Karagiorgos et al. (2016) Multi-vulnerability analysis for flash flood risk management

Identification

• Journal: Natural Hazards
• Year: 2016
• Date: 2016-03-21
• Authors: Konstantinos Karagiorgos, Thomas Thaler, Johannes Hübl, Fotios Maris, Sven Fuchs
• DOI: 10.1007/s11069-016-2296-y

Research Groups

• Institute of Mountain Risk Engineering, University of Natural Resources and Life Sciences, Vienna, Austria
• Flood Hazard Research Center, Middlesex University, London, UK
• Department of Forestry and Management of the Environment and Natural Resources, Democritus University of Thrace, Orestiada, Greece

Short Summary

This study conducts a multi-dimensional vulnerability analysis for flash flood hazards in East Attica, Greece, integrating physical and social vulnerability assessments to bridge the disciplinary gap. It establishes empirical vulnerability functions for buildings and assesses social vulnerability through surveys, finding generally low vulnerability attributed to specific building regulations and socio-economic factors.

Objective

• To undertake a multi-dimensional vulnerability analysis for flash flood hazards in Greece, integrating physical and social vulnerability assessments to bridge the divide between social and natural science perspectives.
• To quantitatively evaluate individual vulnerability components, including elements at risk, their physical exposure, social characteristics, and underlying institutional settings, by coupling the internal (coping capacity) and external (exposure, physical susceptibility) sides of vulnerability.

Study Configuration

• Spatial Scale: East Attica regional unit, Greece, covering an area of 1513 km² between sea level and 1109 m above sea level.
• Temporal Scale: Physical vulnerability data from eight flash flood events between 1996 and 2006; social vulnerability data collected via door-to-door surveys in May-June 2012.

Methodology and Data

• Models used:
  • Nonlinear regression approach to derive cumulative distribution functions for physical vulnerability.
  • Weibull distribution was selected as the best fit for vulnerability functions after testing Weibull, Frechet, Exponential, and Logistic distributions.
  • Root Mean Square Error (RMSE) was used to evaluate model fit.
• Data sources:
  • Physical Vulnerability: Loss assessment reports from the Prefecture of East Attica (for building characteristics and absolute water depths inside houses), data from the Earthquake Recovery Service of Greece (for reconstruction values), and legal amendments from the Greek Ministry for Health and Social Solidarity (for household contents damage compensation).
  • Social Vulnerability: Empirical data collected through a door-to-door survey of 114 flash flood victims across four study sites in East Attica, using 11 variables identified through a literature review.

Main Results

• Physical Vulnerability:
  • 114 buildings were analyzed for flash flood losses. Buildings without cellars showed a mean vulnerability of 0.052 (mean loss of 4810 €), while buildings with cellars showed a mean vulnerability of 0.025 (mean loss of 4208 €).
  • Vulnerability functions, best fitted by Weibull distributions, indicated a relatively sharp increase in vulnerability for buildings without cellars up to 0.5 m flood height, then a flattening curve. Buildings with cellars showed a more gradual flattening.
  • Buildings with pilotis (Category A) were the least susceptible (mean vulnerability 0.0159 without cellar, 0.0071 with cellar). Buildings with enhanced ground plates (Category B) without cellar were less susceptible than Category C (non-enhanced ground plates) without cellar. Conversely, Category C buildings with cellar were less susceptible than Category B with cellar.
  • Both absolute and relative intensity models showed low RMSE values (0.039 and 0.036 for buildings without cellar, 0.055 and 0.022 for buildings with cellar, respectively), suggesting comparable applicability.
• Social Vulnerability:
  • Respondents generally showed low social vulnerability due to high employment rates, education levels, risk perception, private property ownership (over 82.8%), and age distribution.
  • However, a low degree of local integration (low solidarity, trust, and participation in local associations) was observed, despite long residency.
  • The economic crisis significantly impacted household income, preventing investments in individual protection measures (over 85.8% had no savings/investments for property).
  • A high-risk perception was noted due to frequent past flood events, but this did not translate into high individual preparedness (over 50% showed very low/low preparedness).

Contributions

• Established empirical vulnerability functions for buildings exposed to flash floods in a Mediterranean context, suitable for operational risk analyses and spatially explicit valuation within a GIS environment.
• Bridged the disciplinary gap between natural and social sciences by integrating physical and social vulnerability assessments for flash flood hazards.
• Demonstrated that a combination of different vulnerability perspectives leads to a better understanding of actors' perceptions and capacities, aiding in the development of locally embedded coping strategies and alternative flood risk management.
• Provided insights into the specific factors shaping vulnerability in Greece, such as building design principles influenced by anti-seismic regulations and the impact of economic crises on preparedness.

Funding

Not explicitly mentioned in the paper.

Citation

@article{Karagiorgos2016Multivulnerability,
  author = {Karagiorgos, Konstantinos and Thaler, Thomas and Hübl, Johannes and Maris, Fotios and Fuchs, Sven},
  title = {Multi-vulnerability analysis for flash flood risk management},
  journal = {Natural Hazards},
  year = {2016},
  doi = {10.1007/s11069-016-2296-y},
  url = {https://doi.org/10.1007/s11069-016-2296-y}
}