Evaluation of rainfall threshold models for debris flow initiation in the Jiangjia Gully,Yunnan Province,China

Systematically determining the discriminatory power of various rainfall properties and their combinations in identifying debris flow occurrence is crucial for early warning systems.In this study,we evaluated the discriminatory power of different univariate and multivariate rainfall threshold models in identifying triggering conditions of debris flow in the Jiangjia Gully,Yunnan Province,China.The univariate models used single rainfall properties as indicators,including total rainfall(R_(tot)),rainfall duration(D),mean intensity(I_(mean)),absolute energy(Eabs),storm kinetic energy(E_(s)),antecedent rainfall(R_(a)),and maximum rainfall intensity over various durations(I_(max_dur)).The evaluation reveals that the I_(max_dur)and Eabs models have the best performance,followed by the E_(s),R_(tot),and I_(mean)models,while the D and R_(a)models have poor performances.Specifically,the I_(max_dur)model has the highest performance metrics at a 40-min duration.We used logistic regression to combine at least two rainfall properties to establish multivariate threshold models.The results show that adding D or R_(a)to the models dominated by Eabs,E_(s),R_(tot),or I_(mean)generally improve their performances,specifically when D is combined with I_(mean)or when R_(a)is combined with Eabs or E_(s).Including R_(a)in the I_(max_dur)model,it performs better than the univariate I_(max_dur)model.A power-law relationship between I_(max_dur)and R_(a)or between Eabs and R_(a)has better performance than the traditional I_(mean)–D model,while the performance of the E_(s)–R_(a)model is moderate.Our evaluation reemphasizes the important role of the maximum intensity over short durations in debris flow occurrence.It also highlights the importance of systematically investigating the role of R_(a)in establishing rainfall thresholds for triggering debris flow.Given the regional variations in rainfall patterns worldwide,it is necessary to evaluate the findings of this study across diverse watersheds.

Climate Variability & Establishment of Rainfall Threshold Line for Landslide Hazards in Rangamati, Bangladesh

This study aims to evaluate the impact of extreme rainfall events on landslides under current and past climate scenarios. Rainfall-triggered landslides are analyzed by rainfall estimates, derived using statistics of events. It is established that recent climate changes, mainly temperature and rainfall patterns have significantly increased the rainfall-induced landslide hazards in the Rangamati district, Bangladesh. It is also observed that the temperature and rainfall of Rangamati had increased gradually during the last 40 years (1981-2021). On 13 June 2017, a series of landslides triggered by heavy monsoon rains (300 mm/24 h) occurred and killed more than 112 people in the Rangamati hill district, Bangladesh. The highest annual decade rainfall is 3816 mm, recorded in 2010-21. A relationship between causalities and the number of events has also been established. The analysis shows that both antecedent and single-day major rainfall patterns can influence sliding events. It is established that monsoonal rainfall (June-September) can significantly influence catastrophic landslide hazard events. Finally, two rainfall threshold lines for the researched area are constructed based on antecedent and single-day major rainfall occurrences, as well as the number of fatalities caused by landslides. Total rainfall of 100 mm (16.66 mm/day) during six days appears to define the minimum rainfall that has led to shallow landslides/slope failures, while 210 mm (35 mm/day) within six days appears to define the lowest rainfall that could be a cause of catastrophic landslide in Rangamati district.

Estimation of rainfall thresholds for shallow landslides in the Sierra Madre Oriental, northeastern Mexico

Landslides induced by prolonged rainfalls are frequent mass movements along the northeastern portion of the Sierra Madre Oriental in Mexico,causing significant damage to infrastructure.In this work,we have studied the connection between rainfall and landslides in the Santa Rosa Canyon,a catchment located in the northeastern Mexico.A landslide database triggered by major storms and hurricanes that have hit the region over the past 30 years was analyzed.A total of 92 rainfall events in the Santa Rosa Canyon were studied to determine the amount of precipitation needed to trigger shallow landslides.For each event the duration(D,in hours)and the cumulated rainfall event(E,in mm)were determined by using historical rainfall data from weather stations located near the study area.We have proposed an ED threshold for rainfall-induced landslides with durations 0.5<D<120 hours to address the conditions that trigger these events in the study area.On analyzing the obtained threshold,it has been established that almost 60 mm of a daily rainfall accumulation is required to trigger shallow landslides in the study area.This estimation is consistent with other calculations made for areas close to the Santa Rosa Canyon.Finally,we validated the predictive capability of the threshold with a different set of rainfall data that did not result in landslides performing a straightforward receiver operating characteristic analysis.A good approach was obtained,especially for rainfall events with daily measurements.Results could be used as input information in the design of a landslide early warning system for the northeastern Mexico,and replicated for other landslide prone areas in the region.

A multiobjective evolutionary optimization method based critical rainfall thresholds for debris flows initiation

At present,most researches on the critical rainfall threshold of debris flow initiation use a linear model obtained through regression.With relatively weak fault tolerance,this method not only ignores nonlinear effects but also is susceptible to singular noise samples,which makes it difficult to characterize the true quantization relationship of the rainfall threshold.Besides,the early warning threshold determined by statistical parameters is susceptible to negative samples(samples where no debris flow has occurred),which leads to uncertainty in the reliability of the early warning results by the regression curve.To overcome the above limitations,this study develops a data-driven multiobjective evolutionary optimization method that combines an artificial neural network(ANN)and a multiobjective evolutionary optimization implemented by particle swarm optimization(PSO).Firstly,the Pareto optimality method is used to represent the nonlinear and conflicting critical thresholds for the rainfall intensity I and the rainfall duration D.An ANN is used to construct a dual-target(dual-task)predictive surrogate model,and then a PSO-based multiobjective evolutionary optimization algorithm is applied to train the ANN and stochastically search the trained ANN for obtaining the Pareto front of the I-D surrogate prediction model,which is intended to overcome the limitations of the existing linear regression-based threshold methods.Finally,a double early warning curve model that can effectively control the false alarm rate and negative alarm rate of hazard warnings are proposed based on the decision space and target space maps.This study provides theoretical guidance for the early warning and forecasting of debris flows and has strong applicability.

An hourly shallow landslide warning model developed by combining automatic landslide spatial susceptibility and temporal rainfall threshold predictions

Landslide warning models are important for mitigating landslide risks.The rainfall threshold model is the most widely used early warning model for predicting rainfall-triggered landslides.Recently,the rainfall threshold model has been coupled with the landslide susceptibility(LS)model to improve the accuracy of early warnings in the spatial domain.Existing coupled models,designed based on a matrix including predefined rainfall thresholds and susceptibility levels,have been used to determine the warning level.These predefined classifications inevitably have subjective rainfall thresholds and susceptibility levels,thus affecting the probability distribution information and eventually influencing the reliability of the produced early warning.In this paper,we propose a novel landslide warning model in which the temporal and spatial probabilities of landslides are coupled without predefining the classified levels.The temporal probability of landslides is obtained from the probability distribution of rainfall intensities that triggered historical landslides.The spatial probability of landslides is then obtained from the susceptibility probability distribution.A case study shows that the proposed probability-coupled model can successfully provide hourly warning results before the occurrence of a landslide.Although all three models successfully predicted the landslide,the probability-coupled model produced a warning zone comprising the fewest grid cells.Quantitatively,the probabilitycoupled model produced only 39 grid cells in the warning zone,while the rainfall threshold model and the matrix-coupled model produced warning zones including 81 and 49 grid cells,respectively.The proposed model is also applicable to other regions affected by rainfall-induced landslides and is thus expected to be useful for practical landslide risk management.

Rainfall Threshold Calculation Method for Debris Flow Pre-Warning in Data-Poor Areas

Debris flows are the one type of natural disaster that is most closely associated with hu- man activities. Debris flows are characterized as being widely distributed and frequently activated. Rainfall is an important component of debris flows and is the most active factor when debris flows oc- cur. Rainfall also determines the temporal and spatial distribution characteristics of the hazards. A reasonable rainfall threshold target is essential to ensuring the accuracy of debris flow pre-warning. Such a threshold is important for the study of the mechanisms of debris flow formation, predicting the characteristics of future activities and the design of prevention and engineering control measures. Most mountainous areas have little data regarding rainfall and hazards, especially in debris flow forming re- gions. Therefore, both the traditional demonstration method and frequency calculated method cannot satisfy the debris flow pre-warning requirements. This study presents the characteristics of pre-warning regions, included the rainfall, hydrologic and topographic conditions. An analogous area with abundant data and the same conditions as the pre-warning region was selected, and the rainfall threshold was calculated by proxy. This method resolved the problem of debris flow pre-warning in ar- eas lacking data and provided a new approach for debris flow pre-warning in mountainous areas.

Deformation characteristics and thresholds of the Tanjiawan landslide in the Three Gorges Reservoir Area,China

Since the first impoundment of the Three Gorges Reservoir(TGR)in China in 2003,more than 5000 landslides including potential landslides were identified.In this paper,a deep-seated active landslide in TGR area was analyzed.Fourteen years’monitoring data and field investigations from 2006 to 2020 were used to analyze the deformation characteristics,influencing factors,and meteohydrological thresholds.The landslide showed a none-overall periodic movement pattern featuring acceleration during long-duration rainfall and rapid transition to constant creep after rainfall events.Two secondary sliding masses,No.1 and No.2,were defined via field investigation.The reservoir has no impact on the deformation whereas long-duration-low-intensity rainfall is the main factor.At present,the cumulative displacements of the main sliding mass range from 0.9 to 3.2 m,and the deformation during the rainy season is gradually increasing.The boundary of this landslide was formed,and the boundary of No.2 sliding mass became obvious.The probability of the failure of sliding mass No.2 is very high under the conditions of continuous rainfall.The 15-day antecedent rainfall combined with 4-day cumulative rainfall could be the rainfall threshold which could be associated with the groundwater level S1 of 294 m above sea level for forecasting large deformation of Tanjiawan landslide.

Field observation of debris-flow activities in the initiation area of the Jiangjia Gully, Yunnan Province, China

The Jiangjia Gully, which is located in Dongchuan District, Yunnan Province, China, is a watershed prone to debris flows and has long-term recorded data of debris-flow occurrence. However, the initiation mechanism has mainly been studied by experiments in this watershed. To further reveal debris-flow formation mechanism in the Jiangjia Gully, debris-flow activities in the initiation zone were observed with hand-held video cameras in the summer of 2016 and 2017. In these two years, six debris-flow events were triggered in Menqian Gully, a major tributary of the Jiangjia Gully, while debrisflow activities in some sub-watersheds of Menqian Gully were recorded with video cameras in four events. The video recording shows that landslides constituted an important source for sediment supply in debris flow. Some landslides directly evolved into debris flows, while the others released sediment into rills and channels, where debris flows were generated for sediment entrainment by water flow. Therefore, debris-flow occurrence in the Jiangjia Gully is influenced both by infiltration-dominated processes and by runoff-dominated processes. In addition, rainfall data from four gauges installed in Menqian Gully were analyzed using mean intensity(I), duration(D), peak 10-minute rainfall(R10min) and antecedent rainfall(AR) up to 15 days prior to peak 10-minute rainfall. It reveals that debris-flow triggering events can be discriminated from nontriggering events either by an I-D threshold or by an R10min-AR threshold. However, false alarms can be greatly reduced if these two kinds of thresholds are used together. Moreover, behaviors including intermittency of debris flow, variance in moisture content and volume among surges, and coalescence of multiple surges by temporary damming were observed, indicating the complexity of debris-flow initiation processes. These findings are expected to enhance our knowledge on debris-flow formation mechanism in regions with similar environmental settings.