How Mixing Surveys, Maps and Conversations Could Transform Disaster Risk Assessment on Tenerife

On the volcanic island of Tenerife, where wildfires, coastal storms, landslides and the looming threat of eruptions all compete for attention, a team of researchers has argued that the way scientists assess disaster risk is due for a rethink. In an essay published in PLOS Climate, Guillermo García Álvarez and Marleen C. de Ruiter of the Institute for Environmental Studies at Vrije Universiteit Amsterdam make the case that quantitative models alone cannot capture the messy, interlocking realities of multi-hazard risk at the local scale. Their proposed remedy is a mixed-methods approach that weaves together desktop analysis, geographic information system mapping, field observations, household questionnaires and stakeholder interviews into a single iterative framework, one in which each method continuously refines the others.

The urgency of the problem is grounded in a shifting hazard landscape. Climate change is intensifying climate-driven natural hazards, and these events increasingly strike simultaneously, in quick succession, or trigger one another in cascading chains. Recent studies cited by the authors show that multi-hazard events are becoming more frequent across Europe and globally. The United Nations Office for Disaster Risk Reduction defines multi-hazards as the selection of multiple major hazards a country faces, considering the specific contexts where hazardous events may occur simultaneously, cascadingly or cumulatively over time, taking into account their potential interrelated effects. In response, risk research has been shifting from single-hazard to multi-hazard thinking, yet most large-scale assessments remain dominated by quantitative methods such as machine learning, and data gaps often force studies into narrow approaches that exclude low- and middle-income countries.

The difficulty, the authors argue, is that multi-hazard risk assessment is inherently complex. It demands data, a deep understanding of how risks interact across space and time, and awareness of the local context in which disasters actually unfold. Decision makers face competing adaptation priorities and potential conflicts between measures designed for different hazard types. Moreover, the effectiveness of any disaster risk reduction strategy depends on how communities perceive and respond to risk, a dimension that standardized quantitative methods aimed at larger scales routinely miss. Risk perception, the essay notes, does not always align with modelled risk. A study in Bangladesh combining media narratives, interviews and field observations found that risk perception is often ignored in local adaptation planning, while research on coastal property markets in the United States shows that despite growing hazards, coastal properties continue to rise in value.

The temporal dynamics of perception are equally striking. Immediately after a flood event, property prices in inundated areas drop by an average of 24.9 percent compared to non-flooded properties, yet that discount loses statistical significance within five years. In multi-hazard contexts there is a further complication: focusing on one hazard can decrease people’s attention to the risk posed by another. To structure their investigation of these dynamics, the researchers anchored their survey work in Protection Motivation Theory, a psychological framework with two key components. Threat appraisal addresses how likely people believe an event is and how severely they expect to be hit, while coping appraisal probes whether people believe anything can be done, whether it is affordable, and whether government adaptation measures would actually work.

Mixed-methods research, defined as projects that bring together qualitative approaches such as interviews with quantitative techniques such as risk modelling and survey analysis, is not new. Greene and colleagues laid the terminological foundations in 1989, reviewing studies that already combined methodologies, and interest has grown notably over the past decade, particularly in the social sciences. The main designs include convergent parallel, explanatory sequential and exploratory sequential approaches. What García Álvarez and de Ruiter propose is a simultaneous, iterative process in which quantitative and qualitative strands inform each other continuously. A qualitative interview can improve the interpretation of structured survey data or reshape the survey itself for later stages, as stakeholders identify problems the researchers had not anticipated. Quantitative results, in turn, can guide which questions matter most in interviews and which adaptation solutions to model.

The framework was tested during a ten-day fieldwork trip to Tenerife in March 2025, an island the authors describe as a complex multi-hazard environment exposed to volcanic eruptions, coastal hazards, wildfires and landslides. Recent events, including the devastating 2023 wildfires and the 2021 eruption on the neighboring island of La Palma, have highlighted gaps in preparedness. The team applied a sequential design in which quantitative methods influenced qualitative instruments and vice versa: GIS-based observations and survey results shaped interview questions, while exploratory interviews determined which key questions to address quantitatively and which adaptation solutions to model. A total of 174 respondents completed a 25-question survey, administered in person and online in Spanish and English versions, coded in KoboToolbox and analyzed in Python to test Protection Motivation Theory hypotheses. Five semi-structured interviews targeted three key economic sectors highly exposed to climate risk: agriculture, tourism and critical infrastructure, with thematic analysis covering risk perception, adaptation levels and trust in authorities.

The complementarity of the methods surfaced repeatedly. Desktop analysis and GIS-based exposure mapping provided an initial understanding of each region and identified locations to verify in the field, where observations revealed missing infrastructure, unregistered buildings and local bottlenecks. Several regions of the island contained buildings absent from Google Maps that would have been neglected in modelling otherwise. Questionnaire results showed that respondents were unaware of any emergency plan for most climate hazards despite recent wildfires and coastal storms, yet residents knew volcanic eruption procedures because drills had been conducted by some municipalities after the La Palma eruption. That contrast, the authors suggest, could directly shape future emergency communication plans. In the northern town of Tegueste, nestled in a valley between forested mountains and exposed to wildfires and flooding from steep ravines called barrancos, field observations confirmed residential areas and infrastructure inside high-risk zones, along with accumulated vegetation and debris that could block water flows.

Interviews then revealed opportunities invisible to any dataset. Conversations with locals and field observations uncovered large areas of abandoned agricultural terraces near Taganana, offering low-conflict sites for reforestation, since converting productive farmland to nature-based solutions would require compensating farmers. For Tegueste specifically, the survey confirmed that locals ranked wildfires as the top perceived hazard following the nearby 2023 blaze, and econometric analysis showed that residents with prior hazard experience were more likely to have taken adaptive measures. Yet overall worry levels were low, a contrast between perceived and modelled risk that the authors say reveals communication and awareness gaps pure modelling would overlook. The final recommendations combined these strands: clearing vegetation at barranco bottlenecks, installing five small retention basins to slow runoff, and introducing grazing herbivores to manage vegetation regrowth.

The approach carries particular weight for nature-based solutions, defined by the European Commission as solutions inspired and supported by nature that are cost-effective while providing environmental, social and economic benefits. Their feasibility depends on community acceptance and perceived co-benefits, dimensions that are difficult to quantify but critical for uptake, and the lack of standardized indicators has limited both the evidence base and private investment. Mixed-methods can monetize non-market co-benefits through tools such as choice experiments and contingent valuation, feeding into cost-benefit analysis, while also exposing inequities, since nature-based solutions can raise property prices and displace lower-income residents. In one Dutch study cited by the authors, conversations with waterboard members revealed land-use conflicts that led researchers to add agricultural land conversion as a survey attribute, uncovering a significant decline in respondent utility as more farmland was converted, an insight that would otherwise have been missed. The framework also aligns with the IUCN Global Standard for Nature-based Solutions, addressing criteria on societal challenges, local context, economic viability, inclusive governance and trade-offs.

The authors are candid about the costs and pitfalls. Fieldwork is resource-intensive and time-bounded; access to key locations was sometimes blocked by private property, forcing contingency plans. Time constraints limited questionnaire numbers and raised sampling bias risks, since environmentally engaged residents were more willing to participate, and any mid-fieldwork survey revision cost a full day of data. Language barriers complicated interviews with stakeholders uncomfortable speaking English, and technical terms such as multi-hazards and nature-based solutions required clarification. Integrating diverse data types, from ecological measurements to soil stability and tourism impacts, into a coherent framework risks inconsistencies, and local results resist generalization. Their research agenda responds with concrete prescriptions: interdisciplinary teams familiar with the region, standardized mixed-methods protocols for local multi-hazard risk, pre-testing surveys before fieldwork begins, preliminary analysis during the trip to catch data problems while they can still be fixed, and genuine incorporation of local knowledge. Done well, they conclude, the iterative integration of methods bridges the gap between scientific modelling and community realities, producing adaptation strategies that are simultaneously evidence-based, socially equitable and, crucially, accepted by the people they are meant to protect.

Subject of Research: Mixed-methods approaches to local multi-hazard climate risk assessment and nature-based solutions in Tenerife

Article Title: Breaking the methodological silos: Integrating Mixed-Methods in Multi-Hazard Risk Assessment Lessons learnt from fieldwork in Tenerife

Article References: García Álvarez, G., & de Ruiter, M. C. (2026). Breaking the methodological silos: Integrating Mixed-Methods in Multi-Hazard Risk Assessment Lessons learnt from fieldwork in Tenerife. PLOS Climate, 5(10), e0001070. https://doi.org/10.1371/journal.pclm.0001070

Image Credits: AI Generated

DOI: 10.1371/journal.pclm.0001070

Keywords: multi-hazard risk, mixed methods, Tenerife, nature-based solutions, climate adaptation, risk perception, disaster risk reduction, fieldwork, stakeholder interviews, GIS mapping, PLOS Climate, Canary Islands