20 research outputs found
Correction to: Monitoring strategies for local landslide early warning systems (Landslides, (2019), 16, 2, (213-231), 10.1007/s10346-018-1068-z)
No full textThe original version of this article was revised: Unconverted data in Figure 3; Tables 1 and 2 caption has error in both PDF and XML; Table 6 needs to be organized and structured so it would be more readable
Standards for the performance assessment of territorial landslide early warning systems
No full textLandslide early warning systems (LEWS) can be categorized into two groups: territorial and local systems. Territorial landslide early warning systems (Te-LEWS) deal with the occurrence of several landslides in wide areas: at municipal/regional/national scale. The aim of such systems is to forecast the increased probability of landslide occurrence in a given warning zone. The performance evaluation of such systems is often overlooked, and a standardized procedure is still missing. This paper describes a new Excel user-friendly tool for the application of the EDuMaP method, originally proposed by (Calvello and Piciullo 2016). A description of indicators used for the performance evaluation of different Te-LEWS is provided, and the most useful ones have been selected and implemented into the tool. The EDuMaP tool has been used for the performance evaluation of the “SMART” warning model operating in Piemonte region, Italy. The analysis highlights the warning zones with the highest performance and the ones that need threshold refinement. A comparison of the performance of the SMART model with other models operating in different Te-LEWS has also been carried out, highlighting critical issues and positive aspects. Lastly, the SMART performance has been evaluated with both the EDuMaP and a standard 2 × 2 contingency table for comparison purposes. The result highlights that the latter approach can lead to an imprecise and not detailed assessment of the warning model, because it cannot differentiate among the levels of warning and the variable number of landslides that may occur in a time interval
Calibration of rainfall thresholds for landslide early warning purposes: Applying the EDuMaP method to the system deployed in Campania region (Italy)
No full textAssessing the performance of regional landslide early warning models: the EDuMaP method
No full textA schematic of the components of regional early warning systems for
rainfall-induced landslides is herein proposed, based on a clear distinction
between warning models and warning systems. According to this framework an
early warning system comprises a warning model as well as a monitoring and
warning strategy, a communication strategy and an emergency plan. The paper
proposes the evaluation of regional landslide warning models by means of an
original approach, called the "event, duration matrix,
performance" (EDuMaP) method, comprising three successive steps:
identification and analysis of the events, i.e., landslide events and
warning events derived from available landslides and warnings databases;
definition and computation of a duration matrix, whose elements
report the time associated with the occurrence of landslide events in
relation to the occurrence of warning events, in their respective classes;
evaluation of the early warning model performance by means of
performance criteria and indicators applied to the duration matrix. During
the first step the analyst identifies and classifies the landslide and
warning events, according to their spatial and temporal characteristics, by
means of a number of model parameters. In the second step, the analyst
computes a time-based duration matrix with a number of rows and columns
equal to the number of classes defined for the warning and landslide events,
respectively. In the third step, the analyst computes a series of model
performance indicators derived from a set of performance criteria, which
need to be defined by considering, once again, the features of the warning
model. The applicability, potentialities and limitations of the EDuMaP
method are tested and discussed using real landslides and warning data from
the municipal early warning system operating in Rio de Janeiro (Brazil)
Modeling of propagation and entrainment phenomena for landslides of the flow type: the May 1998 case study
No full textRainfall-induced landslides of the flow type are dangerous phenomena due to their high velocities and large run-out distances. Indeed, proper modeling of their propagation stage is a fundamental issue for risk analysis and management. To this aim, several factors must be taken into account to properly estimate the run-out distance and the landslide magnitude that are strongly related to an appropriate choice of the rheological properties of the moving mass. Moreover, several distinct processes must be adequately tackled such as: i) relative movement of the interstitial fluid relative to the solid fraction, ii) vertical consolidation process, iii) entrainment of material along the landslide path. All the above mentioned processes can be consistently simu-lated through the use of a depth-integrated coupled SPH model which revealed to be appropriate in simulating landslides of the flow type. In this paper, taking into account the entrainment phenomena, the above model is applied to the May 1998 Sarno-Quindici case history (Southern Italy) for which an advanced data-set is avail-able. The numerical analyses provide a satisfactory simulation of the observed propagation path, deposition heights and velocities that are strongly influenced by the entrainment rate and the extent of the erodible area
Regional-scale spatiotemporal landslide probability assessment through machine learning and potential applications for operational warning systems: a case study in Kvam (Norway)
No full textThe use of machine learning models for landslide susceptibility mapping is widespread but limited to spatial prediction. The potential of employing these techniques in spatiotemporal landslide forecasting remains largely unexplored. To address this gap, this study introduces an innovative dynamic (i.e., space–time-dependent) application of the random forest algorithm for evaluating landslide hazard (i.e., spatiotemporal probability of landslide occurrence). An area in Norway has been chosen as the case study because of the availability of a comprehensive, spatially, and temporally explicit rainfall-induced landslide inventory. The applied methodology is based on the inclusion of dynamic variables, such as cumulative rainfall, snowmelt, and their seasonal variability, as model inputs, together with traditional static parameters such as lithology and morphologic attributes. In this study, the variables’ importance was assessed and used to interpret the model decisions and to verify that they align with the physical mechanism responsible for landslide triggering. The algorithm, once trained and tested against landslide and non-landslide data sampled over space and time, produced a model predictor that was subsequently applied to the entire study area at different times: before, during, and after specific landslide events. For each selected day, a specific and space–time-dependent landslide hazard map was generated, then validated against field data. This study overcomes the traditional static applications of machine learning and demonstrates the applicability of a novel model aimed at spatiotemporal landslide probability assessment, with perspectives of applications to early warning systems
Performance analysis of landslide early warning systems at regional scale
No full text2014 - 2015Landslide early warning systems are non-structural risk mitigation strategies aiming at dealing with intolerably high probabilities of landslide occurrence by reducing risk through the reduction of the exposed elements. The majority of landslide early warning systems deal with rainfall-induced landslides. The systems can be classified, as a function of the scale of analysis, into: “local” and “regional” systems. Several differences exists among these two different types of warning systems, such as: the actors involved in the process, the monitoring tools, the variables selected to define triggering thresholds, the way the warnings are issued and spread to the public. This work exclusively deals with regional landslide early warning systems (ReLEWSs). These systems are used to assess the probability of occurrence of landslides over appropriately-defined homogeneous alert zones of relevant extension, typically through the prediction and monitoring of meteorological variables, in order to give generalized warnings to administrators and the population. At first, a detailed review of the structure and the functioning of these systems is presented. The information has been gathered mainly from the literature, with the exception of the regional system operating in Campania region, Italy, the municipal system of Rio de Janeiro, Brazil, and the national Norwegian landslide early warning system. The functioning and the structure of the latter two systems have been analyzed in greater depth thanks to research periods spent, respectively, at the GEO-Rio foundation in Rio de Janeiro and at The Norwegian Water Resources and Energy Directorate (NVE) in Oslo. In literature, several authors provided a general description of the structure of a landslide early warning system. Starting from the analysis of these contributions, an original scheme and the main components of such systems for rainfall-induced landslides forecast is proposed. The scheme is based on a clear distinction among the following components: correlation laws, decisional algorithm and warning management. Subsequently, the functioning of the reviewed ReLEWSs has been described according to these components, with a special attention on how the performance of the various warning models was assessed. It is straightforward that a periodical assessment of the technical performance of a landslide early warning system, in terms of evaluation of the warning issued in relation to the landslides occurred, is a required task in order to continuously keep the system reliable. Nevertheless, no standard requirements exist for assessing the performance of regional warning models (ReWaMs) and, typically, this is evaluated by computing the joint frequency distribution of landslides and warnings, both considered as dichotomous variables. Herein, an original methodology to assess the performance of ReWaMs, called the “Event, Duration Matrix, Performance” (EDuMaP) method, is proposed. The performance is evaluated taking into account: the possible occurrence of multiple landslides in the warning zone; the duration of the warnings in relation to the time of occurrence of the landslides; the warning level issued in relation to the landslide spatial density in the warning zone; the relative importance system managers attribute to different types of errors. The applicability of EDuMaP method is tested considering three different ReLEWSs: the municipal early warning system operating in Rio de Janeiro (Brazil); the Norwegian landslide early warning system; the landslide early warning system for hydro-geological risk management of the Campania region, Italy. The main differences among these systems are discussed in great detail, mainly dealing with the functioning and the databases available for the three case studies. The LEWS operational in Rio de Janeiro is employed to issue a certain level of warning in four warning zones in which the municipality is divided. The warnings can be issued at any time during the day if the monitored rainfall exceeds pre-identified thresholds. The Norwegian landslide early warning system is employed to issue daily warnings adopting variable warning zones. In the LEWS of the Campania region each municipality has a reference rain gauge for which three different rainfall threshold are specified for the activation of 3 warning levels. The EDuMaP method was successfully employed to assess the performance for all these case studies, thus underlying the wide applicability of the method, which can be easily adopted to evaluate the performance of any regional landslide early warning systems for which landslides and warnings data are available. For the three case studies, sensitivity analyses are also conducted by varying some of the input parameters of the EDuMaP method. The results of these analyses indicate that the input parameters most affecting the performance of the warning models are: i) the landslide density criterion used to differentiate among the classes of landslide events; ii) the database on landslides considered in the simulations; iii) the time set
xvii
as the minimum time interval between landslide events; iv) the area of analysis; v) the time frame of the analysis. In conclusion, the analyses prove the applicability of the EDuMaP method in evaluating the performance of real case studies related to ReLWaMs characterized by different decisional algorithms, components and input parameters. The method can also be used as an effective tool to calibrate a warning model by back-analysing landslide and warning data in test area with the aim of defining the set of warning criteria which maximises the model performance. [edited by author]XIV n.s
Calibration of rainfall thresholds for landslide early warning purposes: Applying the EDuMaP method to the system deployed in Campania region (Italy)
No full textPreface: Landslide early warning systems: monitoring systems, rainfall thresholds, warning models, performance evaluation and risk perception
No full text
