Uncertainties in local and regional mass movement prediction using rainfall

The accurate prediction of debris flows occurrence that will allow the reduction or prevention of economic losses and human casualties is presently the most difficult aspect of debris flows studies but also the aspect that receives most attention. Most prediction methods are based on rainfall as the basic parameter, with the moment of occurrence as only result, and without a prediction of debris flow travel time and size. This paper takes Jiangjia Gully in Dongchuan of Yunnan Province as an example, and considers, on the basis of the fulfillment of the essential condition: the abundant availability of loose materials, the conditions for the formation of debris flows. Based on the mechanism of the initiation of debris flows in channels and the volume of rainfall in the basin, this paper also gives a systematic analysis on the travel time and size of the debris flow and suggests that the hydrological condition for forming debris flow is the unit discharge of the flood ≥0.35 m3/s.m. It uses the 10-min rainfall intensity to calculate both the run-off of the rainfall and the unit discharge caused by the run-off, thus predicting the occurrence of debris flows. The velocity and the travel time of a debris flow can also be determined using the unit discharge of the run-off. The total volume of debris flows can be calculated using the 10-min intensity of rainfall and the total volume of the run-off, together with the volume concentration of the sediment in a debris flow.

To predict the potential geographic distribution of Oncomelania hupensis in Yunnan Province using random forest (RF) and maximum entropy (MaxEnt) models, so as to provide insights into O. hupensis surveillance and control in Yunnan Province. The O. hupensis snail survey data in Yunnan Province from 2015 to 2016 were collected and converted into O. hupensis snail distribution site data. Data of 22 environmental variables in Yunnan Province were collected, including twelve climate variables (annual potential evapotranspiration, annual mean ground surface temperature, annual precipitation, annual mean air pressure, annual mean relative humidity, annual sunshine duration, annual mean air temperature, annual mean wind speed, ≥ 0 ℃ annual accumulated temperature, ≥ 10 ℃ annual accumulated temperature, aridity and index of moisture), eight geographical variables (normalized difference vegetation index, landform type, land use type, altitude, soil type, soil textureclay content, soil texture-sand content and soil texture-silt content) and two population and economic variables (gross domestic product and population). Variables were screened with Pearson correlation test and variance inflation factor (VIF) test. The RF and MaxEnt models and the ensemble model were created using the biomod2 package of the software R 4.2.1, and the potential distribution of O. hupensis snails after 2016 was predicted in Yunnan Province. The predictive effects of models were evaluated through cross-validation and independent tests, and the area under the receiver operating characteristic curve (AUC), true skill statistics (TSS) and Kappa statistics were used for model evaluation. In addition, the importance of environmental variables was analyzed, the contribution of environmental variables output by the models with AUC values of > 0.950 and TSS values of > 0.850 were selected for normalization processing, and the importance percentage of environmental variables was obtained to analyze the importance of environmental variables. Data of 148 O. hupensis snail distribution sites and 15 environmental variables were included in training sets of RF and MaxEnt models, and both RF and MaxEnt models had high predictive performance, with both mean AUC values of > 0.900 and all mean TSS values and Kappa values of > 0.800, and significant differences in the AUC (t = 19.862, P < 0.05), TSS (t = 10.140, P < 0.05) and Kappa values (t = 10.237, P < 0.05) between two models. The AUC, TSS and Kappa values of the ensemble model were 0.996, 0.954 and 0.920, respectively. Independent data verification showed that the AUC, TSS and Kappa values of the RF model and the ensemble model were all 1, which still showed high performance in unknown data modeling, and the MaxEnt model showed poor performance, with TSS and Kappa values of 0 for 24%(24/100) of the modeling results. The modeling results of 79 RF models, 38 MaxEnt models and their ensemble models with AUC values of > 0.950 and TSS values of > 0.850 were included in the evaluation of importance of environmental variables. The importance of annual sunshine duration (SSD) was 32.989%, 37.847% and 46.315% in the RF model, the MaxEnt model and their ensemble model, while the importance of annual mean relative humidity (RHU) was 30.947%, 15.921% and 28.121%, respectively. Important environment variables were concentrated in modeling results of the RF model, dispersed in modeling results of the MaxEnt model, and most concentrated in modeling results of the ensemble model. The potential distribution of O. hupensis snails after 2016 was predicted to be relatively concentrated in Yunnan Province by the RF model and relatively large by the MaxEnt model, and the distribution of O. hupensis snails predicted by the ensemble model was mostly the joint distribution of O. hupensis snails predicted by RF and MaxEnt models. Both RF and MaxEnt models are effective to predict the potential distribution of O. hupensis snails in Yunnan Province, which facilitates targeted O. hupensis snail control.

The Ms 8.0 Wenchuan earthquake greatly altered the threshold for rainfall-triggered debris flows in the affected areas. It is of both scientific and practical significance to determine the rainfall thresholds. This study examines one of the regions most prone to debris flows to analyze the characteristics of rainfall that caused debris flows, and to explore local rainfall thresholds. We applied the relation between rainfall intensity and duration, peak intensity and event amount, and other single factor approaches. Comparison of effectiveness and accuracy indicates that the event rainfall is the most sensitive factor for forecasting. Analysis of the annual rainfall thresholds showed that the rainfall conditions required for debris flows have increased on the continent during the past 6 years. Besides the rainfall fluctuations over the past few years, material changes were the primary reason for threshold variability. Recovery of vegetation plays an important role in reducing potential loose material that supplies volume for debris flows. Natural solidification, decrease of the potential erosion depth, and surface coarsening make it more difficult to initiate a debris flow, and ultimately increased rainfall conditions required. The change in rainfall thresholds can be predicted and verified for the entire earthquake-affected region.

Probabilistic rainfall thresholds for debris flows occurred after the Wenchuan earthquake using a Bayesian technique

Constructing rainfall thresholds for debris flow initiation based on critical discharge and S-hydrograph

Climate change has increased the amount and intensity of rainfall, increasing debris flow risk. Debris flow damage has become a social issue in Korea, and the need for research related to rainfall thresholds for debris flow has increased. In this study, an automatic calculation algorithm was developed to derive criteria suitable for the Korean region for the rainfall thresholds of debris flow. The study area selected was Gangwon-do, where a significant amount of debris flow occurred recently, and information on rainfall and the occurrence of debris flow were collected. As a result of applying an algorithm based on the R programming language, an equation for the relationship between the rainfall intensity and duration in Gangwon-do was automatically derived using debris flow occurrence point information and meteorological rainfall information.

Culicoides biting midges exhibit a global spatial distribution and are the main vectors of several viruses of veterinary importance, including bluetongue (BT) and African horse sickness (AHS). Many environmental and anthropological factors contribute to their ability to live in a variety of habitats, which have the potential to change over the years as the climate changes. Therefore, as new habitats emerge, the risk for new introductions of these diseases of interest to occur increases. The aim of this study was to model distributions for two primary vectors for BT and AHS (Culicoides imicola and Culicoides bolitinos) using random forest (RF) machine learning and explore the relative importance of environmental and anthropological factors in a region of South Africa with frequent AHS and BT outbreaks. Culicoides capture data were collected between 1996 and 2022 across 171 different capture locations in the Western Cape. Predictor variables included climate-related variables (temperature, precipitation, humidity), environment-related variables (normalised difference vegetation index-NDVI, soil moisture) and farm-related variables (livestock densities). Random forest (RF) models were developed to explore the spatial distributions of C. imicola, C. bolitinos and a merged species map, where both competent vectors were combined. The maps were then compared to interpolation maps using the same capture data as well as historical locations of BT and AHS outbreaks. Overall, the RF models performed well with 75.02%, 61.6% and 74.01% variance explained for C. imicola, C. bolitinos and merged species models respectively. Cattle density was the most important predictor for C. imicola and water vapour pressure the most important for C. bolitinos. Compared to interpolation maps, the RF models had higher predictive power throughout most of the year when species were modelled individually; however, when merged, the interpolation maps performed better in all seasons except winter. Finally, midge densities did not show any conclusive correlation with BT or AHS outbreaks. This study yielded novel insight into the spatial abundance and drivers of abundance of competent vectors of BT and AHS. It also provided valuable data to inform mathematical models exploring disease outbreaks so that Culicoides-transmitted diseases in South Africa can be further analysed.

Impact of Chi-Chi earthquake on the occurrence of landslides and debris flows: example from the Chenyulan River watershed, Nantou, Taiwan

BackgroundLumbar drainage is widely used in the clinic; however, forecasting lumbar drainage-related meningitis (LDRM) is limited. We aimed to establish prediction models using supervised machine learning (ML) algorithms.MethodsWe utilized a cohort of 273 eligible lumbar drainage cases. Data were preprocessed and split into training and testing sets. Optimal hyper-parameters were archived by 10-fold cross-validation and grid search. The support vector machine (SVM), random forest (RF), and artificial neural network (ANN) were adopted for model training. The area under the operating characteristic curve (AUROC) and precision-recall curve (AUPRC), true positive ratio (TPR), true negative ratio (TNR), specificity, sensitivity, accuracy, and kappa coefficient were used for model evaluation. All trained models were internally validated. The importance of features was also analyzed.ResultsIn the training set, all the models had AUROC exceeding 0.8. SVM and the RF models had an AUPRC of more than 0.6, but the ANN model had an unexpectedly low AUPRC (0.380). The RF and ANN models revealed similar TPR, whereas the ANN model had a higher TNR and demonstrated better specificity, sensitivity, accuracy, and kappa efficiency. In the testing set, most performance indicators of established models decreased. However, the RF and AVM models maintained adequate AUROC (0.828 vs. 0.719) and AUPRC (0.413 vs. 0.520), and the RF model also had better TPR, specificity, sensitivity, accuracy, and kappa efficiency. Site leakage showed the most considerable mean decrease in accuracy.ConclusionsThe RF and SVM models could predict LDRM, in which the RF model owned the best performance, and site leakage was the most meaningful predictor.

The major landslide events that occurred in Chittagong City in 2007 and 2008 have caused the death of about 140 people and incurred a considerable amount of property loss. At least ten landslide events occurred during the period of 2000–2009, with mainly rainfall as the triggering factor. The critical rainfall condition of the landslide events was analyzed with Gumbel’s extreme value distribution using the historical rainfall–landslide data. The antecedent rainfalls were used for Gumbel’s distribution for the critical rainfall intensity–durations of each historical landslide event. Landslides were correlated with the monthly average as well as the percentile rank of monthly median rainfall. The regression relationship between 30 days of absolute antecedent rainfall intensity (I) and corresponding duration (D) of rainfall responsible for the landslides in the study area was found as I = −15.8ln(D) + 60.49. The calibrated antecedent rainfall analysis showed a relation between I and D as I = −40.5ln(D) + 162.5. The major recent landslide events were related to the extreme rainfall intensities (>40 mm/day) having short period of time (2–7 days). All the major landslide events occurred at much higher rainfall amount compared to the monthly average. However, these rainfall thresholds were statistically based and were not defined hydrological response of rainfall to the soil, as well as other geological conditions associated with slopes.

In mountainous areas in the Republic of Korea, landslides occur primarily due to heavy rainfall in the summer monsoon season. While these landslides tend to occur seasonally in summer, the rainfalls that have triggered landslides are complex and varied. Temporal prediction of landslides based on the variables of rainfall events mainly uses two variables: intensity-duration or cumulative event rainfall-duration. However, it does not consider the antecedent rainfall, another critical factor that could characterize various complex rainfalls regarding landslide occurrences. Here, we first attempted to determine critical rainfall variables and their threshold values for landslide occurring using the decision tree method necessary to consider multiple rainfall variables simultaneously. We then classified landslide-triggering rainfall based on the identified critical variables using the K-means clustering method. We chose as the study area Chuncheon in the middle of the Korean Peninsula, an eroded granite basin surrounded by schist and gneiss mountains, since it has not been affected hardly by earthquakes and thus is suitable for studying rainfall-induced landslides. According to the decision tree analysis, cumulative rainfall and 5-day antecedent rainfalls were determined as critical variables, implying that considering antecedent and cumulative rainfall simultaneously is significant for landslide prediction. The K-means clustering analysis classified landslide-triggering rainfalls into four types: 1) low cumulative rainfall (198.6 &amp;#177; 90.9 mm) with high antecedent rainfall for seven days prior to the landslide, 2) medium cumulative rainfall (308.3 &amp;#177; 81.1 mm) with a gradual increase in antecedent rainfall for four weeks, 3) high cumulative rainfall (534.5 &amp;#177; 85.7 mm) with low antecedent rainfall for four weeks, and 4) high cumulative rainfall (538.4 &amp;#177; 59.8 mm) with a gradual decrease in antecedent rainfall for four weeks. In particular, the high cumulative rainfall after gradually decreased antecedent rainfall caused the most frequent landslides. Our results suggest that the threshold of cumulative rainfall varies with the antecedent rainfall pattern and that antecedent rainfall data of at least four weeks have meaningful information in forecasting and preparedness for landslide occurrence.

Debris flow is one of the most common types of geological disasters in China. Owing to the influence of topography, geomorphology, geological conditions, human activity, and rainfall debris flow disasters frequently occur in the mountainous areas of Beijing. The research on debris flow in the Beijing area focuses on rainfall and risk evaluation, material sources, and early warning and prevention of debris flow. However, there are few studies on the development characteristics of single-gully debris flow and the prediction of hazardous areas in the Beijing area. Therefore, we chose the debris flow of Bangou Village in Yanqing District of Beijing as the research object. We analyzed the recharge conditions in the ditch domain and predicted the extent of the hazardous area around the gully, providing suggestions for control measures. The dynamic reserves of the loose deposits in the debris flow gully, currently in the development stage, were estimated as 15.48 × 104 m3, representing four supply sources: artificial deposits, alluvium and diluvium, residual slope deposits, and collapse. The peak flow is 24.49 m3/s for a 10-year rainfall event, 27.64 m3/s for a 20-year rainfall event, 31.79 m3/s for a 50-year rainfall event, and 34.93 m3/s for a 100-year rainfall event. The total amounts of solids washed out by a debris flow from the preceding events are 0.70 × 104 m3, 0.79 × 104 m3, 0.91 × 104 m3, and 1.00 × 104 m3, respectively. The size of the debris flow is small, with a maximum hazardous area of 0.2810 km2. We conclude that a small debris flow outbreak in the Bangou Village gully is possible. We expect that the results of this study will provide basic information and help improve debris flow research in Beijing.

Critical rainfall thresholds can be the key to ensuring effective debris-flow forecasting. They are significant for the study of the trigger mechanisms of debris flows, for the forecast of the characteristics of future events, and for the development of engineering guidance for mitigation. Using a hydrological approach, we first calculated the flood peak discharge at different frequencies and then the corresponding rainfall thresholds for the initiation of different scales of debris flows in Zhouqu County, China. This was followed by the establishment of a functional relation between the intensity and duration of rainfall events that trigger debris flows of different warning levels in two initial soil conditions (dry and moist). This, in turn, yielded four warning levels and two preliminary warning levels. For the two early soil conditions (dry and moist) in the Sanyanyu Gully of Zhouqu County, the Level I (red) warning values for rainfall triggering debris flow are 56 and 51 mm h −1 , respectively; the Level II (orange) warning values are 41 and 38 mm h −1 , respectively; the Level III (yellow) warning values are 32 and 30 mm h −1 , respectively; and the Level IV (blue) warning values are 24 and 22 mm h −1 , respectively. The Level V preliminary warning values are 17 and 16 mm h −1 , respectively; and the Level VI preliminary warning values are 10 and 9.5 mm h −1 , respectively. The rainfall intensity and duration were found to exhibit a power function relation, I = αD β , where the values of α and β vary with the warning levels. Rainfall events capable of triggering debris flows in the new thresholds and intensity–duration relations presented here can be used for forecasting purposes and in operational geohazard warning systems. These research results also provide a scientific basis for regional hazard mitigation and reduction in Zhouqu County.

Micro accelerometers are crucial tools for monitoring and providing early warning of geological disasters. This study focuses on the potential debris flow in Wangjiayuan, Fang County, Hubei Province, aiming to explore the relationship between clay content and debris flow motion characteristics. Physical model experiments were designed with varying clay content to simulate debris flow motion, and micro accelerometers were employed to monitor and record vibration acceleration changes during debris flow initiation and motion. Frequency spectra were generated using Fourier transform to analyze the motion characteristics in the frequency domain. The results showed that the predominant period of debris flow initiation ranged from 15 Hz to 29 Hz, while the period during motion in the middle section of the flow area ranged from 36 Hz to 54 Hz. Higher clay particle content was associated with a decrease in both the frequency range and mean of the predominant period, as well as a reduction in the Fourier amplitude extremum. These findings provide a foundation for monitoring and predicting debris flows, and the vibration signal characteristics of debris flow initiation and motion can be used to infer the clay content, offering insights into early warning systems for such geological hazards.

On predicting debris flows in arid mountain belts