normal recovery coefficient, tangential recovery coefficient, and friction
18 coefficient, which remain constants during the simulation process.
19 The expression for calculating the rockfall trajectory can be written as
   
 X 0  Vx 0 t   X 0  Vx 0   0 
       
20 L Y0  Vy 0t    Y0   Vy 0  t   0  (2)
   Z  V   1 
 Z 0  Vz 0 t  gt 2   0   z 0   gt 2 
1
 2   2 
21 where, X 0 , Y0 and Z 0 are individually the initial positions of the rockfall in x , y and z direction,
1 t is the time, g is the acceleration of gravity whose value is taken as 9.8 m/s2, Vx 0 , Vy 0 , Vz 0 are the
2 initial velocities in x , y and z direction, respectively. The formula for calculating the rockfall velocity
3 v v can be expressed as
 Vx 0  Vx 0   0 
     
4 v v   Vy 0   Vy 0    0  (3)
V  gt  V    gt 
 z0   z0   
5 When the rockfall undergoes collision and bounces, its inclination, trend, and normal velocity
6 components VDip  , and VN are affected by the tangential recovery coefficient and normal recovery
 , VTrend
7 coefficient, respectively. The corresponding expressions for calculating the inclination, trend, and normal
8 velocity components of rockfall are
9 VDip   RTVTrend , VN  RNVN

  RTVDip , VTrend (4)
10 where, VDip , VTrend , and VN are individually the inclination, trend, and normal velocity components of
11 the elastic collision velocity without energy loss, RT and RN are individually the tangential recovery
12 coefficient and normal recovery coefficient, whose values are in the range of [0, 1].
13 By using the Rockfall Analyst software, not only the trajectory and velocity of avalanche rockfall can
14 be simulated, but also the physical parameters, such as the frequency, height, and energy can be calculated.
15 One this basis, different weights are assigned to the physical parameters of frequency, height, and energy,
16 then these physical parameters are weighted and combined to evaluate the hazard of avalanche rockfall.
17 4. Results and discussion
18 4.1 Results of potential rockfall
19 Since the studied site is in Jinfeng Street, Zhijin County, Bijie City, Guizhou Province, the study area
20 is relatively small and the lithology within the area is similar. Therefore, the region can be regarded as a
21 homogeneous morphology area. To minimize errors, based on the combination of topographical maps, the
22 calculated result is compared with the historical rockfall events to ensure the prediction accuracy of
1 avalanche rockfall source. The slope angle thresholds for potential rockfall are shown in Table 1.
2 Table 1 The slope angel thresholds for potential rockfall in the Jinfeng Street.
Critical slope angle
HMA Area of potential rockfall (m2)
Cliff A Steep slope B
1 38.3° 22.1° 522316.3
3 It can be seen from Table 1 that the critical slope angle of potential rockfall in Jinfeng Street is 38.3°.
4 By using GIS, the slope angles in the potential avalanche rockfall zone are mapped, and the regions with
5 the slope angles greater than this threshold are identified. Fig. 7 show the potential rockfall sources in the
6 Jinfeng Street. It can be seen form the Fig. 7 that the historical disaster sites are all in the identified
7 potential rockfall zones, and the current rockfall in the Baiyan Village is also in the identified potential
8 rockfall zones, which verifies the accuracy of using GIS to identify the potential rockfall sources.
10 Fig. 7 The potential rockfall sources in the Jinfeng Street.
11 4.2 Results of rockfall hazard assessment
12 The simulation and analysis for the avalanche rockfall hazard in the Jinfeng Street, Zijin County, Bijie
13 City, Guizhou Province is conducted by using the Rockfall Analyst software. The identified potential
14 rockfall source information is imported into GIS, and then the corresponding DEM information is inputted.
15 DEM, which reflects the real terrain, is one of the keys to influence the accuracy of simulation. The higher
16 the accuracy of the DEM, the more it can restore the real collapse of rockfall. The commonly used DMEs
1 are 30 m and 12.5 m. The DEM used in this study is 12.5 m. After all the data are imported into GIS, the
2 parameters in the simulation are started to be set. The parameters in the simulation are divided into two
3 parts. The first part of simulation parameters are the parameters of rockfall, in which the mainly parameters
4 are the quality and quantity of rockfall. It is found that the change of rockfall mass has a greater effect on
5 the simulation of the energy results of rockfall, while has a small effect on the simulation of the trajectory
6 of rockfall. To make the simulation results match with the actual situation, the mass of the rockfall is
7 determined according to the field measurement for rockfalls. The measurement masses of the rockfalls
8 mostly range from 200 kg to 400 kg. Therefore, the mean mass, which is 300 kg, is adopted as the mass of
9 rockfall. If the number of rockfall simulations is too small, then the rockfall coverage will not be correctly
10 reflected. While if the number of rockfall simulations is too large, then the simulation time will increase
11 significantly and many rockfall simulation trajectories will overlap. Since the potential rockfalls in the
12 studied area are of 0.5 m apart, setting the number of simulated rockfalls to 1 will correctly reflect the
13 coverage region of rockfall. The second part of the simulation parameters are the parameters of slope
14 property, such as the normal recovery coefficient, tangential recovery coefficient, friction coefficient. To
15 obtain the simulation results that are closely resemble the actual rockfall, the slope characteristics of the
16 entire studied area are investigated by using the aerial photographs, then according to Table 2, the
17 determined material property parameters for the specific slope in the studied area are inputted to simulate
18 the avalanche rockfall hazard.
19 Table 2 Material property parameters of the slope surfaces in the studied area.
Characteristics Normal restitution (RN) Tangential restitution (RT) Friction angle (°)
Slopes covered by grasses and shrubs 0.30 0.80 25
Soft and strongly weathered rocks 0.35 0.85 20
Hard rock and masonry 0.40 0.90 15
Soft soil slope 0.30 0.80 30
Water 0.00 0.00 80
20 The landslide map of the Baiyan Village is shown in Fig. 8. Comparing with the actual rockfall data in
21 the Baiyan Village rockfall, the accuracy of the simulation results is verified. The simulated trajectory of
1 rockfall in the Baiyan village is shown in Fig. 8, and the trajectory of A-A’ and the velocity profile in the
2 collapse rockfall simulation are shown in Fig. 9. It can be seen from Fig. 8 that comparing the simulated
3 results with the actual data, the trajectory and range of the collapse rockfall are roughly fitted with the
4 actual ones, and the rockfall accumulation region is matched with the actual shooting situation.
5 Simultaneously, it can be found from Fig. 10 that the rockfall distance in Baiyan Village is nearly 450 m,
6 and the maximum velocity of the avalanche rockfall reaches 52 km/h. These simulated results are
7 consistent with the actual measured data, which verifies the reliability of the simulation model.
8 (a) (b)
9 Fig. 8 Landslide map of Baiyan Village. (a) UAV view of the affected area; (b) Google Earth image of the area.
10 (a) (b)
11 Fig. 9 Rockfall simulation in Baiyan village: (a) rockfall accident simulation; (b) simulation with rock retaining wall.
1
2 Fig. 10 Trajectory and velocity profile of the simulated avalanche rockfall in Baiyan Village.
3 After the accuracy of the simulation model is verified, the parameters of trajectory, velocity, energy,
4 height, and frequency for the avalanche rockfalls in the Jinfeng Street are identified by using the Jinfeng
5 Street’s DEM map with precision of 12.5 m. The simulated results of the avalanche rockfalls in the Jinfeng
6 Street are shown in Fig. 11.
7 (a) (b) (c)
8 (d) (e) (f)
9 Fig. 11 The simulated results of the avalanche rockfalls in the Jinfeng Street: (a) rockfall hazard zone; (b) rockfall trajectory;
10 (c) rockfall frequency; (d) rockfall energy; (e) rockfall velocity; (f) rockfall height.
1 The various weights are individually assigned to the simulated parameters of velocity, frequency,
2 height, and energy of the collapse rockfalls, and then the sum for these parameters with various weights are
3 conducted. Then, to realize the rockfall hazard assessment, the hazard levels are graded by using the natural
4 breakpoint method. The formula for the rockfall hazard assessment is written as
5 H (i )  Wf  Pf (i )  Wh  Ph (i )  We  Pe (i ) (5)
6 where, H (i ) is the hazard degree of the ith grid, Pf (i ) , Ph (i ) , Pe (i ) are individually the frequency,
7 height, and energy of the rockfall in the ith grid, and Wf , Wh , We are the weights for the parameters of
8 frequency, height and energy, respectively. In this paper, the weight coefficients for the parameters of
9 frequency, height and energy are taken as 0.5, 0.2, and 0.3, respectively.
10 The rockfall hazard zone identified by using the simulation method is shown in Fig. 11(a). It can be
11 seen from Fig. 11(a) that the rockfall hazard zone can be divided into four level regions, which are called as
12 the low hazard region, the medium hazard region, the high hazard region, and the ultra-high hazard region.
13 According to the experts’ advisements, for the ultra-high hazard region, and the high hazard region as well,
14 the residents in this area must be relocated because the rockfall hazard may be seriously affect their safety
15 of lives and properties. While for the medium hazard region, or the low hazard region, so long as
16 appropriate governances are conducted for these regions, the residents can still live in these areas. Within
17 the overall studied area, only the general situation of the landslide rockfall hazard zone can be observed,
18 whereas, the hazard level of each village cannot be directly observed. Therefore, the rockfall zone in the
19 Jinfeng Street is divided into four regions. Thus, the affected situations for every village can be detailly
20 observed from each of the rockfall hazard regions.

1 (a) (b)
2 (c) (d)
3 Fig. 12 The simulated results of the rockfall hazard area in detail: (a-d) corresponding to A-D Region in Fig. 11.
4 The simulated results of the rockfall hazard area in detail, which are individually corresponding to A,
5 B, C, D regions, are shown in Fig. 11. It can be seen from Fig. 11 that in the whole Jinfeng Street, there are
6 still a lot of houses in the high hazard and ultra-high hazard regions, which has a considerable impact on
7 the local residents’ safety of lives and properties. How to reduce the impact of rockfall hazards, this is a
8 question facing to the rockfall hazard governance in the Jinfeng Street. According to the Code for Design of
9 Rock Retaining Wall Engineering in Geological Hazards (T/CAGHP060-2019), under the steep cliffs and
10 steep slopes, a rock retaining wall, whose height is not large than 5 m, can be used for the protection of
11 hazardous rock bodies with a certain width of gentle slopes. Therefore, at a location where the rockfalls
12 maybe threaten the houses, a 5-meter-high rock retaining wall is built to protect the houses from rockfalls.
13 The rock retaining wall with a depth of 1 m, is made of C35 concrete, and no slope is designed for the rock
14 retaining wall. According to the Code, the checking calculations of the shear resistance, impact resistance,
15 foundation bearing capacity, and the stabilities of overturning resistance and sliding resistance as well, are
1 necessary to be done. In this study, the property of impact resistance for the rock retaining wall is mainly
2 investigated. Therefore, the limit situation of the rockfalls must to be considered. According to the
3 simulation results for the rockfalls, before colliding with the rock retaining wall, the fastest rockfall with
4 velocity of 9.5 m/s, is selected as the studied object. In this case, if it is verified to be safe, then the rock
5 retaining wall is validated to be safe. The impact force of rockfall is calculated as the following expression,
6 that is
   
7 p  2 Z  2 tan 4  45o    1 (6)
  2 
G 1
8 Z  vi  (7)
2 g S  
2 tan 4  45o    1
 2
9 where, p is the equivalent impact load of rockfall,  is the weight per unit volume of the buffer layer
10 weight , Z is the impact plunging depth,  is the slope angle of the rock retaining wall, g is the
11 acceleration of gravity, and G is the weight of rockfall.
12 After adding the 5-meter-high rock retaining wall, the diagram of rockfall hazard zone is shown in Fig.
13 13. The numerical calculation results are obtained by using the simulation model. The impact plunging
14 depth is calculated as 0.42 m, and the equivalent impact load is calculated as 38.5 kN/m. The compressive
15 strength of rockfall on the rock retaining wall is derived as 96.4 kPa, which is lower than the compressive
16 strength of C35 concrete, so the rock retaining wall will not be damaged.
17
18 Fig. 13 The diagram of rockfall hazard zone in the Jinfeng Street with the rock retaining wall.
1 The cumulative distribution of rockfall hazard zones in the absence of a rock retaining wall is shown
2 in Fig. 14. It can be found from Fig. 14 that in the case without protective measures, the houses in the
3 central part of Region A are all in the low to medium hazard zone, while the house area in the high and
4 ultra-high hazard zones in the southwestern part of Region A account for 23.4% of the whole house area,
5 and the rest of the houses are in the low to medium hazard zone. In the Region B, although the proportion
6 of the high-hazard and ultra-high-hazard areas to the total area reaches 38.6%, and that of the high-hazard
7 and ultra-high-hazard areas to the total hazard zone area reaches 39.9%, the housing area to the whole area
8 in Region B only accounts for 1.6%, which is very small. In the Region C, the proportion of the high and
9 ultra-high hazard area is 32.2%, and that of the house area in the high and ultra-high hazard zones to the
10 total house area in the whole hazard zones is 9.6%. In the Region D, the proportion of the high and
11 ultra-high hazard areas is 4.2%, and that of the house area in the high and ultra-high hazard areas to the
12 total house area in the whole hazard area accounts for 20.6%.
13 The cumulative distribution of the rockfall hazard zone in the presence of a rock retaining wall is
14 shown in Fig. 15. It can be found from Fig. 15 that the total area of the hazard zone reduces to 2,580,371
15 m2 with the additional 5-meter-high rock retaining wall, which is 39.3% lower than that without a rock
16 retaining wall, and the area of the high and ultra-high hazard zones reduces to 966,397 m2, which is 2.9%
17 lower than that without a rock retaining wall. Simultaneously, after adding a 5-meter-high rock retaining
18 wall, the total house area in the hazard zone reduces from 86,556 m2 to 50,432 m2, which is 41.7% to the
19 original hazard zone area, and the proportion of the house area in the high and ultra-high hazard areas
20 reduces from 25% to 17.8%. The above analysis indicates that the measure of adding a 5-meter-high rock
21 retaining wall can reduce the rockfall hazard to a certain degree, but it cannot eliminate the rockfall hazard
22 completely. Therefore, after adding the stone retaining wall, for the residents who are still in the high
23 hazard zone, a unified relocation strategy is recommended to ensure the safety of residents’ lives and
1 property.
Low hazard
Low hazard Medium hazard
Medium hazard High hazard
5×106 High hazard 9×104 Ultra-high hazard
Ultra-high hazard
8×104
4×106
House area in hazard（m2）

7×104
6×104
Hazard area（ m2）
3×106
5×104
4×104
2×106
3×104
1×106 2×104
1×104
0 0
A B C D Total A B C D Total
Hazard zone Hazard zone
2 (a) (b)
3 Fig. 14 The cumulative distribution of rockfall hazard zones in the absence of a rock retaining wall: (a) The cumulative
4 distribution of hazard area; (b) The cumulative distribution of house area in the rockfall hazard zones.
6×104 Low hazard

3×106 Low hazard Medium hazard
High hazard 5×104
House area in hazard（ m2）
4×104
2×106
Hazard area (m2)
3×104
1×106 2×104
1×104
0 0
Hazard zone Hazard zone
5 (c) (d)
6 Fig. 15 The Cumulative distribution of rockfall hazard zones in the presence of a rock retaining wall: (a) The cumulative
8 4.3 Discussions
9 There are many methods for identifying the potential collapse rockfall sources, such as the traditional
10 manual measurement method, the UAV photogrammetric technique, and the three-dimensional scanning
11 technique. Whereas, when these methods are selected to identify the rockfall sources, there are inevitably a
12 certain degree of limitations. In this paper, to identify the potential rockfall zones, the slope angle
13 distribution method is adopted. Firstly, each homogeneous morphology zone is determined. Then, the slope
14 angle distribution of each homogeneous morphology zone is obtained by using the Gaussian distribution,
1 and the critical slope angles of the potential rockfall sources are calculated. The rockfall sources in the
2 studied area are identified by using the slope angles and rock outcrops obtained from the aerial
3 topographical maps, and the accuracy of the simulated results is verified by comparing with the historical
4 information of the landslide rockfall hazards in the Jinfeng Street. In the case without taking the influence
5 of the rock mass structure into account, the topographic slope distribution method can be utilized to identify
6 the potential rockfall resources on a regional scale. Therefore, it is applicable not only to the alpine regions,
7 but also to the regions described by all topographic features.
8 The results of identifying the avalanche rockfall sources are strongly influenced by the DEM [18].
9 Since the DEM is a simplification of the real terrain, some detailed features may be missing. the lower the
10 accuracy of the DEM, the smaller the slope angle of steep slope is, and its critical slope angle decreases
11 with decreasing the accuracy of the DEM, which will result in the identification range of the avalanche
12 rockfall source much larger than the actual range [14]. Meanwhile, the accuracy of the DEM also affects
13 the results of the rockfall simulation. By comparing the rockfall simulation results of DEM with an
14 accuracy of 1 m and DEM with an accuracy of 20 m, Lan [28] found that the trajectories of the rockfalls
15 simulated by the DEM maps with an accuracy of 1 m are more complex, and the velocity changes of the
16 rockfalls are more frequent as well. In addition, a DEM with an accuracy of 1 m can reflect more
17 micro-geomorphic features, which makes it possible to simulate a lower energy and velocity for an
18 avalanche rockfall during a collision with the ground than that calculated by using a DEM with an accuracy
19 of 20 m. In this study, since a 12.5 m DEM is used as the DEM with the highest accuracy obtained free of
20 charge, to obtain the higher accuracy results, a precise measurement in the field is necessary, which
21 requires a lot of efforts. Therefore, except for the special circumstances required the use of higher accuracy
22 DEMs, for the general situations, the 12.5 m DEM is sufficient.
23 During the rockfall simulation, the slope material properties, such as the normal recovery coefficient,
1 tangential recovery coefficient, and friction coefficient or friction angle, will affect the interaction between
2 the rockfalls and the slope, which will result in a change in the kinematic characteristics of the rockfalls.
3 The slope with vegetation cover has a smaller recovery coefficient and a larger friction angle, which results
4 in a faster decreasing velocity and a shorter moving distance for the rockfall during the falling motion than
5 the general exposed slope. Therefore, the suitable setting for the slope material properties is one of the keys
6 to realize the accurate simulation of rockfalls. By using the field measurement method, the most precise
7 normal and tangential recovery coefficients, and friction angle can be obtained. However, when the studied
8 area is too large, it will lead to difficulties in conducting the field measurement work. In this paper, the
9 material properties of the slopes in the studied area are set by using the aerial photographs. To ensure the
10 accuracy of the numerical modeling, the simulation of the rockfall accident in the Baiyan Village is
11 conducted, and then the simulation results are verified by comparing with the actual data of collapse
12 rockfall accident. Finally, the slope material properties in the whole studied area are set by using the same
13 method.
14 The occurrence of the collapse rockfall accident is also affected by the rainfall [30]. A large amount of
15 rain will fill the rock cracks, which results in the increasing water pressure, reducing the shear strength and
16 frictional resistance, and more likely occurring the collapse rockfall. Furthermore, the scouring effect of the
17 rainwater on the rock bodies and deposits can lead to the instability of the rock bodies, and results in the
18 occurring of collapse rockfall. The rainfall will affect the frequency of the collapse rockfalls, while the
19 rainfall is not the only influencing factor. Except for the factor of rainfall, the other factors, such as the
20 structures of the rock bodies, and the vegetation cover, all will influence the collapse rockfalls. As for the
21 specific qualitative and quantitative relationship between the frequency of rockfall and these factors, the
22 further studies are needed to conduct.

1 5. Conclusions
2 In this paper, the Jinfeng Street in Zhijin Country, Bijie City, Guizhou Province is selected as the
3 studied object. The topographical data of the studied area are obtained by the field survey, the critical slope
4 angles are calculated by using the SAD method, the potential collapse rockfall resource zones are identified
5 by using the topographical maps, and the calculated results are verified by comparing with the historical
6 collapse rockfall incidents. Under the conditions of determining the avalanche rockfall zones, the Rockfall
7 Analyst software is applied to identify the parameters of trajectory, frequency, height, and energy of the
8 slope landslide rockfalls. The modeling parameters of frequency, height, and energy are given various
9 weights, the sum for these parameters with different weights are conducted, and the hazard assessment for
10 the avalanche rockfalls in the Jinfeng Street is carried out. The accuracy of the rockfall model is verified by
11 comparing the simulation results with the actual rockfall accident, and the corresponding governance
12 measures for the rockfall hazard in the Jinfeng Street are proposed. The main conclusions obtained from
13 this study are as follows:
14 (1) Adopting the SAD method, combining with the rock outcrops identified from the statistical
15 topography map, the potential avalanche rockfall source zones are determined, and the area of the rockfall
16 source zone in the Jinfeng Street is calculated as 522,316 m2, which accounts for 0.8% of the total area of
17 the Jinfeng Street.
18 (2) The numerical simulation for the avalanche rockfalls in the Baiyan Village is carried out by using
19 the Rockfall Analyst software, and then the hazard assessment for the rockfalls in the whole Jinfeng Street
20 is conducted. The numerical simulation results indicate that there are 81,105 m2 of house area in the Jinfeng
21 Street within the rockfall hazard zone, with the proportion of high and ultra-high hazard areas accounting
22 for 25.6%.
23 (3) To reduce the hazards caused by the slope landslide rockfalls, according to the Code for Design of
1 Rock Retaining Wall Engineering in Geological Hazards (T/CAGHP 060-2019), a 5-meter-high rock
2 retaining wall is built, and the stability for such a rock retaining wall is verified. The results indicate that
3 after adding the rock retaining wall, the area of the rockfall hazard zone decreases to 2,580,371 m2, and
4 there is no ultra-high hazard zone. The house area in the rockfall hazard zone reduces to 50,432 m2, and the
5 house area in the high hazard zone reduces to 17.8% of the total one in the hazard zone.
6 (4) After adding the rock retaining wall, for the residents who are still in the high hazard zone, a
7 unified relocation strategy is recommended to ensure the safety of residents' lives and property.
8 Acknowledgments
9 This research was funded by the Science and Technology Infrastructure Program of Guizhou Province
10 (2020-4Y047), the Science and Technology Project of Transportation and Communication Ministry of
11 Guizhou Province (2023-312-030), the Opening Fund of Key Laboratory of Geohazard Prevention of Hilly
12 Mountains, Ministry of Natural Resources (KLGHZ202104), the Natural Science Foundation of Hunan
13 Province of China (2023JJ30135), and the Changsha Municipal Natural Science Foundation of China
14 (kq2208031).
15 Conflicts of interests
16 The authors declare no conflict of interest.
17 References
18 1 N. Matsuoka., H. Sakai. Rockfall activity from an alpine cliff during thawing periods. Geomorphology, 28
19 (1999) 309-328.
20 DOI:10.1016/S0169-555X(98)00116-0
21 2 O. Hungr, S. Leroueil, L. Picarelli. The Varnes classification of landslide types: an update. Landslides, 11
22 (2014) 167-194.
1 DOI:10.1007/s10346-013-0436-y
2 3 T. Chalachew, W. Kifle. Characteristics and susceptibility zonation of landslides in Wabe Shebelle Gorge,
3 south eastern Ethiopia. Journal of African Earth Sciences, 182 (2021) 104275.
4 DOI:10.1016/J.JAFREARSCI.2021.104275
5 4 J. Yan, J. Chen, C. Tan, S. Yan, Y. Liu, X. Zhao, Q. Wang. Rockfall source areas identification at local scale
6 by integrating discontinuity-based threshold slope angle and rockfall trajectory analyses. Engineering
7 Geology, 313 (2023) 106993.
8 DOI:10.1016/J.ENGGEO.2023.106993
9 5 C. Dussauge, J. Carine, A. Helmstetter. Statistical analysis of rockfall volume distributions: Implications for
10 rockfall dynamics. Journal of Geophysical Research: Solid Earth, 108 (2007) 2286.
11 DOI:10.1029/2001JB000650
12 6 P. Frattini, G. Crosta, A. Carrara, F. Agliardi. Assessment of rockfall susceptibility by integrating statistical
13 and physically-based approaches. Geomorphology, 94 (2008) 419-437.
14 DOI:10.1016/j.geomorph.2006.10.037.
15 7 M. Xia, B. Hai, N. Jiang, J. Chen, J Zhou. Risk assessment and mitigation evaluation for rockfall hazards at
16 the diversion tunnel inlet slope of Jinchuan Hydropower Station by using three-dimensional terrestrial
17 scanning technology. KSCE Journal of Civil Engineering, 27 (2022) 181-197.
18 DOI:10.1007/S12205-022-1679-8
19 8 X. Wang, H. Liu, J. Sun. A new approach for identification of potential rockfall source areas controlled by
20 rock mass strength at a regional scale. Remote Sensing, 13 (2021) 938.
21 DOI:10.3390/rs13050938
22 9 M. Alvioli, M. Santangelo, F. Fiorucci, M. Cardinali, I. Marchesini, P. Reichenbach, M. Rossi, F. Guzzetti,
23 S. Peruccacci. Rockfall susceptibility and network-ranked susceptibility along the Italian railway.
24 Engineering Geology, 293 (2021) 106301.
25 DOI:10.1016/J.ENGGEO.2021.106301
26 10 M. Cignetti, D. Godone, D. Bertolo, M. Paganone, P. Thuegaz, D. Giordan. Rockfall susceptibility along
27 the regional road network of Aosta Valley Region (northwestern Italy). Journal of Maps, 17 (2021) 54-64.
28 DOI:10.1080/17445647.2020.1850534
29 11 A. Abellan, J. Vilaplana, J. Martinez. Application of a long-range Terrestrial Laser Scanner to a detailed
30 rockfall study at Vallde Nuria (Eastern Pyrenees, Spain). Engineering Geology, 88 (2006) 136-148.
31 DOI:10.1016/j.enggeo.2006.09.012
1 12 A.N.S.D. Albarelli, C.O. Mavrouli, P. Nyktas. Identification of potential rockfall sources using
2 UAV-derived point cloud. Bulletin of Engineering Geology and the Environment, 80 (2021) 1-23.
3 DOI:10.1007/S10064-021-02306-2
4 13 A.M. Fanos, B. Pradhan, A. Alamri, C.W. Lee. Machine learning-based and 3D kinematic models for
5 rockfall hazard assessment using LiDAR Data and GIS. Remote Sensing, 12 (2020) 1755.
6 DOI:10.3390/rs12111755
7 14 A. Loye, M. Jaboyedoff, A. Pedrazzini, Identification of potential rockfall source areas at a regional scale
8 using a DEM-based geomorphometric analysis. Natural Hazards and Earth System Science, 9 (2009)
9 1643-1653.
10 DOI:10.5194/nhess-9-1643-2009
11 15 X. Wang, P. Frattini, D. Stead, J. Sun, H. Liu, A. Valagussa, L. Li. Dynamic rockfall risk analysis.
13 DOI:10.1016/j.enggeo.2020.105622
14 16 J. Zhan, J. Chen, W. Zhang, X. Han, X. Sun, Y. Bao. Mass movements along a rapidly uplifting river valley:
15 An example from the upper Jinsha River, southeast margin of the Tibetan Plateau. Environ. Environmental
16 Earth Sciences, 77 (2018) 634.
17 DOI:10.1007/s12665-018-7825-4
18 17 F. Guzzetti, P. Reichenbach, G.F. Wieczorek. Rockfall hazard and risk assessment in the Yosemite Valley,
19 California, USA. Natural Hazards and Earth System Sciences, 3 (2003) 491.
20 DOI:10.5194/nhess-3-491-2003
21 18 J. Zhan, Z. Yu, Y. Lv, J. Peng, S. Song, Z. Yao. Rockfall Hazard Assessment in the Taihang Grand Canyon
22 Scenic Area Integrating Regional-Scale Identification of Potential Rockfall Sources. Remote Sensing, 14
23 (2022) 3021.
24 DOI:10.3390/RS14133021
25 19 F.Ferrari, A. Giacomini, K. Thoeni. Qualitative rockfall hazard assessment: A comprehensive review of
26 current practices. Rock Mechanics and Rock Engineering, 49 (2016) 2865-2922.
27 DOI:10.1007/s00603-016-0918-z
28 20 K.M. Ansari, M. Ahmad, R. Singh, T.N. Singh, Rockfall Hazard Rating System along SH-72: a case study
29 of Poladpur–Mahabaleshwar road (Western India), Maharashtra, India. Geomatics, Natural Hazards and
30 Risk, 7 (2015) 649-666.
31 DOI:10.1080/19475705.2014.1003416
1 21 S. William, A. Yonathan. Using Google Earth and Google Street View to rate rock slope hazards.
2 Environmental and Engineering Geoscience, 24 (2018) 237-250.
3 DOI:10.2113/eeg-1922
4 22 A.M. Fanos, B. Pradhan. Multi-scenario rockfall hazard assessment using LiDAR Data and GIS. Geotech.
5 Geotechnical and Geological Engineering, 34 (2016) 1375-1393.
6 DOI:10.1007/s10706-016-0049-z
7 23 W.D. Stevens, RocFall, a Tool for Probabilistic Analysis, Design of Remedial Measures and Prediction of
8 Rockfalls. Master Thesis University of Toronto, Toronto, Canada. (1998).
9 24 L. Li, H. Lan. Probabilistic modeling of rockfall trajectories: a review. Bulletin of Engineering Geology
10 and the Environment, 74 (2015) 1163-1176.
11 DOI:10.1007/s10064-015-0718-9
12 25 A. Radtke, D. Toe, F. Berger, S. Zerbe, F. Bourrier. Managing coppice forests for rockfall protection:
13 lessons from modeling. Annals of Forest Science, 71 (2014) 484-494.
14 DOI:10.1007/s13595-013-0339-z
15 26 L. Dorren, F. Berger. New approaches for 3D rockfall modelling with or without the effect of forest in
16 Rockyfor3D. In Proceedings of the EGU General Assembly, Vienna, Austria, 2-7 May (2010) 14811.
17 27 H. Lan, C.D. Martin, C.H. Lim. RockFall analyst: A GIS extension for three-dimensional and spatially
18 distributed rockfall hazard modeling. Computers and Geosciences, 33 (2006) 262-279.
19 DOI:10.1016/j.cageo.2006.05.013
20 28 H. Lan, C.D. Martin, C. Zhou, H. Chang. Rockfall hazard analysis using LiDAR and spatial modeling.
21 Geomorphology, 118 (2010) 213-223.
22 DOI:10.1016/j.geomorph.2010.01.002
23 29 A.N. Strahler. Equilibrium theory of erosional slopes approached by frequency distribution analysis: Part II.
24 American Journal of Science, 248 (1950) 800-814.
25 30 V. Cristina, C. Marta, P. Roberta, M. Andrea, M. Chiara, C. Graziano, N. Guido. An integrated approach to
26 investigate climate-driven rockfall occurrence in high alpine slopes: the Bessanese glacial basin, Western
27 Italian Alps. Journal of Mountain Science, 17 (2020) 2591-2610.
28 DOI:10.1007/s11629-020-6216-y
29
1 List of Tables
4 List of Figures
5 Fig. 1 Map and geographical location: (a-c) Geographical location of study area; (d) The avalanche
6 lo-cation in the Jinfeng Street.
7 Fig. 2 Google Earth history images of Baiyan Village: (a) Image in January 2014; (b) Image in Sep-tember
8 2014; (c) Image in December 2019; (d) Image in March 2022.
11 Fig. 5 The distributions of cliffs, steep slopes, foot slopes, and plains in the topographical feature map of
12 Jinfeng Street.
15 Fig. 8 Landslide map of Baiyan Village. (a) UAV view of the affected area; (b) Google Earth image of the
16 area.
17 Fig. 9 Rockfall simulation in Baiyan village. (a) rockfall accident simulation; (b) simulation with rock
18 retaining wall.
19 Fig. 10 Trajectory and velocity profile of the simulated collapse rockfall in Baiyan Village.
20 Fig. 11 The simulated results of the avalanche rockfalls in the Jinfeng Street: (a) rockfall hazard zone; (b)
21 rockfall trajectory; (c) rockfall frequency; (d) rockfall energy; (e) rockfall velocity; (f) rockfall height.
22 Fig. 12 The simulated results of the rockfall hazard area in detail: (a-d) corresponding to A-D Region in Fig.
1 11.
3 Fig. 14 The cumulative distribution of rockfall hazard zones in the absence of a rock retaining wall: (a) The
4 cumulative distribution of hazard area; (b) The cumulative distribution of house area in the rockfall
5 hazard zones.
6 Fig. 15 The Cumulative distribution of rockfall hazard zones in the presence of a rock retaining wall: (a)
7 The cumulative distribution of hazard area; (b) The cumulative distribution of house area in the
8 rockfall hazard zones.

Manuscript in PDF format Click here to view linked References
1 Rockfall hazard assessment and governing strategy for a cliff feature zone based
2 on the high-resolution online images and terrain analysis
3 Rongguo Zhao* 1,2, Jinchen Yang1,3, Suhua Zhou4, Minghua Huang4, Minghui Yang5
4 1. Key Laboratory of Dynamics and Reliability of Engineering Structures of College of Hunan Province,
5 Xiangtan University, Xiangtan 411105, China;
6 2. School of Mechanical Engineering and Mechanics, Xiangtan University, Xiangtan 411105, China;
7 3. College of Civil Engineering, Xiangtan University, Xiangtan 411105, China;
8 4. College of Civil Engineering, Hunan University, Changsha 410082, China;
9 5. School of Architecture and Civil Engineering, Xiamen University, Xiamen 361005, China.
10 Corresponding authors: zhaorongguo@xtu.edu.cn
11
12 Abstract
13 The geological disaster caused by the slope landslide rockfall has some significant properties, such as
14 suddenness, unpredictability, higher harmfulness. In this paper, a cliff feature zone in Guizhou Province of
15 China is selected as a case, the topographic data in the study area are obtained by the field survey, and the
16 slop angle distribution (SAD) method with the topographic map is adopted to identify the potential rockfall
17 source zone. The Rockfall Analyst software is applied to simulate the trajectory, frequency, height, and
18 energy of the slope landslide rockfalls, and the accuracy of the rockfall model is verified by comparing the
19 simulation results with the actual accident. Then, the rockfall hazard assessment is conducted for the whole
20 Jinfeng street, and relevant governing strategies are proposed. The results show that the rockfall source
21 zone area in the Jinfeng Street validated by using the SAD method with the rock outcrops in the
22 topographic map is identified as 522,316 m2, which is 0.8% of the total area. The building area in the
1 hazard zone are 81,105 m2, and the percentage of building area in the high-hazard and ultra-high-hazard
2 zone reaches 25.6%. To reduce the hazards caused by the slope landslide rockfalls, according to the Code
3 for Design of Rock Retaining Wall Engineering in Geological Hazards (T/CAGHP060 -2019), a stone
4 retaining wall with the height of 5 m is built, and its stability is verified. The results indicate that after
5 adding the stone retaining wall, the rockfall hazard zone area decreases to 2,580,371 m2, there is no
6 ultra-high hazard zone, and the building area in the rockfall hazard zone reduces to 50,432 m2. After adding
7 the stone retaining wall, for the residents who are still in the high hazard zone, a unified relocation strategy
8 is recommended to ensure the safety of residents’ lives and property.
9 Key words
10 rockfall hazard assessment; governing strategy; landslide; terrain analysis; Geographic Information System
11
12 1. Introduction
13 Rockfall means the collapse of rocks on the surface of steep cliffs and slopes under the gravity force
14 [1-4]. Usually, the rockfalls are sudden, unpredictable, fast, high energy and hazardous, and occur
15 frequently around the world [5-7]. Nowadays, many scholars have focused their eyes on the geological
16 hazard caused by the slope landslide rockfalls. Therefore, to reduce the hazards of rockfalls, how to identify
17 the potential rockfalls quickly and accurately, and how to evaluate the impact of avalanche rockfall hazards,
18 have become the continuous concerned questions in the academic community.
19 A rockfall hazard assessment requires the identification of potential rockfall sources [8-10]. Since the
20 potential rockfalls are basically distributed on the cliffs or steep slopes in the mountainous areas, the
21 traditional manual investigation methods are greatly limited. Then, some instruments and techniques, such
22 as the Unmanned Aerial Vehicle (UAV) photography and the three-dimension (3D) scanning, have been
1 utilizing for the rockfall source identification. However, these instruments and techniques have some strict
2 requirements for the application environment, which limits their utilization in the complex mountainous
3 terrain [11-13]. Subsequently, the methods of determining the source of landslide rockfalls by using the
4 aerial photographs, the remote sensing images, and the high-precision Digital Elevation Models (DEMs),
5 which have the characteristics of high precision, high accuracy, and wide applicability, are widely used in
6 the investigation of the area of the potential landslide rockfall source [14-16]. Based on the statistical data
7 on the location of historical avalanche rockfalls, Guzzetti [17] suggested that the cliffs and slopes became
8 the potential sources of the avalanche rockfall when the critical slope angle was greater than 60°. Frattini [6]
9 constructed a 3D rockfall model and mapped the rockfall sources by using the Guzzetti’s model, and
10 empirically identified the potential avalanche rockfall sources when the critical slope angle is greater than
11 37°. Such studies are based on the statistical analysis of the historical data in the specific areas combined
12 with the theoretical assumptions, and some further research works are needed to determine the critical slope
13 angles in the other areas. Thereafter, based on the Slop Angle Distribution (SAD), a method of dividing the
14 homogeneous morphometric areas by using the geomorphic analysis and then determining the critical slope
15 angle of the avalanche rockfalls is proposed, which has an excellent applicability in the area with different
16 geological patterns [14].
17 According to the parameters of rockfalls, such as the trajectory and energy, the rockfall hazard
18 assessment is conducted on estimating the extent of damage [18-19]. The earliest method to assess the
19 rockfall hazard was to analyse the impact factors of rockfalls and established a corresponding evaluation
20 system, but this method only gave a qualitative evaluation of rockfall hazard, which brought some highly
21 subjective factors [20-21]. With the development of computer and calculation software, pursuing the
22 method based on the physical-mechanical models to quantitatively simulate the trajectories of the
23 avalanche rockfalls and assess the rockfall hazard have gradually become a mainstream [22]. Among them,
1 the commonly used two-dimension (2D) simulation software is the Rockfall, while the popularly adopted
2 3D ones are the Rockyfor3D, the Rockfall Analyst, etc. The Rockfall software, which is based on the
3 statistics, is a two-dimensional simulation software that allows almost all of its inputs to be varied
4 randomly [23-24]. However, the profile of rockfall’s motion is artificially specified, so there is a certain
5 degree of human control. Simultaneously, because it is two-dimensional, it cannot consider the lateral
6 displacement of the collapse fallout. The Rockyfor3D software has been widely used to evaluate the
7 rockfall hazard in the forested areas [25-26], which takes the shapes and sizes of individual rock into
8 account in the simulation, and the more accurate the rockfall parameters are, the more accurate the
9 simulated results will be. Yet, it is difficult to obtain the accurate data of the rockfalls, so its accuracy will
10 be affected when the question with a wide range of rockfalls is faced. The Rockfall Analyst software adopts
11 the centralized quality method to simulate the trajectory of rockfall [27-28], which, as an extension of
12 ArcGIS, presents the simulated trajectory, bouncing height, velocity and other distributions of rockfall
13 simulation in the ArcGIS platform by using the vector and raster data, and it has a wide range of
14 applications in the prediction and assessment of the rockfall hazards. Comparing with the other software,
15 the Rockfall Analyst software does not need to consider the size and shape of the rockfall either, nor does it
16 need to consider the profile direction of the rockfall movement. Due to the benefits that there are few
17 human influence factors, it is more suitable for the simulation of the avalanche rockfalls that occurs
18 suddenly and without warning.
19 In this paper, the background of May 8, 2022, Baiyan Village, Chengguan Town, Zijin County, Bijie
20 City, Guizhou Province is taken as the case of the rockfall event. Firstly, the topographic data for the
21 Baiyan Village is obtained via the field survey. Then, according to the topographic data, the potential
22 sources of rockfalls are predicted by using the method of slop angle distribution (SAD) combined with the
23 topographic maps. Following that, according to the potential sources of rockfalls and the high-precision
1 DEM, the avalanche rockfalls are simulated by using the Rockfall Analyst software, the trajectory,
2 frequency, bouncing height and other kinetic characteristics of rockfalls are calculated, and then the
3 rockfall hazard is evaluated by assigning various weights to the frequency, bouncing height, and energy of
4 rockfalls. Finally, to validate the accuracy of the estimated model, the simulation results of the Baiyan
5 Village rockfall event are compared with the actual situation, and then the rockfall hazard in the Jinfeng
6 Street is simulated and evaluated, and the corresponding reasonable strategies for preventing and
7 controlling the rockfall hazard are proposed.
8 2. Geographical location and geological environment
9 2.1 Geographical location
10 Guizhou Province is in the southwestern China, with steep terrain and mountainous areas accounting
11 for 90% of the province's area. Every year, Guizhou is affected by a variety of geologic hazards. According
12 to the Bureau of Natural Resources of Guizhou Province, there were a total of 28,221 potential geologic
13 hazards with various types as of the end of 2021, including 4,812 landslides, 3,091 avalanches and rockfalls,
14 114 mudslides, and 1,377 unstable slopes. At 12:00 a.m. on May 8, 2022, a mountain collapse occurred on
15 the northwest side of the Baiyan Village, Jinfeng Street, Chengguan Town, Zhijin County, Bijie City,
16 Guizhou Province. The map and geographical location of the Jinfeng Street are shown in Fig. 1. The volume of
17 debris from the avalanche was about 25,000 m3, and the rockfalls extended as far as about 950 m, burying
18 three farmers working in the surrounding area and destroying more than 150 m of the local road and several
19 houses. The avalanche occurred on a steep slope with an elevation of about 1,350 m and a gradient of 64.3°.
20 The collapsed rock mass was mainly sandstone, chert, and shale. No rainfall, earthquakes, engineering
21 excavations, or blasting activities were recorded or reported in the vicinity prior to the avalanche. Therefore,
22 it is initially speculated that the avalanche may be the collapse of a broken rock body caused by long-term
1 weathering. In addition, according to the Zijin County Emergency Response Department, there is still a
2 large possibility of collapse destabilization in the mountains around the avalanche site, and the potential
3 volume of rockfall from the avalanche is about 10,000 m3. According to the relevant information from the
4 local government of Zhijin County, the area where the avalanche occurred was identified as a
5 collapse-prone site as early as ten years ago, but its potential area of influence was not delineated.
6 The Google Earth history images of the Baiyan Village in January 2014, September 2014, December
7 2019, and March 2022 were individually shown in Fig. 2(a), 2(b), 2(c) and 2(d). In can be found from Fig.
8 2(a) that in January 2014, the whole mountainous area in the Baiyan Village was covered by the
9 surrounding lush vegetation. However, in September 2014, some exposed rocks began to appear on the
10 slopes in the Baiyan Village, as shown by the yellow block diagram in Fig. 2(b). As time passed, in
11 December 2019, the exposed rocks on the slops in the Baiyan Village developed into the collapsed rocks,
12 and some new exposed rocks appeared. The collapsed rocks and new exposed rocks are shown by the red
13 block diagrams and the yellow one in Fig. 2(c), respectively. In March 2022, the rockfall area in the Baiyan
14 Village was recorded by the Google Earth, and the avalanche rockfall and the exposed rocks are
15 individually shown by the red block diagram and the yellow one in Fig. 2(d).
1
2 Fig. 1 Map and geographical location: (a-c) Geographical location of study area; (d) Avalanche location in the Jinfeng Street.
3 (a) (b)
4 (c) (d)
5 Fig. 2 Google Earth history images of Baiyan Village: (a) Image in January 2014; (b) Image in September 2014; (c) Image in
6 December 2019; (d) Image in March 2022.

1 2.2 Geological environment
2 The Jinfeng Street, at which the avalanche occurred, is dominated by the highland mountainous
3 terrain, and the elevation of the study area ranges from 1,341 m to 2,129 m as shown in Fig. 1(d), with a
4 mean slope of 20.3°, while the steepest area reaches 80°, as shown in Fig. 2(a). In the study area, there are
5 two obvious "northeast-southwest" trending steep slope zones, with a lot of exposed weathered rock bodies
6 distributed on the high steeps or even near-erect slopes, which have a high hazard of collapse. As shown in
7 Fig. 3(b), the main stratigraphy of the study area includes the following three types. The first type is the
8 Upper Permian (P↓2-3): intermediate chert and flint limestone with sandstone shale interbedded with coal
9 at the bottom. The second one is the Shallow Marine Lower Triassic (T↓1↑MS): grayish-green shale at
10 the bottom of stone. The third one is the Carboniferous (CP): dolomite and chert at the bottom of stone.
11
13 3. Methodology for identification and assessment
14 The assessment of the impact area of the rockfall hazard is divided into two main steps, as shown in
15 Fig. 4. Firstly, the rockfall source area needs to be identified. Secondly, after the avalanche occurs, the
16 trajectory of the potential destructive point needs to be analysed.
17 In this paper, based on the terrain slope distribution theory proposed by Loye [14], the areas with the
1 same characteristics of lithology, historical tectonic morphology are divided into homogeneous
2 morphometric areas (HMAs), and then the Gaussian distribution is applied to the terrain morphometric
3 elements in the HMAs to find out the critical value of the slope angle. Combining with the vegetation
4 distribution in the HMAs, the exposed rocks are determined, and the potential sources of avalanche
5 rockfalls are obtained. According to the potential rockfall sources and the site investigation, the Rockfall
6 Analyst software is used to simulate the rockfall, and the trajectory, frequency, speed, height, and energy
7 maps of rockfall are derived. To predict the rockfall hazard zone, the different weights are assigned to the
8 parameters of frequency, height, and energy. The technical flow chart used in this paper is shown in Fig. 4.
11 3.1 Identification of potential rockfall sources
12 Different lithologies, tectonic patterns, and stratigraphic geologies have some effects on rockfall.
13 Therefore, areas with the same lithology, historical tectonic pattern, and geology are classified as the
14 homogeneous morphometric areas [29]. By using ArcGIS, the lithology, historical tectonic morphology,
15 and geology of the study area are superimposed to delineate the homogeneous morphometric areas, and
16 then the SAD method is used to identify the potential avalanche rockfalls in the homogeneous
17 morphometric areas. According to the SAD characteristics, the homogeneous morphometric areas is
18 decomposed into the Gaussian distributions with the specific terrain patterns, and the distributions of the
1 cliffs, steep slopes, foot slopes, and plains in the topographical feature map are shown in Fig. 5.
3 Fig. 5 The distributions of cliffs, steep slopes, foot slopes, and plains in the topographical feature map of Jinfeng Street.
4 Since the effective area of each raster in the DEM depends on the size of the slope angle, the actual
5 terrain area is weighted according to the slope angle of the raster cells [14], and the expression for the
6 frequency of slope angle distribution is written as
Ah 
7 W  (1)
AHM A cos β
8 where, Ah  is the total area of the DEM with a slope angle of β , and AHMA is the sum of the areas of
9 the HMA.
10 Using the slope distribution method, the terrain is divided into several major morphological units, and
11 then the Gaussian distribution is applied to these units. Fig. 6 shows the Gaussian distributions of terrain
12 slope for various geological patterns. The terrain studied in this paper can be divided into the following
13 four geological patterns.
14 (1) Plains: the low-slope areas characterized by the rivers and glacial deposits.
15 (2) Foot slopes: an area of gentle slope angle in the lower portion of a slope characterized by the
16 alluvial fans, debris flows, and landslide deposits.

1 (3) Steep slopes: an area of steeper slopes characterized by the slope deposits and plant-covered rock
2 outcrops.
3 (4) Cliff: an area of steep slopes characterized by the exposed rock outcrops.
Frequency
0.06 Slope Angle Distribution(SAD)
Sum of SAD R2=0.984
Plains
0.05
Foot slopes
B Steep slopes
0.04 Cliff
0.03
0.02
0.01
A
0
0 10 20 30 40 50 60 70 80 90
Slope (?
4
6 In the above morphological elements, the steep slope element and cliff one may contain the rockfall
7 sources. Therefore, for each homogeneous zone's slope angle, the Gaussian distribution is used separately
8 to determine the critical slope angle. The specific method for calculating the critical slope is as follows.
9 (1) The critical slope angle is defined as the point of intersection between the two steepest
10 morphological elements, as shown the point A in Fig. 6, which is the surface of lightly vegetated cliff
11 element and steep slope one, and is recognized as a potential source of avalanche rockfall when the slope
12 angle is greater than the critical slope angle, regardless of lithology and land cover.
13 (2) In some special cases, there may be no cliff element. At this time, the steep slope elements can be
14 considered as the steepest geological morphological elements with the property of the Gaussian
15 distribution.
16 (3) In the steep regions, based on the slope angle analysis, two cliff form elements with the property of
17 the Gaussian distribution are divided. In this case, the cliff element with a lower slope is selected as the
18 research object, and then the critical slope angle is calculated according to (1).
1 (4) If the rockfalls occur in the cliff and steep slope elements, then considering the entire steep slope
2 as a rockfall source will overestimate the region of the potential rockfall source. In this case, only if the
3 angle is higher than the mean of steep slope, as shown the point B in Fig. 6, the rockfall source is identified
4 as a potential rockfall source.
5 Combining with the terrain slope distribution method and the high-precision DEM maps, the various
6 HMAs are investigated by using the Geographic Information System (GIS), and the potential avalanche
7 rockfall resources are identified.
8 3.2 Assessment of rockfall hazards
9 The rockfall simulation is conducted by using the Rockfall Analyst software, which is an extensive
10 software of GIS. The rockfall simulation is divided into two main steps.
11 (1) Constructing a 3D model for the avalanche rockfall;
12 (2) Constructing a grid model for the spatial distribution of the collapse rockfalls.
13 For the Rockfall Analyst software, the centralized mass method is adopted, which means that when
14 using this software to simulate the avalanche of rockfalls, the sizes and shapes of the rockfalls do not need
15 to be considered, but only needs to input the parameters, such as the area, number, and the initial velocity
16 of the avalanche rockfall. The interaction between the avalanche rockfall and the slope is controlled by
17 three coefficients known as the normal recovery coefficient, tangential recovery coefficient, and friction
18 coefficient, which remain constants during the simulation process.
19 The expression for calculating the rockfall trajectory can be written as
   
 X 0  Vx 0t   X 0  Vx 0   0 
       
20 L   Y0  Vy 0t    Y0   Vy 0  t   0  (2)
 1   Z  V   1 
Z0  Vz 0t  gt 2   0   z 0   gt 2 
 2   2 
21 where, X 0 , Y0 and Z 0 are individually the initial positions of the rockfall in x , y and z direction,
1 t is the time, g is the acceleration of gravity whose value is taken as 9.8 m/s2, Vx 0 , V y 0 , Vz 0 are the
2 initial velocities in x , y and z direction, respectively. The formula for calculating the rockfall velocity
3 v v can be expressed as
 Vx 0  Vx 0   0 
     
4 vv   Vy 0   Vy 0    0  (3)
V  gt  V   gt 
 z0   z0   
5 When the rockfall undergoes collision and bounces, its inclination, trend, and normal velocity
6 components VDip  , and VN are affected by the tangential recovery coefficient and normal recovery
 , VTrend
7 coefficient, respectively. The corresponding expressions for calculating the inclination, trend, and normal
8 velocity components of rockfall are
9 VDip   RTVTrend , VN  RNVN

  RTVDip , VTrend (4)
10 where, VDip , VTrend , and VN are individually the inclination, trend, and normal velocity components of
11 the elastic collision velocity without energy loss, RT and RN are individually the tangential recovery
12 coefficient and normal recovery coefficient, whose values are in the range of [0, 1].
13 By using the Rockfall Analyst software, not only the trajectory and velocity of avalanche rockfall can
14 be simulated, but also the physical parameters, such as the frequency, height, and energy can be calculated.
15 One this basis, different weights are assigned to the physical parameters of frequency, height, and energy,
16 then these physical parameters are weighted and combined to evaluate the hazard of avalanche rockfall.
17 4. Results and discussion
18 4.1 Results of potential rockfall
19 Since the studied site is in Jinfeng Street, Zhijin County, Bijie City, Guizhou Province, the study area
20 is relatively small and the lithology within the area is similar. Therefore, the region can be regarded as a
21 homogeneous morphology area. To minimize errors, based on the combination of topographical maps, the
22 calculated result is compared with the historical rockfall events to ensure the prediction accuracy of
1 avalanche rockfall source. The slope angle thresholds for potential rockfall are shown in Table 1.
Critical slope angle
HMA Area of potential rockfall (m2)
Cliff A Steep slope B
1 38.3° 22.1° 522316.3
3 It can be seen from Table 1 that the critical slope angle of potential rockfall in Jinfeng Street is 38.3°.
4 By using GIS, the slope angles in the potential avalanche rockfall zone are mapped, and the regions with
5 the slope angles greater than this threshold are identified. Fig. 7 show the potential rockfall sources in the
6 Jinfeng Street. It can be seen form the Fig. 7 that the historical disaster sites are all in the identified
7 potential rockfall zones, and the current rockfall in the Baiyan Village is also in the identified potential
8 rockfall zones, which verifies the accuracy of using GIS to identify the potential rockfall sources.
11 4.2 Results of rockfall hazard assessment
12 The simulation and analysis for the avalanche rockfall hazard in the Jinfeng Street, Zijin County, Bijie
13 City, Guizhou Province is conducted by using the Rockfall Analyst software. The identified potential
14 rockfall source information is imported into GIS, and then the corresponding DEM information is inputted.
15 DEM, which reflects the real terrain, is one of the keys to influence the accuracy of simulation. The higher
16 the accuracy of the DEM, the more it can restore the real collapse of rockfall. The commonly used DMEs
1 are 30 m and 12.5 m. The DEM used in this study is 12.5 m. After all the data are imported into GIS, the
2 parameters in the simulation are started to be set. The parameters in the simulation are divided into two
3 parts. The first part of simulation parameters are the parameters of rockfall, in which the mainly parameters
4 are the quality and quantity of rockfall. It is found that the change of rockfall mass has a greater effect on
5 the simulation of the energy results of rockfall, while has a small effect on the simulation of the trajectory
6 of rockfall. To make the simulation results match with the actual situation, the mass of the rockfall is
7 determined according to the field measurement for rockfalls. The measurement masses of the rockfalls
8 mostly range from 200 kg to 400 kg. Therefore, the mean mass, which is 300 kg, is adopted as the mass of
9 rockfall. If the number of rockfall simulations is too small, then the rockfall coverage will not be correctly
10 reflected. While if the number of rockfall simulations is too large, then the simulation time will increase
11 significantly and many rockfall simulation trajectories will overlap. Since the potential rockfalls in the
12 studied area are of 0.5 m apart, setting the number of simulated rockfalls to 1 will correctly reflect the
13 coverage region of rockfall. The second part of the simulation parameters are the parameters of slope
14 property, such as the normal recovery coefficient, tangential recovery coefficient, friction coefficient. To
15 obtain the simulation results that are closely resemble the actual rockfall, the slope characteristics of the
16 entire studied area are investigated by using the aerial photographs, then according to Table 2, the
17 determined material property parameters for the specific slope in the studied area are inputted to simulate
18 the avalanche rockfall hazard.
Characteristics Normal restitution (RN) Tangential restitution (RT) Friction angle (°)
Slopes covered by grasses and shrubs 0.30 0.80 25
Soft and strongly weathered rocks 0.35 0.85 20
Hard rock and masonry 0.40 0.90 15
Soft soil slope 0.30 0.80 30
Water 0.00 0.00 80
20 The landslide map of the Baiyan Village is shown in Fig. 8. Comparing with the actual rockfall data in
21 the Baiyan Village rockfall, the accuracy of the simulation results is verified. The simulated trajectory of
1 rockfall in the Baiyan village is shown in Fig. 8, and the trajectory of A-A’ and the velocity profile in the
2 collapse rockfall simulation are shown in Fig. 9. It can be seen from Fig. 8 that comparing the simulated
3 results with the actual data, the trajectory and range of the collapse rockfall are roughly fitted with the
4 actual ones, and the rockfall accumulation region is matched with the actual shooting situation.
5 Simultaneously, it can be found from Fig. 10 that the rockfall distance in Baiyan Village is nearly 450 m,
6 and the maximum velocity of the avalanche rockfall reaches 52 km/h. These simulated results are
7 consistent with the actual measured data, which verifies the reliability of the simulation model.
8 (a) (b)
9 Fig. 8 Landslide map of Baiyan Village. (a) UAV view of the affected area; (b) Google Earth image of the area.
10 (a) (b)
11 Fig. 9 Rockfall simulation in Baiyan village: (a) rockfall accident simulation; (b) simulation with rock retaining wall.
1
2 Fig. 10 Trajectory and velocity profile of the simulated avalanche rockfall in Baiyan Village.
3 After the accuracy of the simulation model is verified, the parameters of trajectory, velocity, energy,
4 height, and frequency for the avalanche rockfalls in the Jinfeng Street are identified by using the Jinfeng
5 Street’s DEM map with precision of 12.5 m. The simulated results of the avalanche rockfalls in the Jinfeng
6 Street are shown in Fig. 11.
7 (a) (b) (c)
8 (d) (e) (f)
9 Fig. 11 The simulated results of the avalanche rockfalls in the Jinfeng Street: (a) rockfall hazard zone; (b) rockfall trajectory;
10 (c) rockfall frequency; (d) rockfall energy; (e) rockfall velocity; (f) rockfall height.
1 The various weights are individually assigned to the simulated parameters of velocity, frequency,
2 height, and energy of the collapse rockfalls, and then the sum for these parameters with various weights are
3 conducted. Then, to realize the rockfall hazard assessment, the hazard levels are graded by using the natural
4 breakpoint method. The formula for the rockfall hazard assessment is written as
5 H (i)  Wf  Pf (i)  Wh  Ph (i)  We  Pe (i) (5)
6 where, H (i) is the hazard degree of the ith grid, Pf (i) , Ph (i) , Pe (i) are individually the frequency,
7 height, and energy of the rockfall in the ith grid, and Wf , Wh , We are the weights for the parameters of
8 frequency, height and energy, respectively. In this paper, the weight coefficients for the parameters of
9 frequency, height and energy are taken as 0.5, 0.2, and 0.3, respectively.
10 The rockfall hazard zone identified by using the simulation method is shown in Fig. 11(a). It can be
11 seen from Fig. 11(a) that the rockfall hazard zone can be divided into four level regions, which are called as
12 the low hazard region, the medium hazard region, the high hazard region, and the ultra-high hazard region.
13 According to the experts’ advisements, for the ultra-high hazard region, and the high hazard region as well,
14 the residents in this area must be relocated because the rockfall hazard may be seriously affect their safety
15 of lives and properties. While for the medium hazard region, or the low hazard region, so long as
16 appropriate governances are conducted for these regions, the residents can still live in these areas. Within
17 the overall studied area, only the general situation of the landslide rockfall hazard zone can be observed,
18 whereas, the hazard level of each village cannot be directly observed. Therefore, the rockfall zone in the
19 Jinfeng Street is divided into four regions. Thus, the affected situations for every village can be detailly
20 observed from each of the rockfall hazard regions.

1 (a) (b)
2 (c) (d)
3 Fig. 12 The simulated results of the rockfall hazard area in detail: (a-d) corresponding to A-D Region in Fig. 11.
4 The simulated results of the rockfall hazard area in detail, which are individually corresponding to A,
5 B, C, D regions, are shown in Fig. 11. It can be seen from Fig. 11 that in the whole Jinfeng Street, there are
6 still a lot of houses in the high hazard and ultra-high hazard regions, which has a considerable impact on
7 the local residents’ safety of lives and properties. How to reduce the impact of rockfall hazards, this is a
8 question facing to the rockfall hazard governance in the Jinfeng Street. According to the Code for Design of
9 Rock Retaining Wall Engineering in Geological Hazards (T/CAGHP060-2019), under the steep cliffs and
10 steep slopes, a rock retaining wall, whose height is not large than 5 m, can be used for the protection of
11 hazardous rock bodies with a certain width of gentle slopes. Therefore, at a location where the rockfalls
12 maybe threaten the houses, a 5-meter-high rock retaining wall is built to protect the houses from rockfalls.
13 The rock retaining wall with a depth of 1 m, is made of C35 concrete, and no slope is designed for the rock
14 retaining wall. According to the Code, the checking calculations of the shear resistance, impact resistance,
15 foundation bearing capacity, and the stabilities of overturning resistance and sliding resistance as well, are
1 necessary to be done. In this study, the property of impact resistance for the rock retaining wall is mainly
2 investigated. Therefore, the limit situation of the rockfalls must to be considered. According to the
3 simulation results for the rockfalls, before colliding with the rock retaining wall, the fastest rockfall with
4 velocity of 9.5 m/s, is selected as the studied object. In this case, if it is verified to be safe, then the rock
5 retaining wall is validated to be safe. The impact force of rockfall is calculated as the following expression,
6 that is
   
7 p  2 Z 2tan 4  45o    1 (6)
  2 
G 1
8 Z  vi  (7)
2g S  
2tan4  45o   1
 2
9 where, p is the equivalent impact load of rockfall,  is the weight per unit volume of the buffer layer
10 weight , Z is the impact plunging depth,  is the slope angle of the rock retaining wall, g is the
11 acceleration of gravity, and G is the weight of rockfall.
12 After adding the 5-meter-high rock retaining wall, the diagram of rockfall hazard zone is shown in Fig.
13 13. The numerical calculation results are obtained by using the simulation model. The impact plunging
14 depth is calculated as 0.42 m, and the equivalent impact load is calculated as 38.5 kN/m. The compressive
15 strength of rockfall on the rock retaining wall is derived as 96.4 kPa, which is lower than the compressive
16 strength of C35 concrete, so the rock retaining wall will not be damaged.
17
1 The cumulative distribution of rockfall hazard zones in the absence of a rock retaining wall is shown
2 in Fig. 14. It can be found from Fig. 14 that in the case without protective measures, the houses in the
3 central part of Region A are all in the low to medium hazard zone, while the house area in the high and
4 ultra-high hazard zones in the southwestern part of Region A account for 23.4% of the whole house area,
5 and the rest of the houses are in the low to medium hazard zone. In the Region B, although the proportion
6 of the high-hazard and ultra-high-hazard areas to the total area reaches 38.6%, and that of the high-hazard
7 and ultra-high-hazard areas to the total hazard zone area reaches 39.9%, the housing area to the whole area
8 in Region B only accounts for 1.6%, which is very small. In the Region C, the proportion of the high and
9 ultra-high hazard area is 32.2%, and that of the house area in the high and ultra-high hazard zones to the
10 total house area in the whole hazard zones is 9.6%. In the Region D, the proportion of the high and
11 ultra-high hazard areas is 4.2%, and that of the house area in the high and ultra-high hazard areas to the
12 total house area in the whole hazard area accounts for 20.6%.
13 The cumulative distribution of the rockfall hazard zone in the presence of a rock retaining wall is
14 shown in Fig. 15. It can be found from Fig. 15 that the total area of the hazard zone reduces to 2,580,371
15 m2 with the additional 5-meter-high rock retaining wall, which is 39.3% lower than that without a rock
16 retaining wall, and the area of the high and ultra-high hazard zones reduces to 966,397 m2, which is 2.9%
17 lower than that without a rock retaining wall. Simultaneously, after adding a 5-meter-high rock retaining
18 wall, the total house area in the hazard zone reduces from 86,556 m2 to 50,432 m2, which is 41.7% to the
19 original hazard zone area, and the proportion of the house area in the high and ultra-high hazard areas
20 reduces from 25% to 17.8%. The above analysis indicates that the measure of adding a 5-meter-high rock
21 retaining wall can reduce the rockfall hazard to a certain degree, but it cannot eliminate the rockfall hazard
22 completely. Therefore, after adding the stone retaining wall, for the residents who are still in the high
23 hazard zone, a unified relocation strategy is recommended to ensure the safety of residents’ lives and
1 property.
Low hazard
Low hazard Medium hazard
5×106 High hazard 9×104 Ultra-high hazard
Ultra-high hazard
8×104
4×106

7×104
6×104
Hazard area（m2）
3×106
5×104
4×104
2×106
3×104
1×106 2×104
1×104
0 0
2 (a) Hazard zone
(b) Hazard zone
3 Fig. 14 The cumulative distribution of rockfall hazard zones in the absence of a rock retaining wall: (a) The cumulative
6×104 Low hazard

3×106 Low hazard Medium hazard
High hazard 5×104
4×104
2×106
Hazard area (m2)
3×104
1×106 2×104
1×104
0 0
Hazard zone
5 (c) Hazard zone
(d)
6 Fig. 15 The Cumulative distribution of rockfall hazard zones in the presence of a rock retaining wall: (a) The cumulative
8 4.3 Discussions
9 There are many methods for identifying the potential collapse rockfall sources, such as the traditional
10 manual measurement method, the UAV photogrammetric technique, and the three-dimensional scanning
11 technique. Whereas, when these methods are selected to identify the rockfall sources, there are inevitably a
12 certain degree of limitations. In this paper, to identify the potential rockfall zones, the slope angle
13 distribution method is adopted. Firstly, each homogeneous morphology zone is determined. Then, the slope
14 angle distribution of each homogeneous morphology zone is obtained by using the Gaussian distribution,
1 and the critical slope angles of the potential rockfall sources are calculated. The rockfall sources in the
2 studied area are identified by using the slope angles and rock outcrops obtained from the aerial
3 topographical maps, and the accuracy of the simulated results is verified by comparing with the historical
4 information of the landslide rockfall hazards in the Jinfeng Street. In the case without taking the influence
5 of the rock mass structure into account, the topographic slope distribution method can be utilized to identify
6 the potential rockfall resources on a regional scale. Therefore, it is applicable not only to the alpine regions,
7 but also to the regions described by all topographic features.
8 The results of identifying the avalanche rockfall sources are strongly influenced by the DEM [18].
9 Since the DEM is a simplification of the real terrain, some detailed features may be missing. the lower the
10 accuracy of the DEM, the smaller the slope angle of steep slope is, and its critical slope angle decreases
11 with decreasing the accuracy of the DEM, which will result in the identification range of the avalanche
12 rockfall source much larger than the actual range [14]. Meanwhile, the accuracy of the DEM also affects
13 the results of the rockfall simulation. By comparing the rockfall simulation results of DEM with an
14 accuracy of 1 m and DEM with an accuracy of 20 m, Lan [28] found that the trajectories of the rockfalls
15 simulated by the DEM maps with an accuracy of 1 m are more complex, and the velocity changes of the
16 rockfalls are more frequent as well. In addition, a DEM with an accuracy of 1 m can reflect more
17 micro-geomorphic features, which makes it possible to simulate a lower energy and velocity for an
18 avalanche rockfall during a collision with the ground than that calculated by using a DEM with an accuracy
19 of 20 m. In this study, since a 12.5 m DEM is used as the DEM with the highest accuracy obtained free of
20 charge, to obtain the higher accuracy results, a precise measurement in the field is necessary, which
21 requires a lot of efforts. Therefore, except for the special circumstances required the use of higher accuracy
22 DEMs, for the general situations, the 12.5 m DEM is sufficient.
23 During the rockfall simulation, the slope material properties, such as the normal recovery coefficient,
1 tangential recovery coefficient, and friction coefficient or friction angle, will affect the interaction between
2 the rockfalls and the slope, which will result in a change in the kinematic characteristics of the rockfalls.
3 The slope with vegetation cover has a smaller recovery coefficient and a larger friction angle, which results
4 in a faster decreasing velocity and a shorter moving distance for the rockfall during the falling motion than
5 the general exposed slope. Therefore, the suitable setting for the slope material properties is one of the keys
6 to realize the accurate simulation of rockfalls. By using the field measurement method, the most precise
7 normal and tangential recovery coefficients, and friction angle can be obtained. However, when the studied
8 area is too large, it will lead to difficulties in conducting the field measurement work. In this paper, the
9 material properties of the slopes in the studied area are set by using the aerial photographs. To ensure the
10 accuracy of the numerical modeling, the simulation of the rockfall accident in the Baiyan Village is
11 conducted, and then the simulation results are verified by comparing with the actual data of collapse
12 rockfall accident. Finally, the slope material properties in the whole studied area are set by using the same
13 method.
14 The occurrence of the collapse rockfall accident is also affected by the rainfall [30]. A large amount of
15 rain will fill the rock cracks, which results in the increasing water pressure, reducing the shear strength and
16 frictional resistance, and more likely occurring the collapse rockfall. Furthermore, the scouring effect of the
17 rainwater on the rock bodies and deposits can lead to the instability of the rock bodies, and results in the
18 occurring of collapse rockfall. The rainfall will affect the frequency of the collapse rockfalls, while the
19 rainfall is not the only influencing factor. Except for the factor of rainfall, the other factors, such as the
20 structures of the rock bodies, and the vegetation cover, all will influence the collapse rockfalls. As for the
21 specific qualitative and quantitative relationship between the frequency of rockfall and these factors, the
22 further studies are needed to conduct.

1 5. Conclusions
2 In this paper, the Jinfeng Street in Zhijin Country, Bijie City, Guizhou Province is selected as the
3 studied object. The topographical data of the studied area are obtained by the field survey, the critical slope
4 angles are calculated by using the SAD method, the potential collapse rockfall resource zones are identified
5 by using the topographical maps, and the calculated results are verified by comparing with the historical
6 collapse rockfall incidents. Under the conditions of determining the avalanche rockfall zones, the Rockfall
7 Analyst software is applied to identify the parameters of trajectory, frequency, height, and energy of the
8 slope landslide rockfalls. The modeling parameters of frequency, height, and energy are given various
9 weights, the sum for these parameters with different weights are conducted, and the hazard assessment for
10 the avalanche rockfalls in the Jinfeng Street is carried out. The accuracy of the rockfall model is verified by
11 comparing the simulation results with the actual rockfall accident, and the corresponding governance
12 measures for the rockfall hazard in the Jinfeng Street are proposed. The main conclusions obtained from
13 this study are as follows:
14 (1) Adopting the SAD method, combining with the rock outcrops identified from the statistical
15 topography map, the potential avalanche rockfall source zones are determined, and the area of the rockfall
16 source zone in the Jinfeng Street is calculated as 522,316 m2, which accounts for 0.8% of the total area of
17 the Jinfeng Street.
18 (2) The numerical simulation for the avalanche rockfalls in the Baiyan Village is carried out by using
19 the Rockfall Analyst software, and then the hazard assessment for the rockfalls in the whole Jinfeng Street
20 is conducted. The numerical simulation results indicate that there are 81,105 m2 of house area in the Jinfeng
21 Street within the rockfall hazard zone, with the proportion of high and ultra-high hazard areas accounting
22 for 25.6%.
23 (3) To reduce the hazards caused by the slope landslide rockfalls, according to the Code for Design of
1 Rock Retaining Wall Engineering in Geological Hazards (T/CAGHP 060-2019), a 5-meter-high rock
2 retaining wall is built, and the stability for such a rock retaining wall is verified. The results indicate that
3 after adding the rock retaining wall, the area of the rockfall hazard zone decreases to 2,580,371 m2, and
4 there is no ultra-high hazard zone. The house area in the rockfall hazard zone reduces to 50,432 m2, and the
5 house area in the high hazard zone reduces to 17.8% of the total one in the hazard zone.
6 (4) After adding the rock retaining wall, for the residents who are still in the high hazard zone, a
7 unified relocation strategy is recommended to ensure the safety of residents' lives and property.
8 Acknowledgments
9 This research was funded by the Science and Technology Infrastructure Program of Guizhou Province
10 (2020-4Y047), the Science and Technology Project of Transportation and Communication Ministry of
11 Guizhou Province (2023-312-030), the Opening Fund of Key Laboratory of Geohazard Prevention of Hilly
12 Mountains, Ministry of Natural Resources (KLGHZ202104), the Natural Science Foundation of Hunan
13 Province of China (2023JJ30135), and the Changsha Municipal Natural Science Foundation of China
14 (kq2208031).
15 Conflicts of interests
16 The authors declare no conflict of interest.
17 References
18 1 N. Matsuoka., H. Sakai. Rockfall activity from an alpine cliff during thawing periods. Geomorphology, 28
19 (1999) 309-328.
20 DOI:10.1016/S0169-555X(98)00116-0
21 2 O. Hungr, S. Leroueil, L. Picarelli. The Varnes classification of landslide types: an update. Landslides, 11
22 (2014) 167-194.
1 DOI:10.1007/s10346-013-0436-y
2 3 T. Chalachew, W. Kifle. Characteristics and susceptibility zonation of landslides in Wabe Shebelle Gorge,
3 south eastern Ethiopia. Journal of African Earth Sciences, 182 (2021) 104275.
4 DOI:10.1016/J.JAFREARSCI.2021.104275
5 4 J. Yan, J. Chen, C. Tan, S. Yan, Y. Liu, X. Zhao, Q. Wang. Rockfall source areas identification at local scale
6 by integrating discontinuity-based threshold slope angle and rockfall trajectory analyses. Engineering
7 Geology, 313 (2023) 106993.
8 DOI:10.1016/J.ENGGEO.2023.106993
9 5 C. Dussauge, J. Carine, A. Helmstetter. Statistical analysis of rockfall volume distributions: Implications for
10 rockfall dynamics. Journal of Geophysical Research: Solid Earth, 108 (2007) 2286.
11 DOI:10.1029/2001JB000650
12 6 P. Frattini, G. Crosta, A. Carrara, F. Agliardi. Assessment of rockfall susceptibility by integrating statistical
13 and physically-based approaches. Geomorphology, 94 (2008) 419-437.
14 DOI:10.1016/j.geomorph.2006.10.037.
15 7 M. Xia, B. Hai, N. Jiang, J. Chen, J Zhou. Risk assessment and mitigation evaluation for rockfall hazards at
16 the diversion tunnel inlet slope of Jinchuan Hydropower Station by using three-dimensional terrestrial
17 scanning technology. KSCE Journal of Civil Engineering, 27 (2022) 181-197.
18 DOI:10.1007/S12205-022-1679-8
19 8 X. Wang, H. Liu, J. Sun. A new approach for identification of potential rockfall source areas controlled by
20 rock mass strength at a regional scale. Remote Sensing, 13 (2021) 938.
21 DOI:10.3390/rs13050938
22 9 M. Alvioli, M. Santangelo, F. Fiorucci, M. Cardinali, I. Marchesini, P. Reichenbach, M. Rossi, F. Guzzetti,
23 S. Peruccacci. Rockfall susceptibility and network-ranked susceptibility along the Italian railway.
25 DOI:10.1016/J.ENGGEO.2021.106301
26 10 M. Cignetti, D. Godone, D. Bertolo, M. Paganone, P. Thuegaz, D. Giordan. Rockfall susceptibility along
27 the regional road network of Aosta Valley Region (northwestern Italy). Journal of Maps, 17 (2021) 54-64.
28 DOI:10.1080/17445647.2020.1850534
29 11 A. Abellan, J. Vilaplana, J. Martinez. Application of a long-range Terrestrial Laser Scanner to a detailed
30 rockfall study at Vallde Nuria (Eastern Pyrenees, Spain). Engineering Geology, 88 (2006) 136-148.
31 DOI:10.1016/j.enggeo.2006.09.012
1 12 A.N.S.D. Albarelli, C.O. Mavrouli, P. Nyktas. Identification of potential rockfall sources using
2 UAV-derived point cloud. Bulletin of Engineering Geology and the Environment, 80 (2021) 1-23.
3 DOI:10.1007/S10064-021-02306-2
4 13 A.M. Fanos, B. Pradhan, A. Alamri, C.W. Lee. Machine learning-based and 3D kinematic models for
5 rockfall hazard assessment using LiDAR Data and GIS. Remote Sensing, 12 (2020) 1755.
6 DOI:10.3390/rs12111755
7 14 A. Loye, M. Jaboyedoff, A. Pedrazzini, Identification of potential rockfall source areas at a regional scale
8 using a DEM-based geomorphometric analysis. Natural Hazards and Earth System Science, 9 (2009)
9 1643-1653.
10 DOI:10.5194/nhess-9-1643-2009
11 15 X. Wang, P. Frattini, D. Stead, J. Sun, H. Liu, A. Valagussa, L. Li. Dynamic rockfall risk analysis.
13 DOI:10.1016/j.enggeo.2020.105622
14 16 J. Zhan, J. Chen, W. Zhang, X. Han, X. Sun, Y. Bao. Mass movements along a rapidly uplifting river valley:
15 An example from the upper Jinsha River, southeast margin of the Tibetan Plateau. Environ. Environmental
16 Earth Sciences, 77 (2018) 634.
17 DOI:10.1007/s12665-018-7825-4
18 17 F. Guzzetti, P. Reichenbach, G.F. Wieczorek. Rockfall hazard and risk assessment in the Yosemite Valley,
19 California, USA. Natural Hazards and Earth System Sciences, 3 (2003) 491.
20 DOI:10.5194/nhess-3-491-2003
21 18 J. Zhan, Z. Yu, Y. Lv, J. Peng, S. Song, Z. Yao. Rockfall Hazard Assessment in the Taihang Grand Canyon
22 Scenic Area Integrating Regional-Scale Identification of Potential Rockfall Sources. Remote Sensing, 14
23 (2022) 3021.
24 DOI:10.3390/RS14133021
25 19 F.Ferrari, A. Giacomini, K. Thoeni. Qualitative rockfall hazard assessment: A comprehensive review of
26 current practices. Rock Mechanics and Rock Engineering, 49 (2016) 2865-2922.
27 DOI:10.1007/s00603-016-0918-z
28 20 K.M. Ansari, M. Ahmad, R. Singh, T.N. Singh, Rockfall Hazard Rating System along SH-72: a case study
29 of Poladpur–Mahabaleshwar road (Western India), Maharashtra, India. Geomatics, Natural Hazards and
30 Risk, 7 (2015) 649-666.
31 DOI:10.1080/19475705.2014.1003416
1 21 S. William, A. Yonathan. Using Google Earth and Google Street View to rate rock slope hazards.
2 Environmental and Engineering Geoscience, 24 (2018) 237-250.
3 DOI:10.2113/eeg-1922
4 22 A.M. Fanos, B. Pradhan. Multi-scenario rockfall hazard assessment using LiDAR Data and GIS. Geotech.
5 Geotechnical and Geological Engineering, 34 (2016) 1375-1393.
6 DOI:10.1007/s10706-016-0049-z
7 23 W.D. Stevens, RocFall, a Tool for Probabilistic Analysis, Design of Remedial Measures and Prediction of
8 Rockfalls. Master Thesis University of Toronto, Toronto, Canada. (1998).
9 24 L. Li, H. Lan. Probabilistic modeling of rockfall trajectories: a review. Bulletin of Engineering Geology
10 and the Environment, 74 (2015) 1163-1176.
11 DOI:10.1007/s10064-015-0718-9
12 25 A. Radtke, D. Toe, F. Berger, S. Zerbe, F. Bourrier. Managing coppice forests for rockfall protection:
13 lessons from modeling. Annals of Forest Science, 71 (2014) 484-494.
14 DOI:10.1007/s13595-013-0339-z
15 26 L. Dorren, F. Berger. New approaches for 3D rockfall modelling with or without the effect of forest in
16 Rockyfor3D. In Proceedings of the EGU General Assembly, Vienna, Austria, 2-7 May (2010) 14811.
17 27 H. Lan, C.D. Martin, C.H. Lim. RockFall analyst: A GIS extension for three-dimensional and spatially
18 distributed rockfall hazard modeling. Computers and Geosciences, 33 (2006) 262-279.
19 DOI:10.1016/j.cageo.2006.05.013
20 28 H. Lan, C.D. Martin, C. Zhou, H. Chang. Rockfall hazard analysis using LiDAR and spatial modeling.
21 Geomorphology, 118 (2010) 213-223.
22 DOI:10.1016/j.geomorph.2010.01.002
23 29 A.N. Strahler. Equilibrium theory of erosional slopes approached by frequency distribution analysis: Part II.
24 American Journal of Science, 248 (1950) 800-814.
25 30 V. Cristina, C. Marta, P. Roberta, M. Andrea, M. Chiara, C. Graziano, N. Guido. An integrated approach to
26 investigate climate-driven rockfall occurrence in high alpine slopes: the Bessanese glacial basin, Western
27 Italian Alps. Journal of Mountain Science, 17 (2020) 2591-2610.
28 DOI:10.1007/s11629-020-6216-y
29
1 List of Tables
4 List of Figures
5 Fig. 1 Map and geographical location: (a-c) Geographical location of study area; (d) The avalanche
6 lo-cation in the Jinfeng Street.
7 Fig. 2 Google Earth history images of Baiyan Village: (a) Image in January 2014; (b) Image in Sep-tember
8 2014; (c) Image in December 2019; (d) Image in March 2022.
11 Fig. 5 The distributions of cliffs, steep slopes, foot slopes, and plains in the topographical feature map of
12 Jinfeng Street.
15 Fig. 8 Landslide map of Baiyan Village. (a) UAV view of the affected area; (b) Google Earth image of the
16 area.
17 Fig. 9 Rockfall simulation in Baiyan village. (a) rockfall accident simulation; (b) simulation with rock
18 retaining wall.
19 Fig. 10 Trajectory and velocity profile of the simulated collapse rockfall in Baiyan Village.
20 Fig. 11 The simulated results of the avalanche rockfalls in the Jinfeng Street: (a) rockfall hazard zone; (b)
21 rockfall trajectory; (c) rockfall frequency; (d) rockfall energy; (e) rockfall velocity; (f) rockfall height.
22 Fig. 12 The simulated results of the rockfall hazard area in detail: (a-d) corresponding to A-D Region in Fig.
1 11.
3 Fig. 14 The cumulative distribution of rockfall hazard zones in the absence of a rock retaining wall: (a) The
4 cumulative distribution of hazard area; (b) The cumulative distribution of house area in the rockfall
5 hazard zones.
6 Fig. 15 The Cumulative distribution of rockfall hazard zones in the presence of a rock retaining wall: (a)
7 The cumulative distribution of hazard area; (b) The cumulative distribution of house area in the
8 rockfall hazard zones.

Nhaz D 23 02247

Uploaded by

Copyright:

Available Formats

Nhaz D 23 02247

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nhaz D 23 02247

Uploaded by

Copyright:

Available Formats

Natural Hazards

Manuscript Number: NHAZ-D-23-02247

Article Type: Manuscript

Keywords: rockfall hazard assessment; governing strategy; landslide; terrain analysis;

2 on the high-resolution online images and terrain analysis

5 Xiangtan University, Xiangtan 411105, China;

7 3. College of Civil Engineering, Xiangtan University, Xiangtan 411105, China;

8 4. College of Civil Engineering, Hunan University, Changsha 410082, China;

10 Corresponding authors: zhaorongguo@xtu.edu.cn

8 is recommended to ensure the safety of residents’ lives and property.

18 have become the continuous concerned questions in the academic community.

16 geological patterns [14].

18 suddenly and without warning.

7 controlling the rockfall hazard are proposed.

8 2. Geographical location and geological environment

9 2.1 Geographical location

6 December 2019; (d) Image in March 2022.

13 3. Methodology for identification and assessment

16 trajectory of the potential destructive point needs to be analysed.

11 3.1 Identification of potential rockfall sources

6 frequency of slope angle distribution is written as

13 four geological patterns.

16 alluvial fans, debris flows, and landslide deposits.

5 Fig. 6 Gaussian distributions of slope angle for various geological patterns.

4 as a potential rockfall source.

7 rockfall resources are identified.

8 3.2 Assessment of rockfall hazards

11 (1) Constructing a 3D model for the avalanche rockfall;

18 coefficient, which remain constants during the simulation process.

19 The expression for calculating the rockfall trajectory can be written as

8 velocity components of rockfall are

9 VDip   RTVTrend , VN  RNVN

17 4. Results and discussion

18 4.1 Results of potential rockfall

10 Fig. 7 The potential rockfall sources in the Jinfeng Street.

11 4.2 Results of rockfall hazard assessment

18 the avalanche rockfall hazard.

6 Street are shown in Fig. 11.

7 (a) (b) (c)

8 (d) (e) (f)

20 observed from each of the rockfall hazard regions.

11 acceleration of gravity, and G is the weight of rockfall.

House area in hazard（m2）

6×104 Low hazard

7 but also to the regions described by all topographic features.

22 DEMs, for the general situations, the 12.5 m DEM is sufficient.

22 further studies are needed to conduct.

13 this study are as follows:

17 the Jinfeng Street.

16 The authors declare no conflict of interest.

7 Geology, 313 (2023) 106993.

12 6 P. Frattini, G. Crosta, A. Carrara, F. Agliardi. Assessment of rockfall susceptibility by integrating statistical

13 and physically-based approaches. Geomorphology, 94 (2008) 419-437.

17 scanning technology. KSCE Journal of Civil Engineering, 27 (2022) 181-197.

20 rock mass strength at a regional scale. Remote Sensing, 13 (2021) 938.

22 9 M. Alvioli, M. Santangelo, F. Fiorucci, M. Cardinali, I. Marchesini, P. Reichenbach, M. Rossi, F. Guzzetti,

24 Engineering Geology, 293 (2021) 106301.

26 10 M. Cignetti, D. Godone, D. Bertolo, M. Paganone, P. Thuegaz, D. Giordan. Rockfall susceptibility along

29 11 A. Abellan, J. Vilaplana, J. Martinez. Application of a long-range Terrestrial Laser Scanner to a detailed