Next Article in Journal
Enhancing the Performance of Landslide Susceptibility Mapping with Frequency Ratio and Gaussian Mixture Model
Previous Article in Journal
Construction of Ecological Security Patterns and Evaluation of Ecological Network Stability under Multi-Scenario Simulation: A Case Study in Desert–Oasis Area of the Yellow River Basin, China
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Land
Volume 13
Issue 7
10.3390/land13071038
land-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Landslide Distribution and Development Characteristics in the Beiluo River Basin

by
Fan Liu
1,2,3,
Yahong Deng
1,2,*,
Tianyu Zhang
4,
Faqiao Qian
1,2,
Nan Yang
1,2,
Hongquan Teng
3,
Wei Shi
3 and
Xue Han
3
1
School of Geological Engineering and Geomatics, Chang’an University, Xi’an 710054, China
2
Key Laboratory of Mine Geological Hazards Mechanism and Control, Ministry of Natural Resources, Xi’an 710054, China
3
Shaanxi Hygrogeology Engineering Environement Geology Survey Center, Xi’an 710068, China
4
Shaanxi Institute of Geological Survey, Xi’an 710068, China
*
Author to whom correspondence should be addressed.
Land 2024, 13(7), 1038; https://doi.org/10.3390/land13071038
Submission received: 18 June 2024 / Revised: 5 July 2024 / Accepted: 7 July 2024 / Published: 10 July 2024
(This article belongs to the Topic Landslides and Natural Resources)
Browse Figures
Versions Notes

Abstract

:
The Beiluo River Basin, situated in the central region of the Loess Plateau, frequently experiences landslide geological disasters, posing a severe threat to local lives and property. Thus, establishing a detailed database of historical landslides and analyzing and revealing their development characteristics are of paramount importance for providing a foundation for geological hazard risk assessment. First, in this study, landslides in the Beiluo River Basin are interpreted using Google Earth and ZY-3 high-resolution satellite imagery. Combined with a historical landslide inventory and field investigations, a landslide database for the Beiluo River Basin is compiled, containing a total of 1781 landslides. Based on this, the geometric and spatial characteristics of the landslides are analyzed, and the relationships between the different types of landslides and landslide scale, stream order, and geomorphological types are further explored. The results show that 50.05% of the landslides have a slope aspect between 225° and 360°, 68.78% have a slope gradient of 16–25°, and 38.97% are primarily linear in profile morphology. Areas with a high landslide density within a 10 km radius are mainly concentrated in the loess ridge and hillock landform region between Wuqi and Zhidan Counties and in the loess tableland region between Fu and Luochuan Counties, with a significant clustering effect observed in the Fu County area. Loess–bedrock interface landslides are relatively numerous in the northern loess ridge and hillock landform region due to riverbed incision and the smaller thickness of loess in this area. Intra-loess landslides are primarily found in the southern loess tableland region due to headward erosion and the greater thickness of loess in this area. Loess–clay interface landslides, influenced by riverbed incision and the limited exposure of red clay, are mainly distributed in the northern part of the southern loess tableland region and on both sides of the Beiluo River Valley in Ganquan County. These results will aid in further understanding the development and spatial distribution of landslides in the Beiluo River Basin and provide crucial support for subsequent landslide susceptibility mapping and geological hazard assessment in the region.

1. Introduction

Landslides are a prevalent geological hazard, causing significant casualties and property damage annually [1,2]. Detailed lists of landslide data are helpful to analyze the spatial distribution of landslides [3,4,5,6] and their impact on landforms [7,8,9,10]. At the same time, landslide data lists are the basis for the evaluation of landslide sensitivity, hazard, and risk in specific areas [11,12,13,14,15,16], and they provide suggestions and a basis for scientific research, land planning, engineering construction, and emergency management [17,18,19,20].
Many scholars have established lists of landslides in different regions. For instance, Zemin Gao [21] evaluated the hazard in Southwest China based on 13,886 historical landslides. Similarly, Tianjun Qi [22] modeled and analyzed correlations in the Western Qinling Mountains in China using 2765 historical landslides. In Europe, 22 out of 37 countries have developed their own landslide inventories [16]. Xiaobo Li et al. [23] compiled an inventory of loess landslides induced by the Ms 8.0 earthquake in Tianshui, China, in 1654. Iris Bostjančić et al. [24] conducted a landslide survey in Slavonski Brod, located in northeastern Croatia, covering an area of 55.1 square kilometers, resulting in an inventory of 854 landslides. Gómez, D et al. [15] established a global landslide catalog with 37,946 entries, covering 161 countries and regions, and they analyzed the triggering factors of landslides. Rosser, B [25] used remote sensing interpretation, field surveys, and media accounts to compile a record of 22,575 landslides in New Zealand. Damm, B [26] developed the German landslide database, which includes data on over 4200 landslides. Sun, JJ [27] developed a landslide database for the Yinshan region, recording a total of 10,968 landslides. Peng et al. [28] compiled a landslide distribution map for the Loess Plateau, encompassing 14,544 landslides.
The method of establishing a landslide database is commonly used in news reports and newspapers [29]. This method is simple and efficient, and a landslide database can initially be established as the basis for remote sensing and field investigations. Remote sensing visual interpretation [30,31,32] is one of the commonly used methods to establish a landslide database. With the development of satellite technology, multispectral images are widely used, especially Google Earth; this provides not only high-resolution satellite images but also a three-dimensional visual model, which can allow researchers to make a more intuitive landslide observation. A field investigation [31] is essential, being the most traditional means of landslide investigation. A field investigation can provide more detailed landslide parameters and is a necessary method to improve the database.
The Beiluo River Basin, located in the heart of the Loess Plateau, experiences severe soil erosion, and it is one of the area’s most critically affected by water and soil loss, fragile ecological environments, and frequent geological disasters globally [33,34,35,36,37]. It is a typical landslide-prone area. To date, the landslides in this region have not been classified, and detailed statistics on them have not been obtained. To address this gap, this study utilizes high-resolution satellite imagery from Google Earth and ZY-3 satellite images taken in China [38,39,40], combined with a historical landslide inventory and field investigations, to create a landslide map of the Beiluo River Basin and establish a landslide database. This database includes information on the morphology, distribution characteristics, and geological features of the landslides. Finally, a statistical analysis of the characteristics and distribution patterns of the landslides in the Beiluo River Basin is conducted to characterize the development trends, clustering features, and spatial distribution patterns of the different types of landslides in the region. The research results provide a basis for the early identification and prevention of geological disasters in the Beiluo River Basin and offer guidance for further geological hazard risk assessment.

2. Study Area

The Beiluo River Basin is located in the central Loess Plateau. Since the Cenozoic era, due to the influence of oscillatory neotectonic movements characterized primarily by uplift, combined with continuous loess erosion and deposition, the geomorphology of the region has evolved into one marked by deeply incised gullies and a fragmented terrain. Initially, the deposition rate exceeded the erosion rate, but, in later stages, the erosion rate far surpassed the deposition rate [41]. The topography of the Beiluo River Basin is higher in the northwest and lower in the southeast (Figure 1), with an elevation ranging from 289 m to 1896 m a.s.l.
The study area lies between the 400 mm and 800 mm isohyets, falling within a warm temperate semi-humid, semi-arid continental monsoon climate zone. It is influenced by the East Asian summer monsoon and the Siberian winter monsoon, leading to distinct seasonal climatic variations. The average annual temperatures range from 7.7 °C to 10.6 °C, with the annual average maximum temperature being 17.2 °C and the annual average minimum temperature being 4.3 °C. July is the hottest month, with an average temperature of 23.1 °C, while January is the coldest month, with an average temperature of −5.5 °C. The extreme maximum temperature recorded is 38.3 °C (on 21 July 2000), and the extreme minimum temperature recorded is −23.0 °C (on 28 December 1991) [42]. The average annual rainfall is approximately 500 mm [43], decreasing gradually from south to north, with most precipitation occurring from June to September, peaking in August. The region often experiences continuous rain and heavy storms during this period. The maximum daily rainfall recorded is 139.9 mm (on 15 August 1981), the highest annual rainfall recorded is 774 mm (in 1981), and the lowest annual rainfall recorded is only 330 mm (in 1974).
Stratigraphy and lithology are intrinsic factors influencing geological hazards, reflected in the rock types and their characteristics, as well as the interface relationships between different geological periods. The main exposed strata in the Beiluo River Basin include the Quaternary (Qp1–Qh), Neogene (N), Cretaceous (K), Jurassic (J), and Triassic (T) [44]. The Quaternary and Neogene strata primarily consist of aeolian loess and clay, while the Neogene, Cretaceous, Jurassic, and Triassic strata are mainly composed of sandstone and mudstone, and they are distributed along river systems (Figure 2).
Based on the presence or absence of firm bonding between soil and rock particles, the masses in the region are divided into two categories: rock masses and soil masses. According to construction type, structural type, and rock strength, rock masses are classified into hard to moderately hard thick-bedded blocky clastic rock, interbedded soft and hard coal-bearing (oil shale) clastic rock, and soft layered clay rock. According to engineering geological properties, soil masses are primarily classified into loess and sandy gravel soils.
Human engineering activities are one of the main factors causing slope instability and inducing collapses, landslides, and other hazards [45]. Located in the central Loess Plateau, the Beiluo River Basin’s industrial and economic structure was originally dominated by agriculture. With the development of petroleum and other mineral resources, related engineering activities have significantly increased. Additionally, urban construction and infrastructure development activities have rapidly progressed. The main economic activities in the basin include agriculture, forestry, animal husbandry, urban and rural construction, infrastructure construction (such as roads and water conservancy), and mineral resource development. A large number of landslides seriously threaten the safety of life and property. Therefore, it is necessary to establish a comprehensive landslide database to provide a basis for disaster prevention and mitigation.

3. Material and Methods

3.1. Material

Landslide data were obtained through remote sensing interpretation, field surveys, and historical landslide inventories. The remote sensing interpretation method is based on a visual interpretation, where landslides exhibit clear armchair-like characteristics on their back walls, with distinctly curved ridgelines. In remote sensing images, they typically appear as tongue-shaped or amphitheater-like features, with the landslide body displaying irregular step-like distributions [46]. The remote sensing interpretation was primarily conducted using the Google Earth platform, which is widely used in landslide investigations [46,47,48]. The resolution and coverage of Google Earth meet the requirements for identifying landslides of a large range (1:10,000). A field investigation was carried out through drones and manual visits (Investigation date: April 2022–November 2022). A list of historical landslides was mainly derived from the geological disaster report of the Shaanxi Provincial Geological Survey Institute, in which the location, depth, and scale of landslides are recorded in detail. Figure 3 shows four examples of large landslides identified within the study area.

3.2. Methods

3.2.1. Investigation Method

Landslide scale is mainly divided into three categories: medium-scale landslides (10–100 × 104 m3), large landslides (100–1000 × 104 m3), and small landslides (0–10 × 104 m3). Landslide age is divided into old landslides (occurred since the Holocene), new landslides (currently active), and ancient landslides (occurred before the Holocene). Landslide types are divided into loess–bedrock interface landslides, intra-loess landslides, and loess–clay interface landslides. The section shape is divided into foveation, convex, and rectilinear shapes. The slope is based on DEM (Digital Elevation Model), with a resolution of 12.5 m, and it is obtained and counted using the slope calculation function of ArcGIS 10.7. The aspect is also based on DEM, with a resolution of 12.5, which is obtained and counted using the aspect calculation function of ArcGIS 10.7. The stream order is based on 12.5 m DEM; the 9-stream order of the Beiluo River Basin is extracted using the ArcGIS 10.7 river network, and the statistics are calculated.
A total of 1781 landslides were identified, covering an investigation area of 27,000 square kilometers. The average landslide area was 72,278 square meters. The smallest landslide, with an area of approximately 2610 square meters, was the Chenbian Landslide #1, located in Chenbian Village, Xuecha Township, Wuqi County, Yan’an City, Shaanxi Province (Figure 4a). The largest landslide, with an area of 893,287 square meters, was the Hou Nijiagou #4 loess landslide, located in Xiaowei Village, Qian’er Township, Fu County, Yan’an City, Shaanxi Province (Figure 4b).

3.2.2. Kernel Density Analysis

The method of finding the overall distribution of a random variable includes parametric and non-parametric methods [49,50,51,52]. The parametric method is known random variables obey a certain distribution, but the specific parameters of the distribution are unknown, need to carry out parameter estimation and hypothesis testing, and finally, the overall distribution. For the case where the overall distribution is unknown, the non-parametric estimation method is used, which has the advantage of not relying on any assumptions about the distribution. The kernel density function estimation method is the more commonly used nonparametric estimation method, first using the kernel function to derive the probability density function, and then based on the kernel estimation of the probability density function to derive the marginal distribution function of the sample.
Suppose that the m sample points x 1 , x 2 , …, x n are from a one-dimensional continuous overall distribution X with an overall probability density function f ( x ) , and its kernel density estimation function can be expressed as:
f ^ h x = ∑ i = 1 m K x − x i h h m
where: f ^ h x is a kernel density estimation function for f x ; K • is the kernel function, ∫ − ∞ + ∞ K ( x ) d x = 1 ; h is the window width or bandwidth (h > 0).

3.2.3. Spatial Autocorrelation

Spatial autocorrelation analysis [53,54,55,56] is a test of whether observations of an element with a spatial location are significantly correlated with observations at its neighboring spatial points. The degree of similarity between the values of a spatial location attribute-valued feature variable and its neighboring spatial location attribute-valued features is used as a measure. Spatial autocorrelation can be expressed by a variety of indicators and methods, of which the most commonly used is Moran’s I index. The specific formula is as follows:
M o r a n ’ s I = n ∑ i = 1 n ∑ j = 1 n w i j ( x i − x ¯ ) ( x j − x ¯ ) ∑ i = 1 n ∑ j = 1 n w i j ∑ i = 1 n ( x i − x ¯ ) 2
where: n is the total number of elements; ( x i − x ¯ ) is the deviation of the observation from the mean on the first study cell; x ¯ is the mean, x ¯ = 1 n ∑ i = 1 n x i ; and w i j is the weight of elements i and j.
After calculating the Moran’s I index, it is generally subjected to a significance Z-test, which is calculated as follows:
Z = I − E ( I ) V a r ( I )
where: E(I) is the expectation of Moran’s I index; Var(I) is the variance.

3.2.4. Hot Spot Analysis

Cold and hot spot analysis [57,58] often uses the local Getis-Ord Gi* index method, which is used in the ArcGIS analysis tool for hot spot analysis, and is mainly used to test whether there are statistically significant high and low values in the local area, and can generally be used to visualize hot and cold spot areas. The Getis-Ord Gi* index is one of the indicators for detecting the local spatial autocorrelation of spatial points. One of the indicators for evaluating the degree of aggregation of points at the local spatial level, thus identifying statistically significant hot and cold areas of spatial aggregation. The local sum of an element and its neighboring elements is compared proportionally to the sum of all elements, and a statistically significant z-score will be generated when the local sum is so different from the desired local sum as to make it impossible to be a random result. The formula is calculated as follows:
G i * = ∑ J = 1 n W i j ( d ) X j ∑ j = 1 n X j ( J ≠ 1 )
Normalization of G i * ( d ) yields:
Z G i * = G i * − E G i * V a r ( G i * )
where E ( G i * ) and E ( G i * ) are the mathematical expectation and variance of G i * , respectively, and W i j is the spatial weight.

3.2.5. Kendall Rank Correlation

Kendall’s correlation [59,60] coefficient is a rank correlation coefficient which: Assuming two random variables X and Y, and the number of elements contained in both variables is N, the ith (1 ≤ i ≤ N) combination is denoted as (Xi, Yi). If Xi > Xj and Yi > Yj occur at the same time, or if Xi < Xj and Yi < Yj occur at the same time, the i-th and j-th combinations are considered to be the same. If Xi > Xj and Yi < Yj, or Xi < Xj and Yi > Yj, the two combinations are not consistent. The formula is as follows:
t = n c − n d 0.5 n ( n − 1 )
where n c and n d denote the number of consistent set pairs and inconsistent set pairs, respectively.

4. Results

4.1. The Distribution of Landslide Types and Characteristics

Identifying the distribution of landslide types and characteristics is crucial for evaluating landslide susceptibility [61,62,63]. The landslide database compiled herein includes 1781 landslides that occurred in the Beiluo River Basin (Figure 5), and it provides information on landslide location, age, scale, type, aspect, slope, profile morphology, and stream order [64]. The preliminary results indicate that 52 landslides, accounting for approximately 2.92% of the total, were caused by human activities (such as slope foot excavation, artificial loading, and slope cutting), while 1729 landslides, accounting for approximately 97.08% of the total, were triggered by natural factors (such as rainfall and erosional scouring).

4.1.1. Landslide Age, Scale, and Type

Out of the 1781 landslides in the Beiluo River Basin, there are primarily three categories based on age: old landslides (occurred since the Holocene) constitute the majority, with 1654 cases, accounting for 92.87% of the total landslides; new landslides (currently active) number 122, making up 6.85% of the total; and ancient landslides (occurred before the Holocene) number nearly 5, accounting for 0.28% of the total.
In terms of landslide scale, medium-scale landslides (10–100 × 104 m3) are predominant, with 1074 occurrences, accounting for 60.3% of the total landslides; large landslides (100–1000 × 104 m3) total 372, representing 20.89%; small landslides (0–10 × 104 m3) number 325, making up 18.25%; and extra-large landslides (>1000 × 104 m3) are rare, with only 10 instances, constituting 0.56% of the total. A distribution histogram is shown in Figure 6.
Regarding landslide types [65,66,67], the distribution is relatively uniform. The main types in the Beiluo River Basin include loess–bedrock interface landslides, with 787 occurrences, accounting for 44.19% of the total; intra-loess landslides, with 571 occurrences, making up 32.06%; and loess–clay interface landslides, with 423 occurrences, representing 23.75%. A distribution histogram is shown in Figure 7.

Loess–Bedrock Interface Landslide

Loess–bedrock interface landslides [68] occur at the toe of slopes, where the underlying bedrock dips outward, typically with a dip angle of 10–20°. The sliding bed consists of middle and upper Pleistocene loess in the middle and rear and Jurassic or Triassic sand–mudstone with weak structural surfaces in the middle and front. The rear of the sliding bed is steep, while the middle and front are relatively gentle, resulting in a generally low slope gradient. The sliding surface is arcuate in the middle and rear and nearly horizontal in the front. The sliding zone soil is dense and reddish brown, with a layered fractured structure in the loess section, appearing smooth and mirror-like on the surface. Some landslides exhibit smooth striations on mudstone surfaces within the sliding zone, with a thickness of 30–50 cm. Under self-weight, the sliding body deforms along the weak structural surface, causing bulging at the toe and tensile cracks at the top. Prolonged creep deformation occurs under the influence of rainfall and groundwater, leading to slope instability without a sudden reduction in shear resistance. Thus, loess–bedrock interface landslides generally have low sliding speeds, short distances, and high recurrence. The debris consists of middle and upper Pleistocene loess and some weathered sand–mudstone, with a distinguishable stratification and a thickness of 15–50 m. Figure 8 shows the evolution process of a loess–bedrock interface landslide. In the figure, I is the original topographic map, II is an early identification map, III is a precursor discriminant map, and IV is a disaster pattern recognition map.

Intra-Loess Landslide

Intra-loess landslides [69] primarily occur in areas with thick loess deposits. The main sliding surface is mostly arcuate, and the rear wall of the landslide is controlled by vertical joints, with slopes reaching 50–70°. The slope body moves towards the free face under self-weight, creating tensile cracks at the top. Rainwater enters these cracks, and, due to the unique properties of loess, the cracks soften and become interconnected, and the toe becomes saturated, intensifying creep deformation. Eventually, the loess body slides along the weak structural surface. Since the shear outlet is at the slope base, the landslide debris accumulates at the front with a relatively short distance and a minor height difference, but it is relatively fragmented. The accumulation range is mainly 2.5–3 times the height difference at the rear. The debris surface has a wavy topography and consists of middle and upper Pleistocene loess landslide debris, with a chaotic structure, mixed or relatively uniform colors (gray-yellow or light yellow), and a thickness of 10–30 m. Figure 9 shows the evolution process of an intra-loess landslide. In the figure, I is the original topographic map, II is an early identification map, III is a precursor discriminant map, and IV is a disaster pattern recognition map.

Loess–Clay Interface Landslide

Loess–clay interface landslides [70] often occur as large landslides along the loess platform edges. The sliding bed has middle Pleistocene loess at the rear and Neogene red clay at the middle and front. The overall slope of the sliding bed is steep at the rear and gentler at the front, forming an arcuate shape. The sliding zone soil is reddish brown to brown, smooth, with striations, dense, and layered with a thickness of 10–30 cm. Loess–clay interface landslides consist of loess at the top and Hipparion red clay at the bottom. Not only do they exhibit characteristics of being soft at the top and hard at the bottom, with the upper layer being loose and permeable and the lower layer being impermeable, but they also often contain weak zones along the slope (surface hipparion red clay affected by intense weathering). During the rainy season, atmospheric rainfall quickly infiltrates through vertical joints in the loess. The surface hipparion red clay, affected by intense weathering, becomes relatively impermeable and saturated. This increases the pore water pressure in the soil and reduces effective stress, leading to the formation of saturated interlayer weak zones and ultimately causing slope instability. The landslide debris consists mainly of middle and upper Pleistocene loess, with discernible stratification, and a thickness of 5–30 m. Figure 10 shows the evolution process of a loess–clay interface landslide. In the figure, I is the original topographic map, II is an early identification map, III is a precursor discriminant map, and IV is a disaster pattern recognition map.

4.1.2. Landslide Slope Gradient, Landslide Aspect, and Profile Morphology

Figure 11 shows the variation in landslide proportions with landslide slope. The highest proportion of landslides occurs on slopes with gradients between 16 and 25°. The landslide area proportion (the ratio of landslide area to total area for each class) reveals a significant concentration of landslides on slopes with gradients between 16 and 20°, indicating that not only is the number of landslides high in this gradient range, but the landslide areas are also relatively large. Figure 12 shows the distribution of landslide proportions by landslide aspect, with the highest proportion of landslides occurring on slopes oriented between 225° and 45° (i.e., from southwest to northeast). Additionally, 24.65% of landslides occur in the northwest direction, and the landslide area proportion is also relatively large in this aspect. Figure 13 illustrates the distribution of landslide profile morphology, with the majority being linear. However, the area proportion distribution reveals that convex landslides have relatively larger areas.

4.1.3. Stream Order

To analyze the relationship between stream development and landslides, the stream order of the landslide locations was statistically examined (Figure 14). The analysis shows that the landslides predominantly occur in the fifth- to eighth-order tributaries of the Beiluo River, with 80.07% of the landslides developing in these tributaries. Furthermore, the area proportion analysis indicates that large landslides mainly develop in fifth- and sixth-order tributaries and in some seventh-order tributaries, while small landslides are primarily distributed in eighth- and ninth-order tributaries. This indicates a certain clustering pattern in landslide scale development.

4.2. Spatial Distribution of the Landslides

4.2.1. Dispersion and Clustering of Landslides

A kernel dnsity analysis can be used to assess the density of point features within a neighborhood, generating a heat map that provides an intuitive way to understand the spatial distribution of phenomena. In kernel density mapping, points within the search area have different weights, with points closer to the search area center being assigned higher weights and weights decreasing with distance from the center [49,50,51,52]. To determine the spatial dispersion and clustering characteristics of landslide points, the ArcGIS 10.7 kernel density analysis tool was used to calculate the density of landslide points within their surrounding neighborhood. Using a circular search window with radii of 3 km and 10 km, a landslide density distribution map of the Beiluo River Basin was generated (Figure 15). The results show that, with a 10 km radius, the landslide points in the Beiluo River Basin exhibit significant dispersion and clustering characteristics. The areas with high landslide density are mainly concentrated in the loess ridge and hillock landform region around Wuqi County and Zhidan County and in the loess tableland region around Fu County and Luochuan County, with notable clustering effects in Fu County.
As the spatial scale changes, the distribution pattern of point features may also change. The distribution results shown in Figure 15b indicate that, with a 3 km search radius, the distribution characteristics of landslide points significantly change, exhibiting a relatively dispersed distribution. A multi-distance spatial cluster analysis (Ripley’s K function) can quantitatively describe the degree of clustering of landslide points at different spatial scales. Using the “Multi-Distance Spatial Cluster Analysis” tool in ArcGIS 10.7 to analyze the loess landslide points in the Beiluo River Basin, the results (Figure 16) show that, within the range of step lengths from 1 to 30 km, the observed K values are greater than the expected values and the confidence intervals. The observed K values exhibit an increasing trend, indicating that the landslide points show a clustered distribution pattern, and the clustering increases with the step length. When the step length exceeds 25 km, the clustering remains stable.

4.2.2. Spatial Autocorrelation of Landslides

Spatial autocorrelation refers to the statistical correlation of attribute values of geographic features distributed at different spatial locations, where peak values closer together tend to have higher correlations. Spatial autocorrelation primarily addresses the issue of clustering and dispersion within the study subject, examining whether a certain attribute of spatial features shows a high–high adjacent distribution or a high–low scattered distribution. According to the research scale, spatial autocorrelation can be divided into global spatial autocorrelation and local spatial autocorrelation [34]. The former (global spatial autocorrelation) describes the overall distribution pattern of the study subject, determining whether there is clustering within the study area but not identifying specific clustered regions. The latter (local spatial autocorrelation) describes whether the spatial autocorrelation of a clustering spatial unit relative to the overall study area is significant. When the significance is high, the unit is a clustering area within the study range. This analysis can quantitatively express clustering and outliers within the study area.
Global spatial autocorrelation is usually represented by Moran’s I index, Global Getis’s G index, and Global Geary’s C index, with Moran’s I index being the most commonly used [53,54,55,56]. Moran’s I value ranges between −1 and 1: “I > 0” indicates a spatial positive correlation; the closer the I value is to 1, the higher the clustering degree. “I < 0” indicates a spatial negative correlation; the smaller the I value, the greater the spatial difference. “I = 0” indicates no spatial autocorrelation, i.e., a random distribution.
Figure 17 and Figure 18 show the global autocorrelation calculation results for landslide scale levels and stream order levels in the Beiluo River Basin. The calculation results for scale levels are Moran’s I = 0.097, with a Z-score of 25.128; the calculation results for stream order levels are Moran’s I = 0.084, with a Z-score of 21.776. This indicates that the spatial distribution of landslides of different scale levels and stream order levels in the Beiluo River Basin exhibits a certain degree of clustering overall.
In this study, we conducted a local spatial autocorrelation analysis on the scale levels of loess landslides (Figure 19). The results indicate that the scale levels of loess landslides exhibit five clustering patterns in their spatial distribution: non-significant clustering, high–high clustering (HH), high–low clustering (HL), low–high clustering (LH), and low–low clustering (LL). Here, high–low clustering (HL) denotes high values primarily surrounded by low values (HL), and low–high clustering (LH) denotes low values primarily surrounded by high values (LH).
In the spatial distribution of landslide scale levels (Figure 19a), non-significant clustering areas are found in the loess ridge and hillock landform region from Wuqi County to Zhidan County in the northern part of the river basin, as well as in the loess tableland region from Fu County to Luochuan County. High–high clustering areas are mainly distributed in the loess tableland region from Fu County to Luochuan County. High–low clustering areas are not significant. Low–high clustering areas are the most prominent within the study area, mainly located in the northern part of the loess tableland region in Fu County. Low–low clustering areas are distributed in major landslide aggregation areas within the river basin but are generally not significant.
In the spatial distribution of landslide stream order levels (Figure 19c), non-significant clustering areas are found in the loess ridge and hillock landform region from Wuqi County to Zhidan County in the northern part of the river basin, as well as in the loess tableland region from Fu County to Luochuan County. High–high clustering areas are primarily distributed in the northern part of the loess tableland region from Fu County to Luochuan County and in the eastern part of the loess ridge and hillock landform region in Zhidan County. High–low clustering areas are mainly distributed in the eastern part of the gentle loess-covered mountains in Ganquan County. Low–high clustering areas are also primarily located in the northern part of the loess tableland region from Fu County to Luochuan County within the study area. Low–low clustering areas are found in the western part of the loess ridge and hillock landform region in Wuqi County.
To statistically quantify the spatial distribution of the different scale levels of loess landslides, we used the ArcGIS 10.7 Hotspot Analysis tool to generate a hotspot distribution map of landslide scale levels in the Beiluo River Basin (Figure 19b). The map clearly shows that the clustering of landslide scale hotspots (high values) and coldspots (low values) is relatively concentrated at different confidence levels. Hotspot (high value) regions are mainly concentrated in the northern part of the loess tableland region from Fu County to Luochuan County, indicating that this area has larger and more frequent loess landslides. Coldspot (low value) regions are mainly concentrated in the eastern part of the loess ridge and hillock landform region from Wuqi County to Zhidan County, indicating that the overall scale of loess landslides in this area is relatively small.
The hotspot distribution map of landslide stream order levels in the Beiluo River Basin (Figure 19d) shows that the clustering of landslide type-level hotspots (high values) and coldspots (low values) is relatively concentrated at different confidence levels. The northern part of the loess tableland region from Fu County to Luochuan County and the eastern part of the loess ridge and hillock landform region from Wuqi County to Zhidan County exhibit 99% confidence interval hotspots, indicating a high and relatively concentrated development of the loess landslide stream order in these areas. The western part of the loess ridge and hillock landform region from Wuqi County to Zhidan County shows a 99% confidence interval coldspot, indicating that loess landslides in this area are primarily developed in the mainstream order.

4.2.3. Spatial Distribution Characteristics of Different Landslide Types

To analyze the correlation between landslide types and landslide scale, stream order levels, and geomorphological types, we performed a Kendall correlation analysis (Table 1). The results indicate a positive correlation between landslide types and landslide scale (r = 0.076), with a strong significance (p < 0.001); a positive correlation between landslide types and stream order levels (r = 0.053), also with a strong significance (p < 0.001); and a positive correlation between landslide types and geomorphological types (r = 0.046), with a relatively strong significance (p < 0.005).
To further analyze the spatial aggregation and spatial autocorrelation characteristics of landslides in the Beiluo River Basin and to clarify the spatial distribution features of different landslide types, we conducted an overlay comparison analysis of various landslide types with geomorphological types, stream order spatial aggregation patterns, and scale spatial aggregation patterns (Figure 20). We found that all three types of landslides exhibit a clustered distribution in the northern ridge and hillock landform region and in the southern tableland region within the river basin.
Loess–bedrock interface landslides are relatively more numerous in the northern ridge and hillock landform region, displaying an east–west distribution trend along the Wuqi–Zhidan line, which corresponds to areas of high geomorphological development. In the loess tableland region, these landslides are mainly distributed along the valleys of the Beiluo River, Hulu River, and Ju River in the Fu County and Huangling areas. By comparing the landslide types with the landslide scale hotspot distribution map, it is observed that small landslides are predominant and relatively concentrated along the Wuqi–Zhidan line in the north, whereas medium and large landslides are predominant and relatively concentrated in the loess tableland region. Furthermore, by simultaneously comparing the hotspot distribution maps of landslide types and stream order, it is found that in the western region along the Wuqi–Zhidan line, landslides are relatively concentrated in the well-developed river network of the mainstream of the Beiluo River. In the eastern region along the Wuqi–Zhidan line, this type of landslide is relatively concentrated in the tributary river network of the Beiluo River. In the loess tableland region, the distribution of landslides across the river network is relatively scattered and lacks statistical significance.
Intra-loess landslides are mainly distributed in the southern loess tableland region. In this region, compared with the other two types, these landslides exhibit a stronger dispersal characteristic, whereas in the northern loess ridge and hillock landform region, they are relatively clustered in the loess–bedrock mountainous areas. By comparing this landslide type with the landslide scale hotspot distribution map, it is found that in the northern part of Huangling–Luochuan–Fu County, medium and large landslides are predominant and relatively concentrated, while in the southern part, small landslides are predominant. Additionally, by comparing the landslide types with the stream order level hotspot distribution map, it is found that in the northern part of Huangling–Luochuan–Fu County, these landslides are relatively concentrated in the tributary river networks of the Beiluo River, while in the western part of the Wuqi–Zhidan line, they are relatively concentrated in the main river network of the Beiluo River.
Loess–clay interface landslides are primarily distributed in the northern part of the southern loess tableland region and along both sides of the Beiluo River valley in Ganquan County. In the northern loess ridge and hillock landform region, they exhibit a dispersed distribution pattern along the river valleys. In the loess tableland region, they show significant clustering, contributing mainly to the overall landslide clustering and hotspot areas within the river basin. By comparing this landslide type with the landslide scale hotspot distribution map, it is observed that medium and large landslides are predominant and relatively concentrated in Luochuan County and Fu County, while small landslides are more concentrated around Wuqi County. Additionally, by comparing this landslide type with the stream order level hotspot distribution map, it is found that in Luochuan County and Fu County, these landslides are mainly distributed in the tributary river networks of the Beiluo River and are relatively concentrated, whereas, around Wuqi County, they are primarily distributed in the main river network of the Beiluo River and are relatively concentrated.

5. Discussion

The evolution of loess landforms in the Beiluo River Basin essentially represents the progressive development of the gully–valley water system from simple to complex and from low grade to high grade, eventually forming the present-day complex water network system with multiple river levels within the basin. The geometry, distribution characteristics, and evolution process of the gully–valley water system are influenced and controlled by various factors such as stratigraphy, lithology, topography, and geological structures. Conversely, the system also exerts significant effects on the surrounding geological environment through downward erosion, lateral erosion, and headward erosion. As a relatively rapid and special geomorphological phenomenon in the geomorphological evolution process, landslides are directly controlled by various erosion processes during the evolution of the gully–valley water system and these processes profoundly influence the macro-spatial distribution of landslides by shaping the geomorphological patterns of the basin.
Through the analysis of the geomorphological evolution process in the Beiluo River Basin, it was determined that the evolution of the Beiluo River system generally progresses upstream from the Luochuan tableland region towards the ridge and hillock landform region, but the historical evolution processes of different river segments within the basin vary. In the Luochuan tableland region, the river valley system, based on the inherited early water system pattern, had developed into a prototype of the modern water system by the early Middle Pleistocene, with the Ju River and Hulu River as the major tributaries already starting to develop. Influenced by the continuous subsidence of the base level of erosion in the Weihe Basin since the Quaternary, the middle reaches of the Beiluo River below Jiaokouhe Town have higher slopes and steepness indices. Coupled with the development of bedrock joints in the riverbed, four phases of downcutting and erosion since the Quaternary have resulted in the deep meandering valleys in the Jiaokou-Baishui section. After long-term natural evolution, the landslides and valley slope forms that developed early in the valley evolution have become stable. Furthermore, due to the low population density and sparse engineering construction on both sides of the valleys, there are fewer active and Holocene landslides today.
Spatial distribution, kernel density, and hotspot analyses of landslides in the Luochuan tableland region indicated that landslides are generally well developed in this area. The kernel density analysis showed strong clustering in Fu and Luochuan Counties. The autocorrelation analysis of landslide scale and causal types revealed high–low and low–high clustering patterns. The hotspot analysis results are also consistent with the spatial analysis results, with anomalous areas concentrated in Fu and Luochuan Counties. By further combining the distribution of different landslide types, it was found that the spatial distribution characteristics of landslides are highly correlated with the landslide types. Among them, intra-loess landslides mainly contribute to the formation of non-significant clustering patterns, primarily because the genesis of intra-loess landslides is mostly influenced by headward erosion and mainly occurs in the upper slopes or gully heads, resulting in a dispersed distribution across the tableland gully regions. Loess–bedrock interface landslides are mainly distributed along the valleys in Fu County, Huangling County, and the Hulu River. These landslides are primarily in the youthful or mature stages of valley development, with strong downcutting and erosion effects, relatively open valley development, and dense human engineering activities. For example, in the tributaries of Fu County, the gully width generally ranges from 600 to 800 m, with a downcutting depth greater than 200 m. The gully cross-section is U-shaped, with slopes of 15 to 35 degrees. This area is the location of county towns and major settlements, where the density of population and engineering construction is high, leading to frequent disturbances to the slopes on both sides of the valleys and forming exposed faces that are prone to landslides. The spatial distribution characteristics of loess–clay interface landslides are similar to those of loess–bedrock interface landslides, primarily controlled by the intensity of valley downcutting and erosion. However, the spatial distribution characteristics of loess–clay interface landslides are more pronounced, with almost no development in the central and southern parts of the Luochuan tableland region from Luochuan to Huangling. This is mainly because the early development of the Beiluo River in this area caused the red clay to thin or be eroded completely. Field investigations also revealed that red clay is sparsely exposed in the upper reaches of the gullies in this area, where headward erosion is predominant, and downcutting and lateral erosion are relatively weak, making it difficult to form exposed faces conducive to the development of loess–clay interface landslides.
In the landslide distribution map of the Beiluo River Basin, it could be observed that one significant feature of the spatial distribution of landslides is the relatively sparse development of landslides in the loess ridge and hillock landform region from Ganquan to Zhidan County in the upper reaches of the Beiluo River. This area’s geomorphological development is primarily characterized by loess ridges and gullies. According to the quantitative geomorphological differentiation results mentioned earlier, the terrain relief and vertical erosion degree in this area lie between those of the loess tableland region and the northern loess ridge and hillock landform region. The geomorphological dimensions in this region are relatively balanced, indicating a transitional state. The analysis of river terrace development and gully formation characteristics indicated that tectonic uplift in this area is relatively weak, and it is far from the river erosion base level of the Weihe Basin, resulting in relatively small river valley slopes compared to the Luochuan tableland area. The geomorphological development and evolution process in this region leads to a relatively gentle terrain with smaller height differences, making it less conducive for effective free faces to form in gullies, and, hence, fewer landslides develop.
The loess mound hilly gully area around Wuqi and Zhidan Counties in the upper reaches of the Beiluo River is another region with a high concentration of landslide development. The landslide density distribution shows a pattern spreading outward from the center of Wuqi County, without displaying a clear spatial aggregation pattern. The hotspot analysis showed anomalies only in the genesis type around Wuqi County. The geomorphological evolution in this region is influenced by the “Ancient Wuqi Lake”, with the modern Beiluo River system starting to develop after the Middle Pleistocene. Following the breach of the Ancient Wuqi Lake, the erosion base level rapidly decreased, leading to slope wash and gully incision, with the regional gully system developing rapidly. This active geomorphological evolution process created numerous favorable free faces for landslide development. Additionally, the development of five-level base terraces in the headwaters of Toudaochuan, Wuqi County, reflects significant and rapid neotectonic uplift since the middle Pleistocene. The terrace investigation results also show that the riverbed downcutting erosion rate around Wuqi in the upper Beiluo River is higher than that in the middle reaches of the Beiluo River. Controlled by the highly active regional geological tectonic background, landslides are generally well developed in this area. The distribution map of loess genesis types shows a close relationship between landslide distribution and geomorphological types. Intra-loess landslides mainly occur near both sides of the mainstream of the Beiluo River and on the right bank, primarily because the left side of this area around Zhidan features a mountainous terrain covered with thick loess, leading to relatively developed intra-loess landslides. Loess–bedrock interface landslides are the most common type in this area. This is mainly due to the geological tectonic background, with significant terrain incision, strong gully headward erosion, and riverbed downcutting, providing favorable free faces for landslide development. Loess–clay interface landslides are generally less developed and more dispersed, mainly because, during the early evolution stage of the “Ancient Wuqi Lake”, the red clay sedimentation range was limited and suffered severe erosion, resulting in the underdevelopment of potential loess–clay interface landslide structures in the study area. However, the research is still in the degree of combing and summarizing the macro laws, the research understanding is not deep enough, and there is a lack of more refined research.

6. Conclusions

This study compiled a large-scale landslide database in the Beiluo River Basin, Shaanxi Province, China, using methods such as remote sensing interpretation, field surveys, and historical landslide records. It investigated the landslide types and distribution characteristics and derived the following conclusions:
A landslide database was established for the Beiluo River Basin, containing various landslide characteristics and remote sensing images. There are a total of 1781 landslides in the Beiluo River Basin, with the largest area being 893,287 m2 and the smallest area being 40 m2, with an average area of 72,278 m2. The loess landslides in the Beiluo River Basin are predominantly old landslides (occurring since the Holocene); medium-scale landslides are the most common (60.3%); the landslide slopes mainly range from 16 to 25°; the landside aspect is mostly concentrated between 225 and 315°; and the tributary levels are mainly concentrated in the 6–8 levels.
Based on Varnes and Hungr’s landslide classification system and a review of the literature on loess landslide types, loess landslides were classified into three types according to the characteristics of the sliding surface: loess–bedrock interface landslides (44.19%), intra-loess landslides (32.06%), and loess–clay interface landslides (23.75%). The evolution and destruction processes of each type of landslide were categorized and summarized.
The distribution patterns of the landslides in the Beiluo River Basin were analyzed. With a search radius of 10 km, the landslide points in the Beiluo River Basin show significant clustering characteristics throughout the basin. Areas with a high landslide density are mainly concentrated in the loess ridge and hillock landform region around Wuqi and Zhidan Counties and in the loess tableland region around Fu and Luochuan Counties.
The global and local spatial autocorrelation of landslide scale and stream order level was analyzed. The results of the global spatial autocorrelation indicate that the spatial distribution of landslides of different scales and stream order levels in the Beiluo River Basin has certain clustering characteristics overall. The results of the local spatial autocorrelation show that areas with a high–high clustering of landslide scale are mainly distributed in the loess tableland region around Fu and Luochuan Counties. Areas with a high–high clustering of landslide stream order level are mainly distributed in the northern part of the loess tableland region around Fu and Luochuan Counties and in the eastern part of the loess ridge hilly geomorphological area around Zhidan County.
A correlation analysis using Kendall’s method revealed that landslide types are strongly correlated with landslide scale, tributary levels, and geomorphological types. Loess–bedrock interface landslides are relatively more numerous in the northern ridge and hillock landform region, and they are mainly distributed along the valleys of the Beiluo River, Hulu River, and Ju River around Fu County and Huangling County in the loess tableland region. Intra-loess landslides are mainly distributed in the southern loess tableland region, and they are relatively clustered in the northern loess ridge and hillock landform region near the loess–bedrock mountains. Loess–clay interface landslides are mainly distributed in the northern part of the southern loess tableland region and along both sides of the Beiluo River valley around Ganquan County. In the northern loess ridge and hillock region, they are scattered along the valleys.
The loess thickness is large in the loess tableland area, and, due to headward erosion, intra-loess landslides are the most developed, followed by loess–bedrock interface landslides, which mainly develop in the youthful or mature river valleys of the loess tableland area due to human engineering activities; loess–clay interface landslides are the least developed, mainly because the early development of the Beiluo River in this area led to the erosion and thinning of the red clay, leaving only a small amount of exposed red clay.
In the loess ridge and hillock landform region in the upper reaches of the Beiluo River, the loess thickness is relatively small, and the terrain incision depth is significant, with strong headward erosion and riverbed downcutting. Therefore, loess–bedrock interface landslides are the most developed, followed by intra-loess landslides, mainly distributed near both sides of the mainstream of the Beiluo River, primarily due to riverbed downcutting forming free faces; lastly, loess–clay interface landslides are less developed due to the limited extent of red clay deposits from the early “Ancient Wuqi Lake”, leading to only localized development.
The establishment of a landslide database in the Beiluo River Basin and the analysis of its relationship with the loess landform are helpful for land planning and engineering construction, which can effectively reduce the risk of landslides. However, the research still stays at the level of combing and summarizing the macro laws. The research understanding is not deep enough, and there is a lack of more refined research. We believe that the next research focus should be on the following aspects: carrying out landslide sensitivity evaluation in the Beiluo River Basin; analyzing the relationship between the initiation mechanism of different types of landslides and landforms; and establishing a landslide detection method based on remote sensing image.

Author Contributions

Conceptualization, F.L. and T.Z.; methodology, F.L. and Y.D.; software, F.L. and F.Q.; Investigation, F.L., H.T. and W.S.; validation, F.L., N.Y. and X.H.; writing—original draft preparation, F.L.; writing—review and editing, F.L., T.Z. and Y.D.; project administration, T.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the National Key R&D Program of China (2022YFC3003400); the Shaanxi Provincial Public Welfare Geological Survey Project (202101); and the National Natural Science Foundation of China (Grant No. 41772275).

Data Availability Statement

The data presented in this study are available on request from the corresponding author. The data are not publicly available due to legal.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Calvello, M.; Pecoraro, G. FraneItalia: A catalog of recent Italian landslides. Geoenvironmental Disasters 2018, 5, 13. [Google Scholar] [CrossRef]
  2. Haque, U.; Blum, P.; da Silva, P.F.; Andersen, P.; Pilz, J.; Chalov, S.R.; Malet, J.P.; Auflič, M.J.; Andres, N.; Poyiadji, E.; et al. Fatal landslides in Europe. Landslides 2016, 13, 1545. [Google Scholar] [CrossRef]
  3. Xu, C.; Xu, X.; Shyu, J.B.H. Database and spatial distribution of landslides triggered by the Lushan, China Mw 6.6 earthquake of 20 April 2013. Geomorphology 2015, 248, 77. [Google Scholar] [CrossRef]
  4. Roback, K.; Clark, M.K.; West, A.J.; Zekkos, D.; Li, G.; Gallen, S.F.; Chamlagain, D.; Godt, J.W. The size, distribution, and mobility of landslides caused by the 2015 Mw7.8 Gorkha earthquake, Nepal. Geomorphology 2018, 301, 121. [Google Scholar] [CrossRef]
  5. Marc, O.; Hovius, N. Amalgamation in landslide maps: Effects and automatic detection. Nat. Hazards Earth Syst. Sci. 2015, 15, 723. [Google Scholar] [CrossRef]
  6. Bucci, F.; Santangelo, M.; Cardinali, M.; Fiorucci, F.; Guzzetti, F. Landslide distribution and size in response to Quaternary fault activity: The Peloritani Range, NE Sicily, Italy. Earth Surf. Process. Landf. 2016, 41, 711. [Google Scholar] [CrossRef]
  7. Parker, R.N.; Densmore, A.L.; Rosser, N.J.; de Michele, M.; Li, Y.; Huang, R.; Whadcoat, S.; Petley, D.N. Mass wasting triggered by the 2008 Wenchuan earthquake is greater than orogenic growth. Nat. Geosci. 2011, 4, 449. [Google Scholar] [CrossRef]
  8. Borgatti, L.; Soldati, M. Landslides as a geomorphological proxy for climate change: A record from the Dolomites (northern Italy). Geomorphology 2010, 120, 56. [Google Scholar] [CrossRef]
  9. Bucci, F.; Cardinali, M.; Guzzetti, F. Structural geomorphology, active faulting and slope deformations in the epicentre area of the MW 7.0, 1857, Southern Italy earthquake. Phys. Chem. Earth 2013, 63, 12. [Google Scholar] [CrossRef]
  10. Xu, C.; Xu, X. Construction of basic earthquake-triggered landslides dataset for several large earthquake events at the beginning of the twenty-first century. Seismol. Geol. 2014, 36, 90. [Google Scholar]
  11. Shao, X.Y.; Ma, S.Y.; Xu, C.; Zhang, P.F.; Wen, B.Y.; Tian, Y.; Zhou, Q.; Cui, Y. Planet image-based inventorying and machine learning-based susceptibility mapping for the landslides triggered by the 2018 Mw6.6 tomakomai, Japan Earthquake. Remote Sens. 2019, 11, 978. [Google Scholar] [CrossRef]
  12. Ma, S.Y.; Xu, C. Assessment of co-seismic landslide hazard using the Newmark model and statistical analyses: A case study of the 2013 Lushan, China, Mw6.6 earthquake. Nat. Hazards 2019, 96, 389. [Google Scholar] [CrossRef]
  13. Xu, C.; Dai, F.; Xu, X.; Yuan, H.L. GIS-based support vector machine modeling of earthquake-triggered landslide susceptibility in the Jianjiang River watershed, China. Geomorphology 2012, 145, 70. [Google Scholar] [CrossRef]
  14. Steger, S.; Brenning, A.; Bell, R.; Glade, T. The propagation of inventory-based positional errors into statistical landslide susceptibility models. Nat. Hazards Earth Syst. Sci. 2016, 16, 2729. [Google Scholar] [CrossRef]
  15. Gómez, D.; García, E.F.; Aristizábal, E. Spatial and temporal landslide distributions using global and open landslide databases. Nat. Hazards 2023, 117, 25. [Google Scholar] [CrossRef]
  16. Van Den Eeckhaut, M.; Hervás, J. State of the art of national landslide databases in Europe and their potential for assessing landslide susceptibility, hazard and risk. Geomorphology 2012, 139, 545. [Google Scholar] [CrossRef]
  17. Devoli, G.; Strach, W.; Chávez, G.; Høeg, K. A landslide database for Nicaragua: A tool for landslide-hazard management. Landslides 2007, 4, 163. [Google Scholar] [CrossRef]
  18. Duman, T.Y.; Çan, T.; Emre, Ö.; Keçer, M.; Doğanet, A.; Ateş, S.; Durmaz, S. Landslide inventory of northwestern Anatolia, Turkey. Eng. Geol. 2005, 77, 99. [Google Scholar] [CrossRef]
  19. Van Westen, C.J.; Van Asch, T.W.J.; Soeters, R. Landslide hazard and risk zonation—Why is it still so difficult? Bull. Eng. Geol. Environ. 2006, 65, 167. [Google Scholar] [CrossRef]
  20. Rosi, A.; Tofani, V.; Tanteri, L.; Stefanelli, C.T.; Agostini, A.; Catani, F.; Casagli, N. The new landslide inventory of Tuscany (Italy) updated with PS-InSAR: Geomorphological features and landslide distribution. Landslides 2017, 15, 5–19. [Google Scholar] [CrossRef]
  21. Gao, Z.M.; Ding, M.T. Application of convolutional neural network fused with machine learning modeling framework for geospatial comparative analysis of landslide susceptibility. Nat. Hazards 2022, 133, 833. [Google Scholar] [CrossRef]
  22. Qi, T.J.; Zhao, Y.; Meng, X.M.; Shi, W.; Qing, F.; Chen, G.; Zhang, Y.; Yue, D.X.; Guo, F.Y. Distribution Modeling and Factor Correlation Analysis of Landslides in the Large Fault Zone of the Western Qinling Mountains: A Machine Learning Algorithm. Remote Sens. 2021, 13, 4990. [Google Scholar] [CrossRef]
  23. Li, X.B.; Yan, L.Y.; Wu, Y.W.; Peng, D.; Duan, J.J.; Bo, J.S. Distribution and characteristics of loess landslides induced by the 1654 Tianshui earthquake, Northwest of China. Landslides 2023, 20, 2775. [Google Scholar] [CrossRef]
  24. Bostjancic, I.; Avanic, R.; Frangen, T.; Pavic, M. Spatial distribution and geometric characteristics of landslides with special reference to geological units in the area of Slavonski Brod, Croatia. Geol. Croat. 2022, 75, 3–16. [Google Scholar] [CrossRef]
  25. Rosser, B.; Dellow, S.; Haubrock, S.; Glassey, P. New Zealand’s National Landslide Database. Landslides 2017, 14, 1949. [Google Scholar] [CrossRef]
  26. Damm, B.; Klose, M. The landslide database for Germany: Closing the gap at national level. Geomorphology 2015, 249, 82. [Google Scholar] [CrossRef]
  27. Sun, J.J.; Xu, C.; Feng, L.Y.; Li, L.; Zhang, X.W.; Yang, W.T. The Yinshan Mountains Record over 10,000 Landslides. Data 2024, 9, 31. [Google Scholar] [CrossRef]
  28. Peng, J.B.; Wang, S.K.; Wang, Q.Y.; Zhuang, J.Q.; Huang, W.L.; Zhu, X.H.; Leng, Y.Q.; Ma, P.H. Distribution and genetic types of loess landslides in China. J. Asian Earth Sci. 2019, 170, 329. [Google Scholar] [CrossRef]
  29. Zhang, S.; Li, C.; Peng, J.Y.; Zhou, Y.L.; Wang, S.R.; Chen, Y.M.; Tang, Y. Fatal landslides in China from 1940 to 2020: Occurrences and vulnerabilities. Landslides 2023, 20, 1243. [Google Scholar] [CrossRef]
  30. Li, W.Y.; Liu, C.; Hong, Y.; Zhang, X.H.; Wan, Z.M.; Saharia, M.; Sun, W.W.; Yao, D.J.; Chen, W.; Chen, S. A public Cloud-based China’s Landslide Inventory Database (CsLID): Development, zone, and spatiotemporal analysis for significant historical events 1949–2011. J. Mt. Sci. 2016, 13, 1275. [Google Scholar] [CrossRef]
  31. Zhang, Z.L.; Wang, T.; Wu, S.R. Distribution and features of landslides in the Tianshui Basin, Northwest China. J. Mt. Sci. 2020, 17, 686. [Google Scholar] [CrossRef]
  32. Shao, X.Y.; Ma, S.Y.; Xu, C.; Shen, L.L.; Lu, Y.K. Inventory, Distribution and Geometric Characteristics of Landslides in Baoshan City, Yunnan Province, China. Sustainability 2020, 12, 2433. [Google Scholar] [CrossRef]
  33. Froude, M.J.; Petley, D.N. Global fatal landslide occurrence from 2004 to 2016. Nat. Hazards Earth Syst. Sci. 2018, 18, 2161. [Google Scholar] [CrossRef]
  34. Dikshit, A.; Sarkar, R.; Pradhan, B.; Segoni, S.; Alamri, A.M. Rainfall induced landslide studies in indian himalayan region: A critical review. Appl. Sci. 2020, 10, 2466. [Google Scholar] [CrossRef]
  35. Li, Y.Y.; Huang, J.S.; Jiang, S.H.; Huang, F.M.; Chang, Z.L. A web-based gps system for displacement monitoring and failure mechanism analysis of reservoir landslide. Sci. Rep. 2017, 7, 17171. [Google Scholar] [CrossRef] [PubMed]
  36. Qiu, H.J.; Cui, P.; Regmi, A.D.; Wang, Y.M.; Hu, S. Slope height and slope gradient controls on the loess slide size within different slip surfaces. Phys. Geogr. 2017, 38, 303. [Google Scholar] [CrossRef]
  37. Dai, W.; Zeng, Y.; Jing, T.G.; Wang, Z.X.; Zong, R.J.; Ni, L.S.; Fang, N.F. Estimation of rainfall erosivity on the Chinese Loess Plateau: A new combination of the ERA5 dataset and machine learning. J. Hydrol. 2023, 624, 129892. [Google Scholar] [CrossRef]
  38. Denys, B. Mass movement; the research frontier and beyond: A geomorphological approach. Geomorphology 1993, 7, 85. [Google Scholar]
  39. Xu, C.; Xu, X.W.; Shyu, J.B.H.; Gao, M.X.; Tan, X.B.; Ran, Y.K.; Zheng, W.J. Landslides triggered by the 20 April 2013 Lushan, China, mw 6.6earthquake from field investigations and preliminary analyses. Landslides 2015, 12, 365. [Google Scholar] [CrossRef]
  40. Pike, R.J. The geometric signature: Quantifying landslide-terrain types from digital elevation models. Math. Geol. 1988, 20, 491. [Google Scholar] [CrossRef]
  41. Wen, B.Q.; Huang, C.L.; Zhou, C.; Zhang, H.; Yang, Q.K.; Li, M.C. Spatiotemporal dynamics and driving factors of soil erosion in the Beiluo River Basin, Loess Plateau, China. Ecol. Indic. 2023, 155, 110976. [Google Scholar] [CrossRef]
  42. He, Y.; Song, J.X.; Hu, Y.Y.; Tu, X.; Zhao, Y. Impacts of different weather conditions and landuse change on runoff variations in the Beiluo River Watershed, China. Sustain. Cities Soc. 2019, 50, 101674. [Google Scholar] [CrossRef]
  43. Wang, S.T.; Cao, Z.; Luo, P.P.; Zhu, W. Spatiotemporal V ariations and Climatological T rends in Precipitation Indices in Shaanxi Province, China. Atmosphere 2022, 13, 744. [Google Scholar] [CrossRef]
  44. Ma, L.F. Chinese Geological Atlas; Geological Publishing House: Beijing, China, 2002. [Google Scholar]
  45. Meng, Z.J.; Ma, P.H.; Peng, J.B. Characteristics of loess landslides triggered by different factors in the Chinese Loess Plateau. J. Mt. Sci. 2021, 18, 3218. [Google Scholar] [CrossRef]
  46. Alkevli, T.; Ercanoglu, M. Assessment of ASTER satellite images in landslide inventory mapping: Yenice-Gokcebey (Western Black Sea Region, Turkey). Bull. Eng. Geol. Environ. 2012, 70, 607. [Google Scholar] [CrossRef]
  47. Luino, F.; Barriendos, M.; Gizzi, F.T.; Glaser, R.; Gruetzner, C.; Palmieri, W.; Porfido, S.; Sangster, H.; Turconi, L. Historical Data for Natural Hazard Risk Mitigation and Land Use Planning. Land 2023, 12, 9. [Google Scholar] [CrossRef]
  48. Cui, Y.L.; Bao, P.P.; Xu, C.; Ma, S.Y.; Zheng, J.; Fu, G. Landslides triggered by the 6 September 2018 mw 6.6 Hokkaido, Japan: An updated inventory and retrospective hazard assessment. Earth Sci. Inform. 2020, 14, 247. [Google Scholar] [CrossRef]
  49. Ding, Z.L. Solid Earth Science Research Methods; Science Press: Beijing, China, 2013. (In Chinese) [Google Scholar]
  50. Livio, F.; Ferrario, M.F. Assessment of attenuation regressions for earthquake-triggered landslides in the Italian Apennines: Insights from recent and historical events. Landslides 2020, 17, 2825. [Google Scholar] [CrossRef]
  51. Qiu, H.J.; Cui, Y.F.; Hu, S.; Yang, D.D.; Pei, Y.Q.; Ma, S.Y.; Liu, Z.J. Size distribution and size of loess slides in response to slope height and slope gradient based on field survey data. Geomatics. Nat. Hazards Risk 2019, 10, 1443. [Google Scholar] [CrossRef]
  52. Zhu, K.X.; Xu, P.H.; Cao, C.; Zheng, L.J.; Liu, Y.; Dong, X.J. Preliminary Identification of Geological Hazards from Songpinggou to Feihong in Mao County along the Minjiang River Using SBAS-InSAR Technique Integrated Multiple Spatial Analysis Methods. Sustainability 2021, 13, 1017. [Google Scholar] [CrossRef]
  53. Lin, J.Y.; Li, X. MFOZ planning of Dongguan based on spatial autocorrelation by using genetic algorithms. Geogr. Res. 2014, 33, 349. [Google Scholar]
  54. He, Z.; He, J.P. Regional difference and Spatial patter evolution of net income of rural households based on spatial autocorrelation. Sci. Surv. Mapp. 2014, 39, 109. [Google Scholar]
  55. Zhang, S.L.; Zhang, K. Comparison between general Moran’s index and Getis-ord general G of spatial autocorrelation. Acta Sci. Nat. Univ. SunYatseni 2007, 46, 93. [Google Scholar]
  56. Getis, A.; Ord, J.K. The Analysis of Spatial Association by Use of Distance Statistics. Geogr. Anal. 2010, 24, 189. [Google Scholar] [CrossRef]
  57. Liu, X.M.; Su, P.C.; Li, Y.; Xu, R.; Zhang, J.; Yang, T.Q.; Guo, X.J.; Jiang, N. Spatial Patterns and Scaling Distributions of Earthquake-Induced Landslides-A Case Study of Landslides in Watersheds Along Dujiangyan-Wenchuan Highway. Front. Earth Sci. 2021, 9, 659152. [Google Scholar] [CrossRef]
  58. Pokharel, B.; Alvioli, M.; Lim, S. Assessment of earthquake-induced landslide inventories and susceptibility maps using slope unit-based logistic regression and geospatial statistics. Sci. Rep. 2021, 11, 21333. [Google Scholar] [CrossRef]
  59. Brophy, A.L. An algorithm and program for calculation of Kendall’s rank correlation coefficient. Behav. Res. Methods Instrum. Comput. 1986, 18, 45. [Google Scholar] [CrossRef]
  60. Witt, A.; Malamud, B.D.; Rossi, M.; Guzzetti, F.; Peruccacci, S. Temporal correlations and clustering of landslides. Earth Surf. Process. Landf. 2010, 35, 1138. [Google Scholar] [CrossRef]
  61. Dai, F.C.; Lee, C.F.; Ngai, Y.Y. Landslide risk assessment and management: An over-view. Eng. Geol. 2002, 64, 65. [Google Scholar] [CrossRef]
  62. Dai, F.C.; Lee, C.F.; Xu, Z.W. Assessment of landslide susceptibility on the natural ter-rain of Lantau island, Hong Kong. Environ. Geol. 2001, 40, 381. [Google Scholar]
  63. Guzzetti, F.; Reichenbach, P.; Ardizzone, F.; Cardinali, M.; Galli, M. Estimating the quality of landslide susceptibility models. Geomorphology 2006, 81, 166. [Google Scholar] [CrossRef]
  64. Horton, R.E. Erosional development of streams and their drainage basins: Hydrophysical approach to quantitative morphology. GSA Bull. 1945, 56, 275. [Google Scholar] [CrossRef]
  65. Duan, Z.; Zhao, F.S.; Chen, X.J. Types and Spatio-Temporal Distribution of Loess Landslides in Loess Plateau Region--A Case Study in Wuqi County. J. Catastrophology 2011, 26, 52. [Google Scholar]
  66. Liang, C.Y.; Zhang, H.; Wang, T. Red clay/mudstone distribution, properties and loess-mudstone landslides in the Loess Plateau, China. Environ. Earth Sci. 2022, 81, 15. [Google Scholar] [CrossRef]
  67. Meng, X.M.; Derbyshire, E. Landslides and their control in the Chinese Loess Plateau models and case studies from Gansu Province, China. Geohazards Eng. Geol. 1998, 15, 141. [Google Scholar] [CrossRef]
  68. Meng, Q.K.; Xu, Q.; Wang, B.C.; Li, W.L.; Peng, Y.; Peng, D.L.; Qi, X.; Zhou, D.D. Monitoring the regional deformation of loess landslides on the Heifangtai terrace using the Sentinel-1 time series interferometry technique. Nat. Hazards 2019, 98, 485. [Google Scholar] [CrossRef]
  69. Tang, Y.M.; Xue, Q.; Li, Z.G.; Feng, W. Three modes of rainfall infiltration inducing loess landslide. Nat. Hazards 2015, 79, 137. [Google Scholar] [CrossRef]
  70. Lian, B.Q.; Wang, X.G.; Liu, K.; Hu, S.; Feng, X. A mechanical insight into the triggering mechanism of frequently occurred landslides along the interface between loess and red clay. Sci. Rep. 2021, 11, 17556. [Google Scholar] [CrossRef]
Land 13 01038 g001
Figure 1. Geographic location map of the Beiluo River Basin and landslides mapped.
Figure 1. Geographic location map of the Beiluo River Basin and landslides mapped.
Land 13 01038 g001
Land 13 01038 g002
Figure 2. Geological map of the Beiluo River Basin.
Figure 2. Geological map of the Beiluo River Basin.
Land 13 01038 g002
Land 13 01038 g003
Figure 3. UAV photos of typical landslides in the study area.
Figure 3. UAV photos of typical landslides in the study area.
Land 13 01038 g003
Land 13 01038 g004
Figure 4. Remote sensing interpretation of landslide images.
Figure 4. Remote sensing interpretation of landslide images.
Land 13 01038 g004
Land 13 01038 g005
Figure 5. Landslide distribution map of the Beiluo River Basin.
Figure 5. Landslide distribution map of the Beiluo River Basin.
Land 13 01038 g005
Land 13 01038 g006
Figure 6. Proportion of landslide scale and number distribution.
Figure 6. Proportion of landslide scale and number distribution.
Land 13 01038 g006
Land 13 01038 g007
Figure 7. Proportion of landslide types by number and area.
Figure 7. Proportion of landslide types by number and area.
Land 13 01038 g007
Land 13 01038 g008
Figure 8. Loess–bedrock interface landslide.
Figure 8. Loess–bedrock interface landslide.
Land 13 01038 g008
Land 13 01038 g009
Figure 9. Intra-loess landslide.
Figure 9. Intra-loess landslide.
Land 13 01038 g009
Land 13 01038 g010
Figure 10. Development model of loess–clay interface landslide.
Figure 10. Development model of loess–clay interface landslide.
Land 13 01038 g010
Land 13 01038 g011
Figure 11. Proportion of landslide slope by quantity and area.
Figure 11. Proportion of landslide slope by quantity and area.
Land 13 01038 g011
Land 13 01038 g012aLand 13 01038 g012b
Figure 12. Distribution and proportion of landslide aspect by quantity.
Figure 12. Distribution and proportion of landslide aspect by quantity.
Land 13 01038 g012aLand 13 01038 g012b
Land 13 01038 g013
Figure 13. Proportion of landslide profile morphology by number and area.
Figure 13. Proportion of landslide profile morphology by number and area.
Land 13 01038 g013
Land 13 01038 g014
Figure 14. Proportion of landslide stream order by number and area.
Figure 14. Proportion of landslide stream order by number and area.
Land 13 01038 g014
Land 13 01038 g015
Figure 15. Landslide density distribution map of the Beiluo River Basin (a) search radius 3 km; (b) search radius 10 km.
Figure 15. Landslide density distribution map of the Beiluo River Basin (a) search radius 3 km; (b) search radius 10 km.
Land 13 01038 g015
Land 13 01038 g016
Figure 16. Results of landslide point multi-distance spatial clustering analysis.
Figure 16. Results of landslide point multi-distance spatial clustering analysis.
Land 13 01038 g016
Land 13 01038 g017
Figure 17. Global autocorrelation results of landslide scale levels in the Beiluo River Basin.
Figure 17. Global autocorrelation results of landslide scale levels in the Beiluo River Basin.
Land 13 01038 g017
Land 13 01038 g018
Figure 18. Global autocorrelation results of landslide stream order in the Beiluo River Basin.
Figure 18. Global autocorrelation results of landslide stream order in the Beiluo River Basin.
Land 13 01038 g018
Land 13 01038 g019aLand 13 01038 g019b
Figure 19. Results of spatial aggregation patterns and hotspot distribution of landslide scale levels and stream order levels in the Beiluo River Basin.
Figure 19. Results of spatial aggregation patterns and hotspot distribution of landslide scale levels and stream order levels in the Beiluo River Basin.
Land 13 01038 g019aLand 13 01038 g019b
Land 13 01038 g020aLand 13 01038 g020b
Figure 20. Comparative results of landslide types with geomorphological types, stream order spatial aggregation patterns, and scale spatial aggregation patterns in the Beiluo River Basin.
Figure 20. Comparative results of landslide types with geomorphological types, stream order spatial aggregation patterns, and scale spatial aggregation patterns in the Beiluo River Basin.
Land 13 01038 g020aLand 13 01038 g020b
Table 1. Kendall correlation calculation results.
Table 1. Kendall correlation calculation results.
ScaleTributary LevelGeomorphological Type
Kendall CorrelationTypeCorrelation Coefficient0.076 **0.053 **0.046 *
Sig. (Two-Tailed)00.080.024
Number178117811781
* Significant at the 0.05 level (two-tailed). ** Significant at the 0.01 level (two-tailed).
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Liu, F.; Deng, Y.; Zhang, T.; Qian, F.; Yang, N.; Teng, H.; Shi, W.; Han, X. Landslide Distribution and Development Characteristics in the Beiluo River Basin. Land 2024, 13, 1038. https://doi.org/10.3390/land13071038

AMA Style

Liu F, Deng Y, Zhang T, Qian F, Yang N, Teng H, Shi W, Han X. Landslide Distribution and Development Characteristics in the Beiluo River Basin. Land. 2024; 13(7):1038. https://doi.org/10.3390/land13071038

Chicago/Turabian Style

Liu, Fan, Yahong Deng, Tianyu Zhang, Faqiao Qian, Nan Yang, Hongquan Teng, Wei Shi, and Xue Han. 2024. "Landslide Distribution and Development Characteristics in the Beiluo River Basin" Land 13, no. 7: 1038. https://doi.org/10.3390/land13071038

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop