Earthquake Research in China  2020, Vol. 34 Issue (1): 50-63     DOI: 10.19743/j.cnki.0891-4176.202001009
Distribution of Landslides in Baoshan City, Yunnan Province, China
GAO Yuxin1, XU Chong1,2, TIAN Yingying1, MA Siyuan1, SHEN Lingling3, LU Yongkun4, RAN Hongliu1     
1. Key Laboratory of Active Tectonics and Volcano, Institute of Geology, China Earthquake Administration, Beijing 100029, China;
2. Institute of Crustal Dynamics, China Earthquake Administration, Beijing 100085, China;
3. Beijing Meteorological Information Center, Beijing Meteorological Service, Beijing 100089, China;
4. Yunnan Earthquake Agency, Kunming 650224, China
Abstract: Using Google Earth software as a platform, this study has established an integrated database of both old and new landslides in Baoshan City, Yunnan Province, China, and analyzed their development characteristics together with distribution rules, respectively. Based on the results, a total of 2 427 landslides occurred in the study area, including 2 144 new landslides and 283 old landslides, with a total area of about 104.8 km2. The new landslides are mostly in small-scales with an area less than 10 000 m2, while the area of individual old landslide is mostly larger than 10 000 m2. By analyzing the relationship between the two types of landslides and eight impact factors (i.e., elevation, slope angle, slope aspect, slope position, lithology, fault, regional Peak Ground Acceleration (PGA), and average annual rainfall), the different individual influencing factors, distribution regularities and mechanisms of the two types of landslides are revealed. In detail, the main influencing factors of new landslides are elevation, slope angle, slope aspect, slope position, lithology, regional PGA and average annual rainfall, while the influencing factors of old landslides are mainly elevation, slope angle, and lithology. This study provides basic data and support for landslide assessment and further disaster reduction in Baoshan City. Besides, it also provides new constraints in deeply understanding the effect of different topographic and geological conditions, historical earthquakes, rainfall and other factors on the occurrence mechanisms of both new landslides and old landslides.
Key words: Baoshan City     Old landslides     New landslides     Landslide density     Distribution     Mechanism    

INTRODUCTION

Affected by several natural disasters such as earthquakes and rainfalls, landslide is a common type of geological hazards in southwestern China, especially in Yunnan Province. Normally, landslides have a wide range of distribution, high frequency of occurrence, and strong destructiveness, which seriously threaten people̓s life and property(Lan Hengxing et al., 2004; Tao Yun et al., 2009; Wang Xueliang et al., 2014; Xu Chong et al., 2014b; Zhang Jie et al., 2014). Baoshan City is located in the southwest of Yunnan Province, with a total area of about 19 000 km2, and the mountain area accounts for more than 90%. There are various types of geological disasters in Baoshan, including landslides, collapses, debris flows, ground subsidence, ground fissures, and unstable slopes. As a result, landslides have brought numerous disasters to the life and property of the local people. For example, on September 1st, 2010, a landslide in Hedong Village, Wama Township, Longyang District, Baoshan City left 8 deaths and 40 missing (Li Weiyue et al., 2017). According to the government report, in 2010, there were 1 825 geological disasters that directly affected human activities and geological environment, including 1 240 landslides, 76 rock falls, 145 debris flows, 335 unstable slopes, 21 ground collapses, and 8 ground fissures (Jian Tao and Duan Xuefeng, 2014). Landslide is one of the most developed disasters in Baoshan City. The studies of spatial distribution law, susceptibility assessment, relationship with precipitation, and the control measures of landslides in Baoshan City (Xu Shiguang et al., 2004; Zhou Songlin et al., 2010; Zhang Ying, 2015) have provided important information for understanding the development of landslide disasters and have been used as important references for mitigation of landslide disasters. However, due to the data and method limitations at that time, there may be significant omissions in existing landslide database covering the whole Baoshan City, leading to some discrepancies between the actual results and those from the previous studies.

Therefore, based on high-resolution optical satellite imagery in Google Earth and Digital Elevation Model (DEM), this study conducts the interpretation and database construction of old landslides and new landslides in Baoshan City and analyzes their relationship with the impact factors. The results show that 2 427 landslides, including 2 144 new landslides and 283 old landslides, occurred in the study area. The relationship analysis between influencing factors and landslides indicates that the two types of landslides have different distribution rules and influencing factors. This study provides scientific and technological support for a more comprehensive and detailed understanding of the development and distribution regularities of landslides in Baoshan City, as well as landslide disaster prevention and reduction.

1 GEOLOGICAL SETTINGS

The study area, Baoshan City, is located in the western part of Yunnan Province, the southern section of Hengduan Mountain (Fig. 1). The terrains and landforms in this region are complex. In general, the local elevation is high in the northwest and low in the southeast. The lowest altitude is 535 m, the highest altitude is 3 780.9 m, and the average altitude is approximately 1 800 m. The rivers are developed and lies mostly along active fault zones. The main streams are Lancang River and Nujiang River, both of which are international rivers. The annual temperature difference is small while the daily temperature difference is large, with an average annual temperature of 14℃ -17℃. Precipitation is abundant and unevenly distributed, with an annual rainfall of 700 mm -2 100 mm. In addition, there are numerous active fault zones in Baoshan City, an earthquake-prone area. In particular, the area is composed of three significant fault zones, including Lancangjiang fault zone in the east, Tengchong and Longling active fault zone in the west, and Shidian fault in the South and Nujiang fault in the north (Deng Qidong, 2007; Xu Xiwei et al., 2016). According to previous records, 53 earthquakes with M≥5.0 has occurred in Baoshan City, including 41 earthquakes with 5.0≤M < 6.0; 10 earthquakes with 6.0≤M < 7.0; 2 earthquakes with 7.0≤M < 8.0. The M7.3 and M7.4 Longling earthquakes occurred at 20 :23 and 22 :00 on May 29, 1976 (Fig. 1).

Fig. 1 Location of the study area (Baoshan City), active faults and historical earthquakes LCJF-Lancangjiang fault, TCF-Tengchong fault, LLF-Longling fault, SDF-Shidian fault, NJF-Nujiang fault, WBNJF-West branch of Nujiang fault
2 DATA AND METHODS

In recent years, Google Earth has been widely used in geoscience because of the public access of the three-dimensional terrain data and the multi-temporal high-resolution (~0.5 m) satellite images that cover most part of the world(Lisle R. J., 2006; Yu Le et al., 2012; Padarian J. et al., 2015; Boardman J., 2016; Liu Kai et al., 2018). It has provided great convenience for landslide interpretation (Sato H. P. et al., 2009; Gorum T. et al., 2013; Li W. L. et al., 2013; Xu Chong et al., 2014a; Xu Chong, 2015).Additionally, Google Earth gives us the opportunity to do visual interpretation and landslide mapping. High-resolution satellite images in multiple periods cover the entire study area. In this study, the acquisition time of the satellite images we used ranges from November 2017 to present. We identify each individual landslides, delineate their boundaries and mark the positions using professional knowledge. Landslides in the study area are generally divided into old landslides and new landslides. In detail, old landslides are mainly identified by the topographic features. This type of landslides are mostly deep-seated landslides that have occurred for a long period and remain clear landslide topographic features (Fig. 2(a)), such as ring-shaped landslide back wall, abnormally curved ridgelines, steep landslide back wall, relatively flat landslide accumulation area, and residential area on the slope bodies and terraces. Because small- and medium-scales old landslides are unrecognizable, most of them that can be identified are large-scale old landslides. Considering the scales, the landslides are more likely to be triggered by earthquakes (Xu Chong et al., 2014c). They often occurred a long period, and some may have existed before the historical earthquake records.New landslides are mainly identified by the differences in color and texture compared with the surrounding environment from the satellite image (Fig. 2(b)). The vegetation coverage of the landslide body and the surrounding environment is significantly different. Considering the climatic characteristics and vegetation restoration capacity of the area, it can be inferred that such landslides have occurred in recent years. Considering the recent weak seismic activities in the area, it can be suggested that the new landslides are more likely to be triggered by rainfall events. Fig. 2 shows a group of satellite images of old and new landslides. In detail, Fig. 2(a) shows two old landslides. It can be seen that the landslide body is not much different from the surrounding environment in image color, but there are obvious chair- and circle-like landslide features and a steeper back wall. Fig. 2(b) shows a group of new landslides. The satellite images of the bare landslide body and the surrounding vegetation are extremely different.

Fig. 2 Landslides shown on the satellite images of Google Earth platform (a) two old landslides located at 24.717°N, 99.1352°E, view to the east, taken on January 11th, 2018; (b)a group of new landslides located at 24.1814°N, 98.7557°E, view to the north, and taken on January 23rd, 2018

Factors affecting the occurrence of landslides include terrain, geological setting, earthquake, rainfall, etc. The slope angle is one of the most important factors affecting the occurrence of landslides because the slope angle determines the effective frontier of a landslide. Generally, with other conditions being similar, landslide may occur at a larger possibility when the slope angle increases. Besides, factors such as elevation, slope aspect, and slope position also affect the occurrence of landslides in different ways. Among the geological factors, the underlying stratum is important in the occurrence of landslides, because it is an essential material condition for landslides. The susceptibility of landslides in different underlying strata areas often varies greatly. Active faults often have a weakening effect on the rock masses on both sides and thus affect the occurrence of landslides. Regional PGA reflects the intensity of historical earthquakes and the hazard of future earthquakes, and is also an important landslide factor. Rainfall affects the occurrence of landslides by changing the strength of rock and soil and loading of the slope Therefore, in this study, eight factors including elevation, slope angle, slope aspect, slope position, lithology, fault, PGA and average annual rainfall are selected to analyze the relationship between old and new landslides and these factors. The elevation data is derived from SRTM DEM data with a resolution of 1 arc second. After projection, it is converted into a DEM data with a resolution of 20 m. Using this DEM data, the slope angle, slope aspect, and slope position data of the study area are produced. The lithological data of the study area is from 1:500 000 geological maps of China. Active fault data are from Active Tectonic Map of China(Deng Qidong, 2007; Xu Xiwei. et al., 2016). Regional PGA data are from the Seismic Ground Motion Parameter Zoning Map of China (GB 18306—2015) (General Administration of Quality Supervision, Inspection and Quarantine of the People̓s Republic of China and China National Standardization Administration, 2016). Rainfall data comes from the interpolation of 28 rainfall stations in and around the study area. In ArcGIS software, these factors are classified, the density of landslide points within each classification is counted, and the relationship between landslide and these factors is analyzed.

3 RESULTS AND ANALYSIS 3.1 Landslide Distribution

The results show that 2 427 landslides have occurred in the 19 050 km2 study area, including 2 144 new landslides and 283 old landslides (Fig. 3). The total area of these landslides is about 104.8 km2, the area of the largest landslide reaches 4.8 km2, the area of the smallest landslide is only 50 m2, and the average area of the landslide is about 43 000 m2. The landslide point density of the study area is 0.127 km-2, and the area density is 0.55%. The landslide distribution is shown in Fig. 4. The results also show that there are 17 landslides with an area greater than 1 km2; 169 landslides with an area between 100 000 m2 and 1 km2; 229 landslides with an area between 10 000 m2 and 100 000 m2; 1 175 landslides with an area between 1 000 m2 and 10 000 m2; and 826 landslides with an area between 100 m2 and 1 000 m2. The remaining 11 landslides have an area less than 100 m2. The scale distribution of old landslides and new landslides shows a clear difference. The number of new landslides with an area less than 10 000 m2 is 2 000, accounting for 93.3% of the total number of recent landslides. For the ancient landslides, there are 271 landslides with an individual area of more than 10 000 m2, accounting for 95.8% of the total number of old landslides.

Fig. 3 Landslide distribution map of the study area

Fig. 4 Distribution of landslide number in different landslide area class
3.2 Analysis of Landslide Influencing Factors

The elevation range in the study area is 534 m-3 758 m, divided into 12 classes at a 200-meter interval. As the elevation increases, the landslide distribution first increases and then decreases (Fig. 5). In general, the density of new landslides tends to increase as the elevation increases and the maximum landslide density (0.196 km-2) appears in the interval where the elevation is larger than 3 000 m. The old landslide point density generally exhibits a decreasing trend, and the maximum point density (0.057 km-2) is in the elevation range of 1 000 m-1 200 m (Fig. 5). This may be associated with the low recoverability in high altitudes. Besides, new landslides occurring in these areas are more easily identified, while new landslides occurring in low altitude areas are not easily identified due to rapid vegetation recovery. Whereas the identified old landslides are generally large-scales, and the positioning point of a landslide is located in its center so that the locations of old landslides are relatively low. In general, although the relationship between old landslides and new landslides and elevations shows certain regularity, the mechanisms still need to be further discussed and studied.

Fig. 5 Relationship between elevation and landslides

Fig. 6 shows the relationship between slope angle and density of old landslides and new landslides in the study area. Most of the landslides have a slope angle below 30°, indicating that the slopes in the study area are mainly gentle. In general, the density of old landslides and new landslides increase with slope increasing. The maximum density of new landslides (0.227 km-2) is in the range of slopes 35°-40°, and the maximum density of old landslides (0.054 km-2) is in the slope angle >45° (Fig. 6). The distribution of the density of the two types of landslides shows that the landslides are more likely to occur in the areas with a larger slope angle. The density of new landslides in the areas with a slope greater than 40° shows a downward trend, which perhaps because of the small area of these classes and thus has less statistical value. In general, the relationship between slope angle and the occurrence of landslides is relatively clear: higher slope renders larger possibility of the landslide occurrence.

Fig. 6 Relationship between slope angle and landslides

Fig. 7 shows the relationship between the slope aspect and landslide density. The results indicate that the densities of the new landslides in the three directions of SE, S, and SW are much higher than those in other classes, whereas the old landslides do not have this trend. This is because the rainfall in the area is mainly affected by the water vapor in the Indian Ocean. Therefore, the south and adjacent slope aspects receive more rainfall than other directions. During the process of rainfall softening and loading, these slopes may slide more easily. On the other hand, the three slopes receive more sunlight than other slopes, so the weathering effect may be more significant and the landslide may be more susceptible.

Fig. 7 Relationship between slope aspect and landslides

Fig. 8 shows the relationship between slope position and landslide density. According to the principles proposed by Weiss(Weiss A.D., 2001) and Jenness (Jenness J. et al., 2013), the study area is divided into ridge, upper slope, middle slope, flat slope, lower slope, and valley. It can be seen from Fig. 8 that the density of new landslides in the valley area is the largest (0.224 km-2), and the density of old landslides in the middle slope area is the largest (0.022 km-2). Combining the relationship between landslide and elevation, it can be seen that new landslides are more likely to occur in high-altitude valley areas, whereas old landslides are more likely to occur in middle slope with low-altitude areas.

Fig. 8 Relationship between slope position and landslides

According to the underlying stratigraphic age and lithology, the study area is divided into 19 categories, including (1) Magmatic rocks; (2) Lower Proterozoic granulite and gneisses; (3) Sinian sandstones and plate rocks; (4) Upper Cambrian shale, limestone; (5) Ordovician shale, sandstone; (6) Silurian gray, fuchsia limestone; (7) Lower Devonian fuchsia, gray grey (8) Middle Carboniferous variegated rocks and glutenite; (9) Upper Carboniferous limestone andargillaceous limestone with calcareous shale; (10) Lower Permian shale, conglomerate, and glutenite; (11) Middle Permian limestone; (12) Upper Permian shale, siltstone intercalated limestone; (13) Lower Triassic thin layer limestone, sandstone; (14) Upper Triassic dolomite; (15) Middle Jurassic fuchsia, gray conglomerate, sandstone; (16) Upper Jurassic fuchsia mudstone; (17) Lower Cretaceous gray, grayish-yellow, and fuchsia limestone; (18) Tertiary sandstone Clay rocks; (19) Quaternary alluvium. The statistical results in Fig. 9 indicate that the susceptibilities of old landslides and new landslides are different. New landslides are more likely to occur in magmatic rocks (1) and Lower Permian shale, conglomerate, and glutenite rocks (10), while old landslides are more likely to occur in the Upper Permian shale, siltstone intercalated limestone rocks (12) and Middle Jurassic fuchsia, gray conglomerate, and sandstone rocks (15). In general, old landslides are more likely to occur in younger strata, while new landslides are more likely to occur in older strata. The possible reason is that the new age strata is tectonically weak and is easy to produce large-scale deep-seated landslides, while the old strata is relatively strong, and long-term weathering may only cause more shallow rainfall-induced landslides.

Fig. 9 Relationship between lithology and landslides

Fig. 10 shows the statistical results of landslide density at different active fault distances in the study area. There is no obvious correspondence between landslide density and active fault distance. This may be due to the fact that the main influencing factor of these landslides is not active faults, which are obtained from relatively low-scale data across China, or are less studied and identified.

Fig. 10 Relationship between active fault distance and landslides

Fig. 11 shows the relationship between the regional PGA and landslides. There are four levels of regional PGA in the study area, which are 0.1 g, 0.15 g, 0.2 g, and 0.3 g, respectively. The PGA in most parts of the study area is 0.2 g. In general, the density of new landslides increases with the increasing PGA, while the density of ancient landslides decreases with the increasing PGA. The maximum density of recent landslides (0.189 km-2) appears in the area with a PGA of 0.3 g, and the maximum density of ancient landslides (0.025 km-2) appears in the area with a PGA of 0.1 g. The regional PGA used in this study comes from the Seismic Ground Motion Parameter Zoning Map of China (GB 18306-2015) (General Administration of Quality Supervision, Inspection and Quarantine of the People̓s Republic of China and China National Standardization Administration, 2016), and this map is mainly based on historical seismic records to determine the hazard of future earthquakes. This is undoubtedly affected by the quality of historical earthquake data records. According to the historical records, 53 historical earthquakes with M≥5.0 earthquakes are recorded in this area, of which only 2 earthquakes with M≥7.0, and the distribution of these major earthquakes does not show spatial correlation with the distribution of old landslides (Figs. 1 and 3). Furthermore, there is no correlation between the old landslides and regional PGA, and it even shows the opposite trend, suggesting most of the earthquake events that triggered these old landslides may not have been recorded. The historical earthquake records in the study area are relatively short. The earliest recorded earthquake is the Tengchong M6 3/4 earthquake on October 18th, 1512. Therefore, there is no correlation between old landslides and historical earthquake records and regional PGA.

Fig. 11 Relationship between regional PGA and landslides

The new landslides and regional PGA show a clear correlation (Fig. 11), which is mainly due to the existence of post-earthquake effects, and after time, the post-effects gradually weaken, and may even restore back to the pre-seismic level. Historical earthquake effects are more significant than those events which have not been recorded and thus lead to more susceptible landslides. Even though these new landslides are more likely to be triggered by rainfall, there is a clear correlation with historical earthquakes, that is, the regional PGA. This phenomenon reminds us that the potential seismic hazards in this area are not consistent with historical earthquake records. Thus, more investigations are needed to reveal the relationship between active faults and paleoearthquakes in this area; the post-earthquake effects may last for hundreds of years and are reflected by the susceptibility of rainfall-induced landslides.

The average annual rainfall in the study area gradually increases from northeast to southwest and the area is divided into 6 categories based on the average annual rainfall (Fig. 12).The result suggests that the density of new landslides generally increases with the increase of rainfall, however, no similar trends appear among old landslides, indirectly illustrating the rationality of our analysis of the trigger factors of the new landslides and old landslides. New landslides are mainly triggered by rainfall events, while old landslides may be triggered by other factors, such as earthquake events.

Fig. 12 Relationship between average annual rainfall and landslides
4 CONCLUSIONS

This study investigates the development of landslides in Baoshan City, Yunnan Province and analyzes their spatial distribution. The results show that 2 427 landslides have developed in Baoshan, including 2 144 new landslides and 283 old landslides, with a total area of about 104.8 km2. Most new landslides are small-scaled, presumably triggered by rainfall in recent years; however, old landslides are large-scaled, presumably triggered by earthquake events. By analyzing the relationship between landslide densities and influencing factors, we find that these landslides and slope angles generally exhibit a positive correlation. In detail, new landslides are more likely to occur on the three slope aspects of SE, S, and SW but old landslides do not have this trend. In addition, rainfall-triggered landslides are more likely to occur in low-altitude areas at high altitudes, whereas old landslides are more likely to occur in low-altitude areas. Older, stronger strata are more prone to new landslides, while younger, weaker strata are more prone to old landslides. Due to the post-earthquake effects of the historical earthquakes and the short recording time of historical earthquakes, there is a positive correlation between new landslides and the regional PGA, while old landslides have a negative correlation with the regional PGA. These results indicate that the main influencing factors of new landslides are elevation, slope angle, slope aspect, slope position, lithology, regional PGA, and annual average rainfall; while the main influencing factors of ancient landslides are only elevation, slope angle, and lithology. Furthermore, even with the same factors, the mechanism of new landslides and old landslides are quite different. This can reminds us that there are obvious differences in the distribution regularities and mechanisms between rainfall-triggered landslides and earthquake-triggered landslides. The study of landslides in this area also reminds us that there are several limitations in the regional PGA data based on historical earthquakes, especially in areas where the historical seismic records and active tectonic studies are insufficient.

REFERENCES
Boardman JBoardman J. The value of Google EarthTM for erosion mapping[J]. CATENA, 2016, 143: 123-127. DOI:10.1016/j.catena.2016.03.031
Deng Qidong. Map of Active Tectonics in China[M]. Beijing: Seismogical Press (in Chinese).
General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China, China National Standardization Administration. GB 18306-2015 Seismic ground motion parameters zonation map of China[S]. Beijing: China Standards Press, 2016 (in Chinese).
Gorum T., van Westen C.J., Korup O., van der Meijde M., Fan Xuanmei, van der Meer F.DGorum T., van Westen C.J., Korup O., van der Meijde M., Fan Xuanmei, van der Meer F.D. Complex rupture mechanism and topography control symmetry of mass-wasting pattern, 2010 Haiti earthquake[J]. Geomorphology, 2013, 184: 127-138. DOI:10.1016/j.geomorph.2012.11.027
Jenness J., Brost B., Beier P. Land facet corridor designer: Topographic position index tools[EB/OL]. (2013-07-09). www.jennessent.com.
Jian Tao, Duan XuefengJian Tao, Duan Xuefeng. Spatial distribution of geological hazards and their susceptibilityzonation and evaluation in Baoshan, Yunnan[J]. Super Science, 2014, 3(214): 215 (in Chinese).
Lan Hengxing, Zhou C.H., Wang L.J., Zhang H.Y., Li R.HLan Hengxing, Zhou C.H., Wang L.J., Zhang H.Y., Li R.H. Landslide hazard spatial analysis and prediction using GIS in the Xiaojiang watershed, Yunnan, China[J]. Engineering Geology, 2004, 76(1/2): 109-128.
Li Weiyue, Liu Chun, Scaioni M., Sun Weiwei, Chen Yu, Yao Dongjing, Chen Sheng, Hong Yang, Zhang Kaihang, Cheng GuodongLi Weiyue, Liu Chun, Scaioni M., Sun Weiwei, Chen Yu, Yao Dongjing, Chen Sheng, Hong Yang, Zhang Kaihang, Cheng Guodong. Spatio-temporal analysis and simulation on shallow rainfall-induced landslides in China using landslide susceptibility dynamics and rainfall I-D thresholds[J]. Science China Earth Sciences, 2017, 60(4): 720-732. DOI:10.1007/s11430-016-9008-4
Li W.L., Huang R.Q., Pei X.J., Zhang X.C. Historical co-seismic landslide inventory with Google Earth: A case study of 1920 Haiyuan Earthquake, China. In: Wu Faquan, Shengwen Qi (Editors). Global View of Engineering Geology and the Environment. Boca Raton: CRCPress, 2013: 179-184.
Lisle R.JLisle R.J. Google Earth:a new geological resource[J]. Geology Today, 2006, 22(1): 29-32.
Liu Kai, Ding Hu, Tang Guoan, Song Chunqiao, Liu Yiwen, Jiang Ling, Zhao Bangyuan, Gao Yunfei, Ma RonghuaLiu Kai, Ding Hu, Tang Guoan, Song Chunqiao, Liu Yiwen, Jiang Ling, Zhao Bangyuan, Gao Yunfei, Ma Ronghua. Large-scale mapping of gully-affected areas:An approach integrating Google Earth images and terrain skeleton information[J]. Geomorphology, 2018, 314: 13-26. DOI:10.1016/j.geomorph.2018.04.011
Padarian J., Minasny B., Malone B., McBratney A.BPadarian J., Minasny B., Malone B., McBratney A.B. Digital soil mapping in the cloud using Google Earth Engine[J]. Computers & Geosciences, 2015, 83: 80-88.
Sato H.P., Harp E.L.Sato H.P., Harp E.L. Interpretation of earthquake-induced landslides triggered by the 12 May M7.9 Wenchuan earthquake in the Beichuan area, Sichuan Province, China using satellite imagery and Google Earth[J]. Landslides, 2009, 6(2): 153-159.
Tao Yun, Tang Chuan, Duan XuTao Yun, Tang Chuan, Duan Xu. Landslide and debris flow hazards in Yunnan and their relationship with precipitation characteristics[J]. Journal of Natural Disasters, 2009, 18(1): 180-186 (in Chinese with English abstract).
Wang Xueliang, Zhang Luqing, Wang Sijing, Lari SWang Xueliang, Zhang Luqing, Wang Sijing, Lari S. Regional landslide susceptibility zoning with considering the aggregation of landslide points and the weights of factors[J]. Landslides, 2014, 11(3): 399-409. DOI:10.1007/s10346-013-0392-6
Weiss A.D. Topographic position and landforms analysis[EB/OL]. (2001). http://www.jennessent.com/downloads/tpi-poster-tnc_18x22.pdf.
Xu Shiguang, Zhang Zhigang, Guo Yuanming, Cui Yali, Zhao Jinfa, Peng Guangyue, Hou LiXu Shiguang, Zhang Zhigang, Guo Yuanming, Cui Yali, Zhao Jinfa, Peng Guangyue, Hou Li. Cause and control of the landslide in Baoshan Teachers' College in Yunnan Province[J]. The Chinese Journal of Geological Hazard and Control, 2004, 15(3): 6-10 (in Chinese with English abstract).
Xu Chong, Shyu J.B.H., Xu XiweiXu Chong, Shyu J.B.H., Xu Xiwei. Landslides triggered by the 12 January 2010 Port-au-Prince, Haiti, MW=7.0 earthquake:visual interpretation, inventory compiling, and spatial distribution statistical analysis[J]. Natural Hazards and Earth System Sciences, 2014a, 14(7): 1789-1818.
Xu Chong, Xu Xiwei, Shen Lingling, Dou Shuai, Wu Saier, Tian Yingying, Li XiXu Chong, Xu Xiwei, Shen Lingling, Dou Shuai, Wu Saier, Tian Yingying, Li Xi. Inventory of landslides triggered by the 2014 MS6.5 Ludian earthquake and its implications on several earthquake parameters[J]. Seismology and Geology, 2014b, 36(4): 1186-1203 (in Chinese with English abstract).
Xu Chong, Xu Xiwei, Yao Xin, Dai FuchuXu Chong, Xu Xiwei, Yao Xin, Dai Fuchu. Three (nearly) complete inventories of landslides triggered by the May 12, 2008 Wenchuan MW7.9 earthquake of China and their spatial distribution statistical analysis[J]. Landslides, 2014c, 11(3): 441-461. DOI:10.1007/s10346-013-0404-6
Xu ChongXu Chong. Preparation of earthquake-triggered landslide inventory maps using remote sensing and GIS technologies:Principles and case studies[J]. Geoscience Frontiers, 2015, 6(6): 825-836. DOI:10.1016/j.gsf.2014.03.004
Xu Xiwei, Han Zhujun, Yang Xiaoping, Zhang Shimin, Yu Guihua, Zhou Bengang, Li Feng, Ma Baoqi, Chen Guihua, Ran Yongkang. Seismotectonic Map in China and Its Adjacent Regions[M]. Beijing: Seismogical Press, 2016 (in Chinese).
Yu Le, Gong PengYu Le, Gong Peng. Google Earth as a virtual globe tool for Earth science applications at the global scale:progress and perspectives[J]. International Journal of Remote Sensing, 2012, 33(12): 3966-3986. DOI:10.1080/01431161.2011.636081
Zhang Jie, Wang Yu, Zhang Hongbing, Zhu Chunlin, Li Jundong, Yan XiangshengZhang Jie, Wang Yu, Zhang Hongbing, Zhu Chunlin, Li Jundong, Yan Xiangsheng. Characteristics of secondary geological hazards induced by Yiliang 9·07 earthquakes in Yunnan Province[J]. Journal of Engineering Geology, 2014, 22(2): 280-291 (in Chinese with English abstract).
Zhang Ying. Research on the development characteristics and countermeasures of highway geological hazards in basin-ridge staggered area of Western Yunnan: A case study from Baoshan to Pupiao section of G320 national road[D]. Kunming: Kunming University of Science and Technology, 2015 (in Chinese with English abstract).
Zhou Songlin, Tan Shucheng, Jiang Shunde, Li JunZhou Songlin, Tan Shucheng, Jiang Shunde, Li Jun. The spatial distribution, prone zoning and evaluation of geologic hazards in Longyang District, Baoshan City, Yunnan Province[J]. Earth and Environment, 2010, 38(3): 351-356 (in Chinese with English abstract).