In recent years, large scale debris flows have occurred frequently in mountainous areas as a result of strong earthquakes, extreme rainfall, and increased human activities. These debris flows often block main river channels and form a multi-hazard chain of rivers dammed by debris flow followed by outburst flooding, which has caused huge losses of life and property worldwide [1
]. In particular, under the complex geological and climatic conditions of the Qinghai-Tibetan Plateau, events have included the rivers dammed by the 2020 Danba debris flow and the 2019 Yarlung Tsangpo debris flow [7
]. Debris flow dams are mostly formed by the accumulation of loose materials, and are much less stable than artificial dams; once debris flow dams breach, they may cause more serious flood disasters downstream. Therefore, rapid evaluation of their stability and the risk of dam failure is particularly important for the emergency response and mitigation of the hazard chain induced by debris flows.
The Bailong River basin is in the confluence zone of the Loess Plateau, the Qingha-Tibetan Plateau, and the Sichuan Basin. It has strong tectonic activities and is one of the four regions that suffer the most from geological hazards in China [9
]. Debris flow disasters occur frequently in the Bailong River basin, which has the characteristics of “every earthquake is inevitable, and prolonged rain disasters” [11
]. Based on the latest survey, there are 1008 debris flow valleys in the Bailong River basin. More than 100 of these are large debris flows that pose a high risk of damming rivers. In 2010, a huge mudslide in Zhouqu blocked the Bailong River. The mudslide inundated 1/3 of the main city, and led to 1765 deaths and a great loss of property. Key issues in the emergency response and risk management of the river damming debris-flow hazard chain are whether a barrier dam is stable and the extent of the impact of an outburst flood caused by a dam breach. Therefore, there is an urgent need to quantitatively predict the dynamic evolution process of dam breaches and the subsequent outburst flood under different scenarios within a short time period for the purpose of emergency decision making.
Barrier dams and related dam break flood disasters are common all over the world, such as in China [12
], Italy [13
], and Central Asia [14
]. Therefore, many scholars have conducted relevant research on debris flow barrier dams. The research on debris flow barrier dams is mainly divided into model experiments and numerical model studies. Researchers have conducted model experiments by reducing the barrier dam to a certain scale [15
]. Dang et al. [16
] used the debris flow blockage event in Tibet as a generalized model to carry out overtop failure experiments for debris flow barrier dams, and studied the process of debris flow barrier dam failure by changing the dam’s body parameters and the hydraulic parameters of the main trench. Li et al. [17
] used scale paper and cameras to record the formation, development, and change process of failure under different conditions for the main factors influencing dam breaks. These studies have all contributed to the study of debris flow barrier dams, but due to the difficulty of model experiments, numerical simulations have become a more important method in the study of debris flow [18
]. With the widespread use of computers, researchers have carried out a large number of quantitative studies on the numerical modeling of dam breaks and outburst flood evolution, which mainly includes parameter-based models and physical process-based models [19
]. For example, Singh et al. [20
] analyzed 20 dam-break cases and made a quantitative assessment of the width of the failure for the first time. Chen et al. [21
] proposed a hyperbolic model of soil erosion and successfully applied it to the failure analysis of the Tangjiashan barrier dam. Fu et al. [22
] developed a simulation model of the overtopping outburst process of a dam based on the physical mechanism. On the basis of the research on the process of barrier dam breakage, scholars have conducted in-depth research into the flood flow of dam failures. Xu and Zhang [23
], Thornton [24
], and Hooshyaripor [25
] respectively established different forms of relational expressions for predicting dam-break flood flow. Numerous mature numerical models have been applied to simulate barrier dam breaks. These studies and models provide an important reference for the quantitative study of debris flow dam failure. However, these studies either focus on the simulation of the dam break process or on dam break flood. Limited quantitative studies based on numerical models have been undertaken regarding the complete debris flow-dam breach-outburst flood process. There are few rapid quantitative evaluation methods for the risk of outburst floods due to dam breaks caused by debris flows.
The occurrence of many major natural disasters is often accompanied by the occurrence of other disasters, and the losses are not caused by a certain kind of disaster, but by the chain reaction of multiple disasters and their complex interactions in time and space. Therefore, the problem of the disaster chain has gradually become a major issue in disaster science [26
]. In the study of debris flow disaster chain, there are few rapid quantitative evaluation methods for determining the risk of outburst floods due to dam breaks caused by debris flows. Many models, such as Hydrologic Engineering Center’s River Analysis System (HEC-RAS) [28
], a modelling system for River and Channels (MIKE 11) [29
], and InfoWorks RS (ISIS) [30
] are used in the simulation of flood evolution. These models provide the possibility for flood evolution simulation after a dam break. Li [31
] used the calculation method of the single reservoir dam break model, one-dimensional flood evolution calculation theory and reservoir flood regulation calculation principle to establish a numerical calculation model for the continuous failure of cascade earth-rock dams. Chen et al. [7
] established a numerical modeling method for the dam breach–outburst flood disaster chain through the Dam Breach Analysis (DABA) model and one-dimensional Saint-Venant equations, and successfully applied it to the Gyalha landslide dam on Yarlung Tsangpo. Fan et al. [32
] established a comprehensive numerical modeling method for the landslide-dam breach-flood disaster chain through Massflow, the DABA, and HEC-RAS models, and successfully applied it to the Baige landslide of the Jinsha River. These studies provide an important basis for the risk evaluation of outburst flooding due to debris flow dam breaks. However, numerical simulation requires more detailed hydraulic parameters for a barrier dam, and the calculation process is relatively complicated. Some numerical models are time-consuming. Choosing appropriate models that can quickly evaluate dam stability and predict the extent of outburst flooding, and establishing a rapid risk assessment method have important practical significance in emergency response to similar chain disasters. Based on the previous works, it is possible to establish a fast and simple model group for evaluating the debris flow weir dam break-flood disaster chain. The recently developed China Institute of Water Resources and Hydropower Research’s Dam Breach (DB-IWHR) spreadsheet can calculate the discharge hydrograph of dam breaks. The China Institute of Water Resources and Hydropower Research’s Dam Breach Slope (DBS-IWHR) worksheet is used to simulate the horizontal expansion of the failure. The combined dam failure analysis method using these two models is physically representative, numerically friendly, and less sensitive to the input parameters [34
]. DB-IWHR requires only a few input parameters and the straight-forward numerical algorithm allows almost instant calculations, where field engineers can perform a dam breach analysis along with a sensitivity study of a target case within 1 h in DB-IWHR, which includes a tutorial [34
]. Some scholars have used DB-IWHR to achieve good results in the inversion analysis of barrier dam events such as Yigong [35
], Hongshiyan [36
], Xiaogangjian [37
], et cetera, and carried out more accurate predictions when the Baige barrier lake incident occurred in 2018 [38
]. The results of DB-IWHR can be input into HEC-RAS to simulate the evolution of floods. The combination of these three models can quickly predict the hazards of dam-break flood disasters. In order to evaluate the failure of debris flow dam-outburst flood hazard chain under different scenarios in the Bailong River basin and establish a method for the rapid evaluation of the risk associated with this hazard chain, this study takes the Zhouqu “8.8” debris flow hazard chain as an example. We achieve this using the DBS-IWHR, DB-IWHR, and HEC-RAS multi-model. This study could provide a relevant reference for the emergency response, risk assessment, and management of debris flow-dam breach-outburst flood hazard chain events in the Bailong River basin and similar areas.
The results presented an analysis of the stability of the Zhouqu “8.8” debris flow dam and reconstructed the entire process of the dam break and outburst flood disaster. According to the evaluation results of the three geomorphological indices, the Zhouqu “8.8” debris flow dam was inferred to be an unstable dam with a high risk of failure. This result has been verified by the existing safety factor. However, compared with the safety factor method, the geomorphology index requires fewer parameters and is easier to obtain, so the geomorphology index method is more suitable for emergency situations. In order to evaluate the possible disasters caused by dam failure under different situations, we simulated different scenarios and formulated a complete debris flow blocking river-outburst flood disaster chain simulation process for the development of the breach, dam-break flood, and flood evolution. By simulating the flood caused by the dam break and the inundation range of the flood under different upstream water conditions, the range of the possible impact of the dam break was estimated. These findings are useful for decision makers to formulate personnel evacuation routes and determine the method of property transfer. In the design of the spillway for the “8.8” debris flow dam, it is necessary not only to ensure that the peak discharge during the flood discharge process is within the flood control standard, but also to gradually expand the flow cross section by using the hydraulic scour ability to accelerate the discharge [43
]. The simulation results of different spillway sizes can provide a reference for the relevant departments responsible for the spillway design.
A comparison of the simulation results under different scenarios showed that, as the upstream water inflow increased, the erosion rate of the dam body accelerated, the flow of the dam break flood gradually increased, and the arrival time of the flood peak was gradually reduced. When the inflow of the barrier lake increased by 506%, the peak discharge of the dam-break flood increased by 243%, while the arrival time of the flood peak reduced by 31%. When the inflow of the barrier lake increased by 707%, the peak discharge of the dam-break flood increased by 350%, while the arrival time of the flood peak reduced by 50%. The increase in the peak discharge and shortening of the peak arrival time would increase the flood risk. Hence, if the inflow of the barrier lake increases, the flood that would form after the dam break would be larger and more urgent, thereby greatly increasing the flood risk. The simulation of the flood inundation range quantitatively revealed the danger of flooding. When the inflow of the barrier lake increased by 506%, the inundation area of the flood within 5 km downstream of the barrier dam increased by 106%, whereas this was 128% when the inflow of the barrier lake increased by 707%. Combining the inundation area of the flood and the population economic data of Zhouqu can intuitively express the impact of the dam-break flood disaster on the downstream residents, and evaluate the potential risks. When the inflow of the barrier lake increases by 506%, the affected population and indirect economic loss within 5 km downstream of the barrier dam will increase by 65.7%, and when the upstream inflow increases by 707%, the affected population and indirect economic loss within 5 km downstream of the barrier dam will increase by 90.2%. Therefore, the results indicate that the hazard of a dam-break flood would increase rapidly with an increased inflow.
For the simulation of different spillway sizes, under the same upstream inflow, increasing the spillway size appropriately could reduce the peak flood discharge and reduce the submerged area downstream. However, increasing the spillway size would mean that it would not be conducive to narrowing the river bed to form a scouring along the way. Therefore, the design of the spillway size is very important for flood discharge and river channel dredging. According to the flood evolution simulation results of different spillway sizes, suitable discharge channels can be determined.
The key to the dam failure risk analysis system established in this study is to use the combined geomorphological index method to judge dam stability, and to simulate and predict the possible impact of disasters through multi-model coupling. The advantage of the geomorphological index method is that it can quickly determine the stability of a barrier dam. In addition, the multi-model coupling method can quickly reconstruct the entire process of debris flow barrier dam failure and the resultant flood disaster. DBS-IWHR simulates the failure expansion process caused by slope instability. The accurate simulation of failure development is a very difficult to achieve, and the simplified Bishop method adopted by DBS-IWHR is the widely accepted sliding surface analysis method in geotechnical engineering. In practical applications, DBS-IWHR simplifies the circular slip surface to the straight slip surface, which makes the calculations more convenient and faster under the condition of less influence on accuracy [35
]. Moreover, DB-IWHR has a high accuracy for simulating the actual spillway discharge flood. The flood evolution simulation of HEC-RAS offers an intuitive display of the disaster area caused by a dam-break flood. The advantage of the failure risk analysis system used in this study is that the calculation is simple, the simulation accuracy is high, and the results can be obtained relatively quickly in an emergency state, which is very important.
The Zhouqu debris flow was used as an example in this work to assess dam stability based on the geomorphological index method. Different models (DBS-IWHR, DB-IWHR, and HEC-RAS) were coupled to simulate dam failure, and to quantitatively analyze the influences of river discharge and spillway size on the dam failure process and scope of the flood disaster. A comprehensive analysis of the failure risk of the Zhouqu “8.8” debris flow dam was realized. The main conclusions are as follows:
(1) The results of the three geomorphological indices indicated that the Zhouqu “8.8” debris flow dam is unstable.
(2) The simulated peak discharge flow of the actual spillway was 317.15 m3/s, which was basically consistent with the actual discharge of 316 m3/s. The model coupling system revealed that, when the inflow of the barrier lake increased by 506% and 707%, the dam-break peak discharge increased by 243% and 350%, respectively, the peak arrival time decreased by 31% and 50%, respectively, and the submerged area increased by 106% and 128%, respectively, and the affected population and indirect economic losses increased by 65.7% and 90.2%, respectively. Under the same upstream inflow, increasing the spillway size appropriately could reduce the peak flood discharge and reduce the submerged area downstream; however, it is not conducive to dredging and restoring the river channel.
(3) This study comprehensively reconstructed the entire process of the failure of debris flow dam-outburst floods disaster chain from the evolution process of barrier dam failure and outburst flooding, and established a rapid quantitative evaluation and analysis system for the risk of such disasters. Taking Zhouqu as an example, the results indicated the instability of the dam body and the magnitude of the impact of an outburst flood. The established risk analysis system for barrier dam failure and flooding provide reference for emergency response, decision making, and risk management relating to the disaster chain of debris flows blocking rivers in the Bailong River basin and similar areas.