Rapidly and quantitatively assessing potential catastrophic debris flow risks over a regional scale is crucial for early-stage risk prevention and control. However, significant differences in the developmental conditions of various debris flow catchments pose considerable challenges to regional risk assessment. To address this, this study proposes a method for the rapid quantitative estimation of the population threatened by debris flows in the Bailong River Basin, aiming to identify potential sites of catastrophic debris flows. First, the SCS hydrological model is used to predict the peak discharge and maximum outflow volume of debris flows under a designed rainfall scenario. Including the innovative application of machine learning to construct a sediment supply prediction model. Then, an empirical formula for debris flow deposition fans in the region is applied to estimate the potential hazard area. Finally, this potential hazard area is overlaid with spatially gridded population distribution data to estimate the number of people potentially threatened by each debris flow catchment. Utilizing publicly available fundamental data, this method constructs a comprehensive estimation process for debris flows, covering rainfall-runoff-outflow volume-hazard area-threatened population. It provides a robust and scalable solution for the rapid quantitative assessment of regional catastrophic debris flow risks.

The traditional theoretical assumption that preferential flow forms within preferential infiltration pathways (PIPs) oversimplifies hydro-mechanical coupling processes. This simplification stems from neglecting the critical influence of the PIPs' occurrence modes on its development, leading to increased systematic deviations when evaluating slow and/or abrupt landslide failure behaviors. In-situ loess slope rainfall simulation under extreme rainfall conditions and multiple in-situ infiltration experiments incorporating different PIP occurrence modes were conducted. These experiments reveal that preferential flow does not always develop within PIPs, and that the occurrence modes and dynamic evolution of PIPs regulate preferential flow formation. Only when loess moisture content exceeds a similar to 21 % threshold, preferential flow becomes independent of path occurrence mode-a condition under which moisture migration is no longer strictly governed by suction control. Thus, pore water pressure associated with preferential flow does not invariably induce rapid landslides. Moreover, even when positive pore water pressure accumulates effectively within the slope, it drops intermittently as the PIP occurrence modes evolve, and may even drop to zero. Instead, the evolution of the loess fabric near the PIP causes significant strength loss. Specifically, the cohesion can decrease by up to approximately 50 % before and after slope failure. This demonstrates that material strength attenuation related to preferential flow exerts a long-term regulatory effect on landslide dynamics. Nevertheless, this crucial mechanism is often ignored. Our findings reevaluate the traditional conceptual framework of PIP hydrological effects, thereby enhancing the understanding of complex feedback mechanisms between environmentally related hydrological effects and landslide dynamics.

Regional identification and risk mapping of landslide and debris flow hazard chains (LDHCs) are essential for disaster prevention and land-use planning in mountainous regions. This study proposes a machine learning framework for the regional identification and risk mapping of LDHCs in southern Gansu Province, China. A spatial database was established using an inventory of 160 landslide hazard chains (LHCs) and 121 debris-flow hazard chains (DHCs), together with eight environmental conditioning factors. Considering the distinct geomorphological characteristics of the two hazard types, grid-based and watershed-based mapping units were adopted for landslide and debris-flow susceptibility mapping, respectively. Five ensemble learning algorithms were evaluated to identify the optimal susceptibility models. Hazard maps were subsequently generated by integrating susceptibility with earthquake- and extreme rainfall-triggering factors, and regional risk maps were produced by incorporating population and Gross Domestic Product (GDP) exposure data. The Random Forest model achieved the best performance for LHC susceptibility mapping, with an accuracy of 84.9% and an AUC of 0.914, whereas the Extra Trees model performed best for DHC susceptibility mapping, with an accuracy of 92.6% and an AUC of 0.978. The resulting risk maps indicate that the middle and lower reaches of the Bailong River, including Wudu, Zhouqu, Qin’an, and Tianshui, represent the highest-risk areas for LDHCs. The proposed framework provides an effective and transferable approach for regional identification and risk mapping of LDHCs, offering valuable support for disaster prevention, emergency planning, and land-use management in mountainous regions.

Landslide caused catastrophic disasters frequently in the Karakoram Mountains wide range. The Hunza Valley, Pakistan in the Northwest of the Karakoram mountains, which is prone to the clustering development of landslide was taken as a case in this study. The updated complete inventory including 53 SBAS-InSAR detected active landslides and optical image interpreted 65 landslides were constructed, based on Sentinel-1A data in 2019–2020 and several field survey until 2023. Twelve factors related to geomorphology, hydrology, vegetation, geology, tectonics, and environment were incorporated into the model training within twelve machine learning models: Generalized Linear models, Navies Bayes, Nearest Neighbors, Support Vector Machines and so on. The Support Vector Classification was selected for landslide susceptibility mapping (LSM) and its characteristics in geomorphologically meaningful landscape partitions called slope units, with the highest accuracy of 0.96, average AUC of 0.99 for tenfold cross-validation, and high computational efficiency of 6.11 s. The results revealed that the areas with moderate landslide susceptibility account for 62.14

Constructing catchment-specific rainfall thresholds is crucial for high-accuracy debris flow forecasting. However, rapid methods for determining thresholds across multiple catchments remain limited due to the difficulty of relating catchment characteristics to peak discharges. This study explores such relationships using hydrological simulations and machine learning techniques, aiming to enable the rapid estimation of catchment-specific rainfall thresholds. We conducted a series of hydrological simulations with varied rainfall amounts to calculate peak discharges for 267 catchments in the Bailong River Basin, China. The rainfall amounts and corresponding peak discharges were fitted using three types of mathematical functions: linear, polynomial, and power-law. Machine learning models were then applied to predict the fitted coefficients for each catchment based on thirteen catchment features related to runoff generation and concentration. The Extra Trees Regressor (ETR) was used for catchments larger than 10 km(2), while CatBoost Regressor (CBR) was employed for smaller ones. This modelling framework allows the derivation of functional relationships between rainfall and discharge for each catchment, which were subsequently used to estimate rainfall thresholds for debris flow initiation. The proposed method shows strong potential for improving forecasting accuracy and regional applicability, and it can be further enhanced with more hydrological observations and long-term monitoring data.

Rainfall plays a crucial role in triggering debris flows, and the rainfall Intensity-Duration (I-D) threshold curve is widely used in early warning systems for debris flows and landslides. However, this model has a notable limitation: it cannot differentiate between rainfall processes with different peaks when the rainfall duration and average rainfall intensity of two rain events are the same. To address this limitation, we proposed the concepts of energy rainfall intensity (Ie) and comprehensive rainfall intensity (Ic) for the first time. By combining Ie, maximum rainfall intensity (Im) and average rainfall intensity (I) in different ways, we constructed a comprehensive rainfall intensity-duration threshold curve (Ic- D). By simulating the differences in peak discharges produced by different rainfall processes using the SCS hydrological model, we explored the superiority of the new rainfall intensity parameters. This newly-developed threshold curve improves the accuracy in predicting debris flow occurrences, reducing false alarms and missed alarms by 23 % compared to a traditional I-D approach, and holds importance for improving early warning models in predicting debris flows.

Rainfall thresholds for debris flow initiation are a common index for debris flow forecasting across the globe. However, as many debris flow catchments have been managed using mitigation measures such as drainage channels and check dams, the occurrence of debris flow may not necessarily lead to damage to properties or lives. Consequently, a higher false alarm rate may be caused when using the initiation threshold in catchments where mitigation measures have been installed. To overcome this limitation, we have defined and constructed a disastrous rainfall threshold over which debris flows may overflow the drainage channel and cause damage to surrounding residents in a typical debris flow catchment in China. This threshold was obtained by using the SynxFlow debris flow model, which is capable of integrating the process of rainfall-runoff-debris flow dynamics. A series of rainfalls with varying intensities and durations were used to simulate debris flow overflow conditions, and the disastrous I-D threshold (I = 8.7D−1.26 + 30.4) was determined using fitting methods. The disastrous rainfall threshold requires much higher rainfall intensity in order to cause a threat to residential areas compared to the initiation threshold. The disastrous threshold intensity for a one-hour duration rainfall is 39.1 mm/h for the study catchment, much larger than the 11.9 mm/h initiation threshold. The study also found that differences in rainfall patterns have a significant impact on the magnitude of the disastrous threshold. The threshold for uniform rainfalls is higher than those for other rainfall patterns (delayed, advanced and intermediate). We propose that the differentiation between disastrous threshold and initiation threshold may help improve the accuracy of debris flow forecasting and early warning.

Under the background of global climate change, shallow landslide clusters induced by extreme rainfall are occurring with increasing frequency, causing severe casualties and economic losses. To address this challenge, this study proposes an integrated approach to support both emergency response and long-term mitigation for rainfall-induced shallow landslides. The workflow includes (1) rapid landslide detection based on time-series image fusion and threshold segmentation on the Google Earth Engine (GEE) platform; (2) numerical simulation of landslide runout using the R.avaflow model; (3) landslide susceptibility assessment based on event-driven inventories and machine learning; and (4) delineation of high-risk slopes by integrating simulation outputs, susceptibility results, and exposed elements. Applied to Qugaona Township in Zhouqu County, Bailong River Basin, the framework identified 747 landslides. The R.avaflow simulations captured the spatial extent and depositional features of landslides, assisting post-disaster operations. The Gradient Boosting-based susceptibility model achieved an accuracy of 0.870, with 8.0% of the area classified as highly susceptible. In Cangan Village, high-risk slopes were delineated, with 31.08%, 17.85%, and 22.42% of slopes potentially affecting buildings, farmland, and roads, respectively. The study recommends engineering interventions for these areas. Compared with traditional methods, this approach demonstrates greater applicability and provides a more comprehensive basis for managing rainfall-induced landslide hazards.

Debris flows are a type of natural disaster induced by vegetation-water-soil coupling under external dynamic conditions. Research on the mechanism by which underground plant roots affect the initiation of gulley debris flows is currently limited. To explore this mechanism, we designed 14 groups of controlled field-based simulation experiments. Through monitoring, analysis, calculation, and simulation of the changes in physical parameters, such as volumetric water content, pore-water pressure, and matric suction, during the debris flow initiation process, we revealed that underground plant roots change the pore structure of soil masses. This affects the response time of pore-water pressure to volumetric water content, as well as hydrological processes within soil masses before the initiation of gully debris flows. Underground plant roots increase the peak volumetric water content of rock and soil masses, reduce the rates of increase of volumetric water content and pore-water pressure, and increase the dissipation rate of pore-water pressure. Our results clarify the influence of underground roots on the initiation of gulley debris flows, and also provide support for the initiation warning of gully debris flow. When the peak value of stable volumetric water content is taken as the early warning value, the early warning time of soil with underground plant roots is delayed by 534 to 1253 s. When the stable peak value of pore-water pressure is taken as the early warning value, the early warning time of soil with underground plant roots is delayed by 193 to 1082 s. This study provides a basis for disaster prevention and early warning of gully debris flows in GLP, and also provides ideas and theoretical basis under different vegetation-cover conditions area similar to GLP.

Debris flow can cause damage only when its discharge exceeds the drainage capacity of the prevention engineering. At present, most rainfall thresholds for debris flows mainly focus on the initiation of debris flow and do not adequately consider the magnitude and drainage measures of debris flows. These thresholds are likely to initiate numerous warnings that may not be related to hazardous processes. This study proposes a method for calculating the rainfall threshold that is related to a defined level of debris flow magnitude, over which certain damage may be caused. This method is constructed by using the transient rainfall infiltration analysis slope stability model (TRIGRS) and the fluid dynamics process simulation model (MassFlow). We first use the TRIGRS model to analyze slope stability in the study area and obtain the distribution of unstable slopes under different rainfall conditions. Afterward, the MassFlow model is employed to simulate the movement process of unstable slope units and to predict the depositional processes at the mouth of the catchment. Lastly a rainfall threshold is constructed by statistically analyzing the rainfall conditions that cause debris flows flushing out of the given drainage ditch. This method is useful to predict debris flow events of a hazardous magnitude, especially for areas with limited historical observational data.

The Bailongjiang River Basin is a high-risk area for debris flow in China. On 17 August 2020, a debris flow occurred in the Shuimo catchment, Wen County, which blocked the Baishui River, forming a barrier lake and causing significant casualties and property damage. In this study, remote sensing, InSAR, field surveys, and unmanned aerial vehicle (UAV) techniques were used to analyze the causal characteristics, material source characteristics, dynamic processes, and disaster characteristics after the debris flow. The results showed that the Shuimo catchment belongs to low-frequency debris flows, with a recurrence cycle of more than 100 years and concealed features. High vegetation coverage (72%) and a long main channel (11.49 km) increase the rainfall-triggering conditions for debris flow occurrence, making it more hidden and less noticed. The Shuimo catchment has a large drainage area of 31.26 km2, 15 tributaries, significant elevation differences of 2017 m, and favorable hydraulic conditions for debris flow. The main sources of debris flow material supply are channel erosion and slope erosion, which account for 84.4% of the total material. The collapse of landslides blocking both sides of the main channel resulted in an amplification of the debris flow scale, leading to the blockage of the Baishui River. The scale of the accumulation fan is 28 × 104 m3, and the barrier lake area is 37.4 × 104 m2. The formation mechanism can be summarized as follows: rainfall triggering → shallow landslides → slope debris flow → channel erosion → landslide damming → dam failure and increased discharge → deposition and river blockage. The results of this study provide references for remote sensing emergency investigation and analysis of similar low-frequency and concealed debris flows, as well as a scientific basis for local disaster prevention and reduction.

The Hunza Valley in the China-Pakistan Economic Corridor(CPEC)in the northern part of Pakistan has a high relief and harsh geo-environment.Villages and towns in this area are prone to geohazard development,and high-risk incidents have been observed from the construction to operation stages of the CPEC.Landslide hazards in the Hunza Valley must be investigated and analyzed via landslide inventories and landslide development tools.This study applied 45 images and 42 images from the ascending and descending Sentinel-1A datasets,respectively,to monitor surface deformation via SBAS-InSAR.The deformation information along the slope direction was subsequently estimated.On the basis of the displacement rates derived from the SAR data,the optical remote sensing images were visually interpreted,and in situ surveys and validations were conducted.A total of 53 potential landslides were detected and delineated.On the basis of the effects of landslide identification and the detected deformation,image interpretation and validation features of typical large landslides Ghulmet and Humarri,11 factors related to geomorphology,geology,hydrology,and vegetation were analyzed for landslide development.Maximum displacement velocities of-311 and-490 mm/a along the slope were detected on the basis of the ascending and descending datasets,respectively.Consequently,an annual deformation velocity of 20 mm/a was set as the threshold for the detection and mapping of potential landslides in the Hunza River Valley.The deformation of large landslides is severe under the influence of Hunza River erosion,and secondary landslides are developed.The validated potential landslides are distributed on the slopes on both sides of the Hunza River and are sometimes on the upper and lower slopes of the road.These active landslides primarily are developed in metamorphic rocks such as phyllite and slate.In the CPEC,landslides preferentially form and deform in areas where the elevation relief is between 200 and 1000 m,the slope is between 30° and 40°,and the aspect is within the southern and southwestern regions.Given the bare area of slope surfaces and sparse vegetation(NDVI＜0.2),weathered and fragmented slopes provide enough provenance and materials for landslide development.The outcomes and results may facilitate hazard management and risk reduction in the Hunza Valley,allowing the operation of the CPEC to be uninterrupted.The findings of this work can also provide scientific references and data support for the monitoring and assessments of major landslide disasters that destroy roads and block rivers and their resulting secondary disaster events.

At 18:00 on July 19, 2019, the ancient Yahuokou flow-like landslide in Dongshan Town, Zhouqu County, Gansu Province, China, was revived. The landslide mass of about 3.92 × 106 m3 was revived step by step, slid down the slope, and finally slipped into the Minjiang river, causing the Minjiang river to become a semi-blocked state, the water level of the river rose, and the riverside road was interrupted. The landslide is 1920 m long, with a height difference of 550 m and an apparent friction coefficient of 0.286. It is a typical push-type large-scale long-striped fault-fractured zone accumulation landslide, which is controlled by the Pingding-Huama fault zone. Under the action of long-term rainfall, the trailing edge of the upper section of the landslide first undergoes slow creep deformation, which eventually leads to the overall instability and decline of the upper section of the landslide, and then loads the upper part of the middle section of the sliding mass, which promotes the instability and decline of the middle section of the sliding mass. Finally, the sliding mass in the middle section exerts a load on the upper part of the sliding mass in the lower section, which revives the sliding mass in the lower section and gradually slides into the Minjiang river. Through field investigation, remote sensing interpretation, and borehole investigation, the basic characteristics and information of landslide deformation and movement are obtained. The limit equilibrium method is used to analyze the revival mechanism and the spring model is proposed to analyze the movement mechanism of the landslide. With the help of DNA-W numerical simulation software, the sliding process of the upper sliding mass is simulated by a frictional rheological model to verify the spring model. The analysis of the dynamic mechanism of the Yahuokou flow-like landslide will help provide a reference for the research on the instability and movement mechanism of the landslide in the Pingding-Huama fault zone, and provide the necessary scientific basis for the prevention and mitigation of landslide disasters.

利用SBAS-InSAR技术对滑坡易发区域进行时序变形监测是研究滑坡机理和预防灾害的重要途径.文章利用SBAS-InSAR技术对甘肃省甘南州舟曲县果耶镇磨里滑坡灾害进行时序变形特征分析,基于2018-2021年光学遥感影像数据获取磨里滑坡所在区域的变形分布图和时序变形特征,再结合其他勘查成果资料验证变形监测结果的准确性,证明SBAS-InSAR技术在特大型滑坡变形监测中的可靠性.结果表明:(1)磨里滑坡滑坡平面形态呈"长舌状",滑动方向最大长度1 500 m,滑坡体上窄下宽,平均宽度240～530 m,面积53×104 m2;(2)磨里滑坡前部变形速度大于中后部,前缘局部形变速率最大可达140 mm/a;(3)磨里滑坡还在不断地发生变形,应进一步加强变形监测,做好应急避险和搬迁避让的措施.研究可为该地区特大型滑坡复活破坏机制和应急处置提供参考.

At 7:50 on April 29, 2015, a large-scale landslide occurred in Heifangtai, Yongjing County, Gansu Province, with a landslide volume of 126.88 × 104 m3, destroying 14 houses and three factories, resulting in a direct economic loss of 54.6 million yuan. Based on a large number of geological investigations on the disaster site, combined with comprehensive investigation means such as remote sensing, three-dimensional laser scanning and on-site video monitoring data, this paper evolution and sliding characteristics of Luojiapo landslide in detail, and analyzes the high-speed remote mechanism of the landslide. The results show that Luojiapo landslide has gone through five stages and five sliding forms from the timeline: loess staggered debris flow, loess mudstone debris flow, loess debris flow, loess mudflow and loess staggered sliding. The landslide movement mode can be divided into two types: block debris flow and loess mud flow. The high-speed and long-distance formation mechanism of block debris flow is closely related to the “concave bed filling effect” of the previous landslide and the underlying surface soil with high water content. The research results have positive guiding significance for further deepening the understanding of the formation mechanism and risk control of high-speed and long-distance landslides in the Heifangtai area.

The China-Pakistan Economic Corridor(CPEC)is a key passage connecting China with South Asia and West Asia.Most parts of CPEC are in areas of the Tibetan Plateau and Pamir Plateau and,with unique and dynamic geomorphology and geo-environment backgrounds,a mass of landslides,debris flow,and glacier surging-related or-induced geohazards,presenting challenges to the normal construc-tion and operation of CPEC,and also bringing disasters and loss for the local society and human life.The study systematically summarized the study progress in the geohazard identification and inventory,hazard assessment,and geohazard prevention and mitigation.Based on that,the shortages of recent studies were pointed out.The study on geohazard development background has been relatively unsubstantial,and early identification of geohazards needed to be further strengthened,the characteristics and mechanism of geo-hazards have not been clearly investigated,the methodology of hazard assessment needed to be advanced,and the comprehensive prevention and control of geological disasters needed to be further enhanced.According suggestions were proposed.The study proposed the prospects in three fields including carry-ing out an environmental geology census and geo-hazard detailed survey to establish a complete database on geological disasters;promoting the establishment of an air-space-ground joint monitoring and identifi-cation technology system suitable for the geohazards study in CPEC,promoting satellite data sharing;car-rying out systematic and comprehensive research on geological disaster prevention and control,risk iden-tification,disaster reduction and post-disaster recovery in CPEC.

Intermittently moving earthflows are widespread geomorphological phenomena in mountainous regions. The Daxiaowan (DXW) earthflow in Zhouqu County, Bailong River Corridor of north-eastern margin of the Qinghai-Tibet Plateau, has been active for over a decade. The earthflow covering has an area of-1.1 km2 and a length of-2.7 km, and it has developed four large secondary landslides and two potential landslides. The slope move-ments pose great threats to the safety of the local residents and the provincial highway S314 at the toe of the DXW. In this study, we used a combination of interferometric synthetic aperture radar technology, unmanned aerial vehicle photogrammetry, geomorphological interpretation, and field investigations to study the defor-mation process and kinematic evolution of the DXW between 2007 and 2020. Field investigations of the earthflow revealed distinct source, transport, and compressional zone with different movement attributes. The displacement time-series results indicated that the deformation process of the DXW earthflow could be divided into two stages: steady deformation with cracks propagation and scarps slumping (during 2007-2016); pro-gressive deformation with surges in the rainy season (during 2017-2020). Rainfall was the primary driver of deformation since 2017. The catchment topography generated by the Pingding-Huama fault has caused flowing water to accumulate on the left side of the catchment, as indicated by the Topographic Wetness Index, leading to the reactivation of DXW. This earthflow was initially reactivated from the middle part of the ancient landslide body, characterized by the gradual development from cracks and scarps to four secondary landslides. Our characterization of the mode of evolution of the DXW earthflow potentially improves our understanding of earthflow movement and landslide hazards in the Bailong River Corridor, and it provides a scientific reference for mitigating disasters risks in this and similar areas.

Due to the heavy rainfall, from February 26th to 28th 2021, the Moli landslide in Guoye Town, Zhouqu County, experienced creeping deformation with the development of slope cracks and obnious signs of deformation. A total of of 92 households with 402 people were affected, causing a direct economic loss of approximately 14.463 million yuan. This study focuses on the Moli landslide in Guoye Town, where the geological environment and physical and mechanical properties of the soil were deeply understood through remote sensing interpretation, aerial photography, and geological investigation. The cause of the landslide formation was analyzed in detail, and the stability of the landslide was calculated theoretically and analyzed for displacement monitoring. Based on field investigation and drilling analysis, the Massflow numerical model was used to simulate and predict the Moli landslide in Guoye Town, Zhouqu County. The range and thickness of the source area were determined to predict the accumulation process of the landslide and the risk of river blocking. The height of the river blocking and the harm caused by the landslide to the upstream and downstream regions were predicted. The results showed that: (1) Moli landslide is a super-large deep landslide with a long tongue shape, clear shape and deformation, and composed of broken phyllite and broken stone soil. The average depth of the slide body is 40m, and the total volume is 21.2 million stere. (2) The main factors of landslide formation are unfavorable topographic conditions, softening of rock and soil, strong tectonic movement, rainfall infiltration, and front river erosion. (3) The landslide has the risk of destroying the landslide houses, roads, and blocking the river. It is suggested that the residents threatened by the Moli landslide should take measures to avoid danger and relocate as soon as possible. This study can provide reference for the formation mechanism and emergency prevention and control of similar landslide geological disaster chains.

The landslide-debris flow-river blockage hazard chain is one of the most common types of geological hazard chain in mountainous areas, seriously threatening human life and property across an extensive geographical area. The development of approaches for the rapid investigation and prediction of multi-hazard chains can play a significant role in reducing the number of casualties and economic losses. Here, taking the Lijie landslide-debris flow-river blockage in the Bailong River basin as a case study, we used remote sensing methods, geophysical technology, numerical simulations and empirical formulas to develop an integrated approach for analyzing the mechanisms, multi-hazard prediction and quantitative risk assessment induced by this geological hazard chain. The reactivation process of the Lijie landslide is composite retrogressive sliding. The steep slopes and highly fractured bedrock with low strength are the principal internal factors, and heavy rainfall in 2020 and spring freeze-thaw effect are the principal triggering factors. The deposited landslide material is eroded and transformed into a channel-type debris flow under high rainfall conditions. The steep terrain and abundant finegrained material provides favorable conditions for a viscous debris flow with a long runout, which is usually followed by secondary disasters associated with river blockage. Our results indicate that under rainfall conditions with a 100-year return period, the Bailong river will be completely blocked by the landslide-induced debris flow. And there will be an estimated 5.84 million CNY of economic losses caused by the multi-hazard chain, which is 2.6 times the losses directly caused by the landslide. Our integrated approach is practicable in an emergency context and useful for local administrators to rapidly develop strategies for multi-hazard chain prevention and mitigation, based on information about the transformation mechanisms of the multi-hazard chain, potential hazard-affected areas, damaged degree of elements at the risk, and incurred economic losses.

This data set is the latest data set of landslide distribution and characteristics along the China-Pakistan Karakoram Highway, which is based on ground deformation rate monitoring, optical remote sensing interpretation, field investigation, and verification, with the comprehensive application of digital elevation model (DEM), geological map, seismic distribution, precipitation, and vegetation data. In the field investigation and verification, according to the preliminary delineation results, the position and area of deformation on the slope, the tensile scarp and deformation cracks of the potential landslide, shearing boundary and cracks, and boundaries are verified, and the necessary verification and modification are made to complete the landslide inventory. The characteristics of morphological, deformation, on the image, and material characteristics of different types of landslides are summarized. In this data set, the overall distribution range of landslides along the Karakoram Highway is 34.5°N -- 39.5°N, 72.5°E -- 76.0°E. It is mainly distributed in Hunza Valley of Pakistan, Tashkurgan Valley, and Gaizi Valley of China, and the landslide scale is small (~3 km2) along the domestic section, while large (maximum area ~11 km2) along the Pakistan section. This dataset cataloged 762 landslides in the 10-km buffer zone on both sides of the China-Pakistan Karakoram Highway, including 57 complex landslides, 126 collapse types, 167 debris flows, and 412 unstable slopes. The data set includes vector boundary files taking the 10-km buffer zone of China-Pakistan Karakoram Highway as the study area; Landslide catalog vector file, documents of attribution characteristics of each landslide, its attribute parameters (fields) include ID, type, acreage, longitude, latitude, relief, slope, aspect, lithology, precipitation, NDVI, distance to fault, distance to the epicenter, distance to drainage, distance to KKH, etc. The landslide inventorying data were obtained by indoor remote sensing monitoring and interpretation, field investigation verification and correction, and the remote sensing identification work was evaluated. It was found that 78.7% of the deformation slopes identified by InSAR and optical remote sensing monitoring were verified accurately. This dataset is the latest one of landslide inventory and characteristics along the whole route of the China-Pakistan Karakoram Highway. The data are of a good current situation and is necessary data for in-depth quantitative study of landslide disasters and risks along the China-Pakistan Karakoram Highway and even the China-Pakistan Economic Corridor in the future.