Assessing the performance of regional landslide early warning models: the EDuMaP method

Abstract

Abstract.A 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).

Adapting the EDuMaP method to test the performance of the Norwegian early warning system for weather-induced landslides

“…The paper proposes the evaluation of the performance of the landslide early warning system operational in Norway by means of the EDuMaP method (Calvello and Piciullo, 2016). This method has been principally employed to analyse the performance of regional early warning model considering fixed warning zones for issuing alerts.…”

“…The method comprises three successive steps: identification and analysis of landslide and warning events (E) from available databases, definition and computation of a duration matrix (DuMa), and evaluation of the early warning model performance (P) by means of performance criteria and indicators. The first step requires the availability of landslides and warnings databases for the preliminary identification of LEs Calvello and Piciullo, 2016). Four classes of warning events (key: no means no warning; M is moderate warning; H is high warning; VH is very high warning) and four classes of landslide events (key: no means no landslides; S is a small event, with few landslides; I is an intermediate event, with several landslides; L is a large event with many landslides).…”

“…A WE is defined as a set of warning levels issued within a given warning zone, grouped by their temporal characteristics. There are 10 parameters that need to be defined to carry out the event analysis: (1) warning levels, W lev ; (2) landslide density criterion, L den(k) ; (3) lead time, t LEAD ; (4) landslide typology, L typ ; (5) minimum interval between landslide events, t LE ; (6) over time, t OVER ; (7) area of analysis, A; (8) spatial discretization adopted for warnings, A (k) ; (9) time frame of analysis, T ; (10) temporal discretization of analysis, t. For more details see Calvello and Piciullo (2016). The second step of the method is the definition and computation of a "duration matrix", whose elements, d ij , report the time associated with the occurrence of LEs in relation to the occurrence of WEs, in their respective classes.…”

“…Cheung et al, 2006;Godt et al, 2006;Martelloni et al, 2012;Staley et al, 2013;Segoni et al, 2014;Lagomarsino et al, 2015;Gariano et al, 2015;Stähli et al, 2015). The method employed herein, which is called Event, Duration Matrix, Performance (EDuMaP; Calvello and Piciullo, 2016), allows us to consider aspects peculiar to territorial LEWSs that are not considered by the joint frequency distribution approach. In particular, the EDuMaP method takes into account the occurrence of concurrent multiple landslides in the warning zone, the duration of the warnings in relation to the landslides, the issued warning level in relation to the landslide spatial density in the warning zone, and the relative importance attributed, by system managers, to different types of errors.…”

“…In particular, the EDuMaP method takes into account the occurrence of concurrent multiple landslides in the warning zone, the duration of the warnings in relation to the landslides, the issued warning level in relation to the landslide spatial density in the warning zone, and the relative importance attributed, by system managers, to different types of errors. Up to now, this method has been applied exclusively to evaluate the performance of regional warning models designed for issuing alerts in fixed warning zones (Calvello and Piciullo, 2016;Piciullo et al, 2016a, b;Calvello et al, 2015). In the present study the EDuMaP method has been adapted to evaluate the performance of the alerts issued for variable size warning zones.…”