Landslide Information Research Articles

Co-seismic landslide mapping using Sentinel-2 10-m fused NIR narrow, red-edge, and SWIR bands

An accurate and timely inventory for major landslides triggered by seismic events is essential for hazard assessment and risk governance. Sentinel-2 MultiSpectral Instrument (MSI) data have been increasingly used for co-seismic landslide mapping in recent years. However, there have been few studies exploring the full use of Sentinel-2 near infrared (NIR) narrow, red-edge, and short-wave infrared (SWIR) bands for landslide mapping. In this study, Sentinel-2 10-m fused NIR narrow, red-edge-1, SWIR-1, and SWIR-2 bands were incorporated into the change detection-based Markov random field (CDMRF) that considers the spectral, spatial, and contextual information of landslides. The effectiveness of the developed method was proved in mapping co-seismic landslides triggered by two major earthquakes in the Yarlung Tsangpo Grand Canyon, China, in 2017 and Hokkaido, Japan, in 2018. The results demonstrated that the mapping performances with the fused Sentinel-2 bands were significantly improved for both seismic events. Such improvement is mainly attributed to the more abundant spectral information provided by the fused Sentinel-2 red-edge, NIR narrow, and SWIR bands. The novelties of this study are (1) the preliminary attempt to use the Sentinel-2 10-m fused bands for landslide mapping, and (2) the synergy of Sentinel-2 red-edge, NIR narrow, and SWIR bands for more reliable landslide mapping.

Study on Extraction of Landslide Information Based on UAV Survey

Traditional analysis of landslide evolution discipline requires particular on-site investigation, which not only consumes a lot of manpower and material resources, but is also very inefficient. With the development of small drone aerial photography technology, a method based on the whole area landslide measurement becomes possible. This paper uses the rapid development of digital photogrammetry technology in recent years to monitor the surface deformation of the Qinyu landslide, and conducts an in-depth study of its evolution. The results show that the evolution of the Qinyu landslide is inherited and different in time and space. The reconstruction of the three-dimensional model greatly simplifies the traditional on-site investigation, and it also provides new methods and ideas for landslide monitoring.

Nepalese landslide information system (NELIS): a conceptual framework for a web-based geographical information system for enhanced landslide risk management in Nepal

Abstract. Comprehensive and sustainable landslide risk management, including the identification of areas susceptible to landslides, requires responsible organisations to collaborate efficiently. Landslide risk management efforts are often made after major triggering events, such as hazard mitigation after the 2015 Gorkha earthquake in Nepal. There is also a lack of knowledge sharing and collaboration among stakeholders to cope with major disaster events, in addition to a lack of efficiency and continuity. There should be a system to allow for landslide information to be easily updated after an event. For a variety of users of landslide information in Nepal, the availability and extraction of landslide data from a common database are a vital requirement. In this study, we investigate the requirements to propose a concept for a web-based Nepalese landslide information system (NELIS) that provides users with a platform to share information about landslide events to strengthen collaboration. The system will be defined as a web GIS (geographic information system) that supports responsible organisations in addressing and managing different user requirements of people working with landslides, thereby improving the current state of landslide hazard and risk management in Nepal. The overall aim of this study is to propose a conceptual framework and design of NELIS. A system like NELIS could benefit stakeholders involved in data collection and landslide risk management in their efforts to report and provide landslide information. Moreover, such a system would allow for detailed and structured landslide documentation and consequently provide valuable information regarding susceptibility and hazard and risk mapping. For the reporting of landslides directly to the system, a web portal is proposed. Based on field surveys, a literature review and stakeholder interviews, a structure of the landslide database and a conceptual framework for the NELIS platform are proposed.

An Optimized Random Forest Model and Its Generalization Ability in Landslide Susceptibility Mapping: Application in Two Areas of Three Gorges Reservoir, China

Numerous researches have been published on the application of landslide susceptibility assessment models; however,they were only applied in the same areas as the models were originated,the effect of applying the models to other areas than the origin of the models has not been explored. This study is purposed to develop an optimized random forest (RF) model with best ratios of positive-to-negative cells and 10-fold cross-validation for landslide susceptibility mapping (LSM),and then explore its generalization ability not only in the area where the model is originated but also in area other than the origin of the model. Two typical counties (Fengjie County and Wushan County) in the Three Gorges Reservoir area,China,which have the same terrain and geological conditions,were selected as an example. To begin with,landslide inventory was prepared based on field investigations,satellite images,and historical records,and 1 522 landslides were then identified in Fengjie County. 22 landslide-conditioning factors under the influence of topography,geology,environmental conditions,and human activities were prepared. Then,combined with 10-fold cross-validation,three typical ratios of positive-to-negative cells,i.e.,1:1,1:5,and 1:10,were adopted for comparative analyses. An optimized RF model (Fengjie-based model) with the best ratios of positive-to-negative cells and 10-fold cross-validation was constructed. Finally,the Fengjie-based model was applied to Fengjie County and Wushan County,and the confusion matrix and area under the receiver operating characteristic (ROC) curve value (AUC) were used to estimate the accuracy. The Fengjie-based model delivered high stability and predictive capability in Fengjie County,indicating a great generalization ability of the model to the area where the model is originated. The LSM in Wushan County generated by the Fengjie-based model had a reasonable reference value,indicating the Fengjie-based model had a great generalization ability in area other than the origin of the model. The Fengjie-based model in this study could be applied in other similar areas/countries with the same terrain and geological conditions,and a LSM may be generated without collecting landslide information for modeling,so as to reduce workload and improve efficiency in practice.

Susceptibility Mapping on Urban Landslides Using Deep Learning Approaches in Mt. Umyeon

In recent years, the incidence of localized heavy rainfall has increased as abnormal weather events occur more frequently. In densely populated urban areas, this type of heavy rain can cause extreme landslide damage, so that it is necessary to estimate and analyze the susceptibility of future landslides. In this regard, deep learning (DL) methodologies have been used to identify areas prone to landslides recently. Therefore, in this study, DL methodologies, including a deep neural network (DNN), kernel-based DNN, and convolutional neural network (CNN) were used to identify areas where landslides could occur. As a detailed step for this purpose, landslide occurrence was first determined as landslide inventory through aerial photographs with comparative analysis using field survey data; a training set was built for model training through oversampling based on the landslide inventory. A total of 17 landslide influencing variables that influence the frequency of landslides by topography and geomorphology, as well as soil and forest variables, were selected to establish a landslide inventory. Then models were built using DNN, kernel-based DNN, and CNN models, and the susceptibility of landslides in the study area was determined. Model performance was evaluated through the average precision (AP) score and root mean square error (RMSE) for each of the three models. Finally, DNN, kernel-based DNN, and CNN models showed performances of 99.45%, 99.44%, and 99.41%, and RMSE values of 0.1694, 0.1806, and 0.1747, respectively. As a result, all three models showed similar performance, indicating excellent predictive ability of the models developed in this study. The information of landslides occurring in urban areas, which cause a great damage even with a small number of occurrences, can provide a basis for reference to the government and local authorities for urban landslide management.

Application of a GIS-Based Slope Unit Method for Landslide Susceptibility Mapping in Helong City: Comparative Assessment of ICM, AHP, and RF Model

Landslides are one of the most extensive geological disasters in the world. The objective of this study was to assess the performances of different landslide susceptibility models information content method (ICM), analytical hierarchy process (AHP), and random forest (RF) model) and mapping unit (slope unit and grid unit) for landslide susceptibility mapping in the Helong city, Jilin province, northeastern China. First, a total of 159 landslides were mapped in the study area based on a geological hazard survey (1:50,000) of Helong city. Then, the slope units of the study area were divided by using the curvature watershed method. Next, eight influencing factors, namely, lithology, slope angle, slope aspect, rainfall, land use, seismic intensity, distance to river, and distance to fault, were selected to map the landslide susceptibility based on geological data, field survey, and landslide information. Afterward, landslide susceptibility modeling of landslide inventory data is performed for extracting and learning the symmetry latent in data patterns and relationships by three landslide susceptibility models and utilizing it to predict landslide susceptibility. Finally, the receiver operating characteristic (ROC) curve was used to compare the landslide susceptibility models. In addition, results based on grid units were calculated for comparison. The AUC (the area under the curve) result for ICM, AHP, and RF model was 87.1%, 80.5%, and 94.6% for slope units, and 83.4%, 70.9%, and 91.3% for grid units, respectively. Based on the overall assessments, the SU-RF model was the most suitable model for landslide susceptibility mapping. Consequently, these methods can be very useful for landslide hazard mitigation strategies.

Integration of Spatial Probability and Size in Slope-Unit-Based Landslide Susceptibility Assessment: A Case Study.

Landslide spatial probability and size are two essential components of landslide susceptibility. However, in existing slope-unit-based landslide susceptibility assessment methods, landslide size has not been explicitly considered. This paper developed a novel slope-unit based approach for landslide susceptibility assessment that explicitly incorporates landslide size. This novel approach integrates the predicted occurrence probability (spatial probability) of landslides and predicted size (area) of potential landslides for a slope-unit to obtain a landslide susceptibility value for that slope-unit. The results of a case study showed that, from a quantitative point of view, integrating spatial probability and size in slope-unit-based landslide susceptibility assessment can bring remarkable increases of AUC (Area under the ROC curve) values. For slope-unit-based scenarios using the logistic regression method and the neural network method, the average increase of AUC brought by incorporating landslide size is up to 0.0627 and 0.0606, respectively. Slope-unit-based landslide susceptibility models incorporating landslide size had utilized the spatial extent information of historical landslides, which was dropped in models not incorporating landslide size, and therefore can make potential improvements. Nevertheless, additional case studies are still needed to further evaluate the applicability of the proposed approach.

DFOS-based association rules analysis on the multi-fields information of Majiagou landslide

The multi-fields information of landslide can be obtained using distributed optical fiber sensing (DFOS) technology. Nevertheless, it is still a quite complicated work to analyze massive amounts of multi-fields information, and to evaluate the slope stability under multifields coupling effect. This paper introduced a data mining method named association rules analysis, and designed a DFOS-based monitoring system to record the multi-field information, including strain field, deformation field, seepage field, temperature field and environmental parameters, respectively. Taking Majiagou landslide as an example, the association rules analysis of the multi-field monitoring information revealed the reservoir water level is the domain factor that influences the deformation of Majiagou landslide. The correlation analysis method can explore the internal relationship among different fields of a landslide and provide scientific basis for slope stability evaluation and landslide early-warning.

Research, development, and field trial of the universal Global Navigation Satellite System receivers

Global Navigation Satellite System (GNSS) can provide continuous, high precision, and absolute displacement information on landslides. It can thus be used in landslide monitoring applications. However, the cost of high-precision GNSS is too high for large-scale deployment. From an application perspective, we reviewed the status and challenges of applying GNSS technology to landslide monitoring. We then investigated the data specification for a landslide monitoring GNSS receiver to better meet specified engineering requirements. We further developed prototype universal GNSS monitoring devices and deployed them in the Heitai area for testing. Approximately two months of testing indicated that the prototype universal GNSS devices well meet the engineering requirements, and the tested GNSS data are in good agreement with the crack data. The hourly monitoring GNSS solution achieves sub-millimeter level accuracy and can be used for extracting landslide information.

National-scale landslide susceptibility map of Romania in a European methodological framework

Landslide susceptibility maps are useful tools for risk analysis and assessment with practical implications because they provide relevant information for territorial planning, land use sustainable management or even forecast and early warning systems. Achievement of accurate assessments of landslide susceptibility for large regions (i.e. including national territories) is still a challenge, mainly because of the lack of proper landslide inventory and monitoring data. Romania represents one of the most landslide-affected countries in Europe. The current study presents an approach for drawing the landslide susceptibility map at a national scale for the Romanian territory, in agreement with the European methodological framework promoted for small-scale evaluations of landslide susceptibility. The methodological approach was adapted to the specific mophostructural, climatic and landuse conditions of the country, as well as to the quantity and quality of the available data, in order to achieve a susceptibility zonation for slides and flows for the national territory. It follows a mixed statistical-heuristic approach based on a Spatial Multi-Criteria Evaluation (SMCE) procedure integrating both landslide information and expert knowledge. The national landslide susceptibility map outlines large areas ranked as having high and very high susceptibility throughout the Subcarpathian chain, the Moldavian and Transylvanian Plateaux and the Getic Piedmont. The prediction performance was examined quantitatively and qualitatively, by making use of regional geomorphical knowledge. The evaluations suggest that, despite uncertainties inherent at this analysis scale, spatially-differentiated models are able to better capture landslide conditioning frameworks and reproduce inter- and, especially, intraregional variability of landslide distribution as compared to a previous version of the national susceptibility map. The study proves that combining statistical and heuristic approaches, calibrated and later on validated for distinct homogeneous morpho-litho-structural units allows to increase the prediction capacity of the national-scale model. The results are useful to public authorities at national, regional, county and municipality levels, providing knowledge for the enhancement of disaster prevention and response plans.

Information Literacy Training on the Preparedness of Garut People in Facing Landslide Threats

The purpose of this study was to analyze the impact of information literacy on the preparedness of the people of Garut Regency in facing the threat of landslide. The data ini this study are the people of Garut Regency who have been affected by landslides based on the 2010 population census based on gender. The research method uses quantitative patterns in the form of surveys. The results showed that the magnitude of the impact of the ability to identify and find the location of information on public awareness in facing the threat of landslides by 38%. The impact of landslide information literacy can increase the preparedness of the people of Garut Regency by 57%. The conclusion is the impact of information literacy has a significant impact on the preparedness of the people of Garut Regency in facing the threat of landslides. 

Investigation of Slow-Moving Artificial Slope Failure with Multi-Temporal InSAR by Combining Persistent and Distributed Scatterers: A Case Study in Northern Taiwan

In Taiwan, landslides pose serious threats to local residents and infrastructures each year due to high mountain relief and distinct seasonal precipitation distribution. Interferometric synthetic aperture (InSAR) provides a powerful tool to map terrain motion and characterize the failure mechanism of unstable slopes. However, it is challenging for the conventional InSAR technique to obtain reliable landslide information in mountainous regions because of insufficient coherent measurements and signal confusion caused by vegetation coverage and rugged terrain. In this study, we adopt an optimized multi-temporal InSAR (MTInSAR) approach to analyze the surface displacement of an artificial side slope along Freeway No. 3, where a catastrophic landslide failure occurred on 25 April 2010, in northern Taiwan. To increase the spatial extent of the deformation signal, we integrate information from both persistent scatterers (PSs) and distributed scatterers (DSs). Topographic residual and height-dependent atmospheric delays are corrected by a component-based method and joint model estimation, respectively. The results reveal the existence of slope movement with a rate of about −30 mm/year prior to the landslide failure. Further analysis shows that the temporal behaviors of downslope movement are correlated with local precipitation. The study demonstrates the need to continuously monitor and verify the stability of artificial slopes to prevent and minimize the probability of a similar landslide occurrence in the future.

RETRACTED: The application of multi-sensor target tracking and fusion technology to the comprehensive early warning information extraction of landslide multi-point monitoring data

• Application of multi-sensor target tracking fusion technology. • Proposes a spatiotemporal registration method and Ensemble Kalman Filte. • The test simulation is carried out by selecting 30 monitoring data. Landslide disasters are the second largest geological disaster activity in addition to earthquake disasters. They cause economic losses of up to several billion yuan each year, resulting in thousands of deaths and injuries. Therefore, it is important to monitor and alert landslide disasters. Significance. In addition, after the occurrence of landslides, a large amount of deposits will be created, which will further breed mudslides. The ability to qualitatively and quantitatively extract landslide information at the first time is particularly important for disaster prevention, mitigation, and disaster management. Therefore, the extraction of landslide information is a significant measure. At the same time, due to the limitations of sensors, how to organically fuse data from multiple sensors to achieve tracking performance that cannot be achieved by a single sensor has become the focus of research and attention on multi-sensor data fusion technology in the field of target tracking. Based on the above background, the research content of this paper is the application of multi-sensor target tracking fusion technology to achieve comprehensive warning information extraction of landslide multi-point monitoring data. This paper proposes a spatiotemporal registration method and Ensemble Kalman Filter (EnKF) for target tracking Algorithm, and the prediction model is established through optimization problem and target tracking. Finally, the test simulation is carried out by selecting 30 monitoring data of a monitoring point of a new construction project as the detection value. The experimental results show that when the moving target model is unknown in actual work, the polynomial. The order of the fitting and the boundary conditions of the spline fitting are not easy to give, but you can optionally use the sensor measurement data in the middle part and then use the spline fitting algorithm to obtain a better time registration effect; In terms of spatial registration, in the Z-axis direction, the measurement error of sensor A and sensor B is large, and the maximum value exceeds 2000 m. The maximum value of sensor A error between the measured values and theoretical values of radar A and B after registration is about 1000 m, which is the Z axis direction, and the maximum error of sensor B is about 1000 m, which is also the Z axis direction. The error has been significantly reduced. EnKF is better than the UKF algorithm under non-Gaussian noise conditions. The PF algorithm itself is suitable for nonlinear and non-Gaussian systems and has the best tracking performance.

Towards a web service for rapid landslide mapping based on Copernicus data

Timely available landslide information is highly valuable for organizing emergency response activities and infrastructure restoration. The collection process for landslide information can benefit from the use of Earth observation (EO) data because they are capable of sensing potentially affected areas quickly and comprehensively. Challenges for an EO-based rapid landslide mapping service lie in developing a reliable and fast information provision process and in appropriate integration of landslide information products in the workflows and working environment of first responders. Our gap analysis for a developed pre-operational web service for EO-based landslide mapping identified a need to implement new components to meet these requirements. These comprise better linkage to the Copernicus data hub, increasing the degree of automation in the landslide information extraction process with machine learning technologies, and improving and testing user interaction with the service.

Assessment of the Sentinel-1 based ground motion data feasibility for large scale landslide monitoring

In this paper, a systematic procedure to assess the feasibility of Advanced Differential Interferometric SAR (A-DInSAR) technique for landslide monitoring using SAR images acquired by Sentinel-1 sensors is presented. The methodology is named “Assessment of the advanced differentiaL interferometric synthetic aperture radar technique Feasibility for large scale lAndslide monitoring – ALFA” and it is structured in two main phases, which includes pre-processing and post-processing elaborations. The methodology was developed and tested in the Alpine sector of the Piedmont region in Italy, which represents a landslide prone area. In particular, ALFA represents a methodology based on previous algorithms available in the literature to assess the a-prior feasibility assessment and post-processing analysis of A-DInSAR data for landslide, which introduces three novel aspects such as (1) a systematic scheme suitable within regional practices; (2) the use of Sentinel-1 data and the development of (3) an index to take into account of the kind of distribution of the measuring points along the landslide. The approach was applied over an area extended about 5300 km2 affected by 5703 landslides mapped in the database of the Piedmont Region (Landslides information system in Piedmont—SIFRAP). Sentinel-1 images covering the period 2014–2018 were analysed. The results show the potential of the Sentinel-1 data for large-scale landslide monitoring. The developed methodology presents reliable tools that could be useful as feasibility for the use of Sentinel-1 data for landslide monitoring at regional and national scale.

Research on Post-Earthquake Landslide Extraction Algorithm Based on Improved U-Net Model

Seismic landslides are the most common and highly destructive earthquake-triggered geological hazards. They are large in scale and occur simultaneously in many places. Therefore, obtaining landslide information quickly after an earthquake is the key to disaster mitigation and relief. The survey results show that most of the landslide-information extraction methods involve too much manual participation, resulting in a low degree of automation and the inability to provide effective information for earthquake rescue in time. In order to solve the abovementioned problems and improve the efficiency of landslide identification, this paper proposes an automatic landslide identification method named improved U-Net model. The intelligent extraction of post-earthquake landslide information is realized through the automatic extraction of hierarchical features. The main innovations of this paper include the following: (1) On the basis of the three RGB bands, three new bands, DSM, slope, and aspect, with spatial information are added, and the number of feature parameters of the training samples is increased. (2) The U-Net model structure is rebuilt by adding residual learning units during the up-sampling and down-sampling processes, to solve the problem that the traditional U-Net model cannot fully extract the characteristics of the six-channel landslide for its shallow structure. At the end of the paper, the new method is used in Jiuzhaigou County, Sichuan Province, China. The results show that the accuracy of the new method is 91.3%, which is 13.8% higher than the traditional U-Net model. It is proved that the new method is effective and feasible for the automatic extraction of post-earthquake landslides.

Beyond Artificial Reality

With billions of active social media accounts and millions of live video cameras, live new big data offer many opportunities for smart applications. However, the main consumers of the new big data have been humans. We envision the research on live knowledge , to automatically acquire real-time, validated, and actionable information. Live knowledge presents two significant and diverging technical challenges: big noise and concept drift. We describe the EBKA (evidence-based knowledge acquisition) approach, illustrated by the LITMUS landslide information system. LITMUS achieves both high accuracy and wide coverage, demonstrating the feasibility and promise of EBKA approach to achieve live knowledge.

Regional susceptibility assessments with heterogeneous landslide information: Slope unit- vs. pixel-based approach

Regional landslide inventories are often prepared by several different experts, using a variety of data sources. This can result in a combination of polygon and point landslide data, characterized by different meanings, uncertainties and levels of reliability. The propagation of uncertainties due to such heterogeneous data is a relevant issue in statistical landslide susceptibility zonation at supra-local scale. In the inhabited highlands of the Rwenzori Mountains, we compare different approaches and mapping units to provide a robust methodology for susceptibility mapping using a combination of landslide point and polygon data. First, the effect of the uncertainty related to a point representation of landslides is assessed comparing slope unit-based and pixel-based analyses, using digital elevation models with different resolutions. Secondly, with regard to landslide polygon inventories, we compare the use of thresholds versus a presence/absence of the depletion centroid or a randomly selected point in the landslide polygon in order to identify slope units with landslides. Based on these results, we prepare regional slope unit-based susceptibility maps using a logistic regression model calibrated with the landslide polygon inventory and validated with the point inventory. Although pixel-based mapping remains the most common approach for statistical landslide susceptibility zonation, our analysis clearly favours the use of slope units as a powerful tool to prepare regional susceptibility maps and, in particular, to exploit heterogeneous information in a consistent way.

Mapping Landslides on EO Data: Performance of Deep Learning Models vs. Traditional Machine Learning Models

Mapping landslides using automated methods is a challenging task, which is still largely done using human efforts. Today, the availability of high-resolution EO data products is increasing exponentially, and one of the targets is to exploit this data source for the rapid generation of landslide inventory. Conventional methods like pixel-based and object-based machine learning strategies have been studied extensively in the last decade. In addition, recent advances in CNN (convolutional neural network), a type of deep-learning method, has been widely successful in extracting information from images and have outperformed other conventional learning methods. In the last few years, there have been only a few attempts to adapt CNN for landslide mapping. In this study, we introduce a modified U-Net model for semantic segmentation of landslides at a regional scale from EO data using ResNet34 blocks for feature extraction. We also compare this with conventional pixel-based and object-based methods. The experiment was done in Douglas County, a study area selected in the south of Portland in Oregon, USA, and landslide inventory extracted from SLIDO (Statewide Landslide Information Database of Oregon) was considered as the ground truth. Landslide mapping is an imbalanced learning problem with very limited availability of training data. Our network was trained on a combination of focal Tversky loss and cross-entropy loss functions using augmented image tiles sampled from a selected training area. The deep-learning method was observed to have a better performance than the conventional methods with an MCC (Matthews correlation coefficient) score of 0.495 and a POD (probability of detection) rate of 0.72 .

Using a Tank Model to Determine Hydro-Meteorological Thresholds for Large-Scale Landslides in Taiwan

Rainfall thresholds for slope failures are essential information for establishing early-warning systems and for disaster risk reduction. Studies on the thresholds for rainfall-induced landslides of different scales have been undertaken in recent decades. This study attempts to establish a warning threshold for large-scale landslides (LSLs), which are defined as landslides with a disturbed area more massive than 0.1 km2. The numerous landslides and extensive rainfall records make Taiwan an appropriate area to investigate the rainfall conditions that can result in LSLs. We used landslide information from multiple sources and rainfall data captured by 594 rain gauges to create a database of 83 rainfall events associated with LSLs in Taiwan between 2001 and 2016. The corresponding rainfall duration, cumulative event rainfall, and rainfall intensity for triggering LSLs were determined. This study adopted the tank model to estimate conceptual water depths (S1, S2, S3) in three-layer tanks and calculated the soil water index (SWI) by summing up the water depths in the three tanks. The empirical SWI and duration (SWI–D) threshold for triggering LSLs occurring during 2001–2013 in Taiwan is determined as SWI = 155.20 − 1.56D and D ≥ 24 h. The SWI–D threshold for LSLs is higher than that for small-scale landslides (SSLs), those with a disturbed area smaller than 0.1 km2. The LSLs that occurred during 2015–2016 support this finding. It is notable that when the SWI and S3 reached high values, the potential of LSLs increased significantly. The rainfall conditions for triggering LSLs gradually descend with increases in antecedent SWI. Unlike the rainfall conditions for triggering SSLs, those for triggering LSLs are related to the long duration–high intensity type of rainfall event.