Spatial and temporal risk assessment of crop yield loss from tephra fall: A case study on Mount Etna

Details

Supervisors

Delmelle, Pierre ; Dib, Hind

Faculty

Faculté des bioingénieurs

Degree label

Master [120] : bioingénieur en sciences agronomiques, à finalité spécialisée

Abstract

enVolcanic soils cover only about 1% of the world’s ice-free land surface but feed at least 10% of the global population thanks to their high fertility. These agriculturally valuable landscapes are subject to multiple volcanic hazards, among which tephra accumulation on crops is the leading cause of agriculture-related losses from volcanic eruptions. Severe and long-lasting tephra fall impacts on agriculture underscore the need for spatial and temporal prioritization of mitigation actions supported by reliable risk models that integrate hazard, exposure and vulnerability. While tephra hazard modelling has advanced substantially, few studies have linked tephra loading to crop damage through empirical relationships or have combined the components of risk in an integrative manner. No study has yet done so at the scale of an entire agricultural area in the vicinity of a volcano. In this work, we provide a methodology to quantify risk of crop yield loss from tephra fall through the case study of Mount Etna agriculture using two models that differ in hazard data treatment and are applied to two eruption magnitudes. The “IM” model uses probabilistic isomass maps to estimate yield loss for various tephra deposit levels, each with an associated exceedance probability. The “1000 simulations” model applies biweekly mean tephra deposits, producing yield loss estimates without probability values but incorporating wind seasonality. This integrative framework outperforms hazard-, exposure- or vulnerability based agricultural impact assessments by highlighting risk patterns across multiple spatial and temporal scales. Results show that incorporating wind seasonality in the “1000 simulations” model reveals greater variability in loss magnitude and spatial distribution than models without this component. These findings represent a contribution and open the way to improvements in volcanic risk management, including resource allocation, land-use planning and targeted preparedness measures. While this research establishes the foundations for a new framework of spatial and temporal risk models for crop yield loss from tephra fall, some limitations remain. Future developments should include refining the vulnerability assessment, adopting a probabilistic biweekly hazard approach for short-term eruptions and incorporating a fully probabilistic framework for eruption scenarios.