Skip to content

Proposal of a combined generalized rainfall threshold for landslide occurrence on São Miguel volcanic island (Azores, Portugal)

Rui Fagundes Silva^1,2^, Rui Marques1,2 and José Luís Zêzere3,4

  • Affiliations:  1 Research Institute for Volcanology and Risk Assessment, University of the Azores, Rua Mãe de Deus, 9500-321 Ponta Delgada, Portugal. 2 Centre for Information and Seismovolcanic Surveilance of the Azores, University of the Azores, Rua Mãe de Deus, 9500-321 Ponta Delgada, Portugal. 3 Centre of Geographical Studies, IGOT, University of Lisbon, Rua Branca Edmée Marques, 1600--276 Lisboa, Portugal. 4 Associate Laboratory TERRA, Instituto Superior de Agronomia, University of Lisbon, Tapada da Ajuda, 1349--017 Lisboa, Portugal. 

  • Presentation type: Poster

  • Presentation time: Thursday 16:30 - 18:30, Room Poster Hall

  • Poster Board Number: 155

  • Programme No: 2.2.19

  • Theme 2 > Session 2

Abstract

Rainfall-triggered landslides pose a significant threat to infrastructures and human lives. Defining accurate rainfall thresholds is crucial for anticipating landslide events and implementing effective early warning systems. This study, focused on São Miguel volcanic island (Azores, Portugal), aims to propose a generalized rainfall threshold, combining preparatory and trigger thresholds, based on data obtained from four different locations on the island. A total of 62 landslide events, along with rainfall data from 1977 to 2020, were analysed. The Generalized Extreme Value (GEV) distribution, with parameters estimated via Maximum Likelihood Estimation (MLE), was used to identify the cumulative rainfall-duration (Rcum-D) pair with the highest return period for each landslide event, forming the preparatory threshold. The triggering threshold was based on the rainfall intensity recorded on the event day (ID1), but it also depends on the duration of the preparatory rainfall period. The combined threshold anticipates landslide events when both thresholds are exceeded. The generalized combined threshold for São Miguel Island was obtained through logarithmic transformation of the normalised thresholds by the Mean Annual Precipitation (MAP) of each rainfall station. Linear regression was used to determine the parameters a and b of the generalized combined threshold. High MAP areas, such as Furnas and Sete Cidades, have thresholds where the rainfall required to trigger landslides is less demanding. In contrast, regions with a lower MAP, such as Ponta Delgada and Santana, have more demanding, reflecting their lower rainfall regimes. The normalized functions of the generalized combined threshold for the island are Rcum=0.034D-0.34 and ID1=0.034D-0.17.