Flood damage models Scope and limitations Forest Resources

Flood damage models – Scope and limitations Forest Resources and Climate Unit Institute for Environment and Sustainability (IES) European Commission - Joint Research Centre José I. Barredo • EU Loss Data Workshop: Progress towards guidelines on sharing loss data among EU countries and state of the art of recording loss data at local level • 30 -31 March 2015 - Ispra

Assessing flood disaster losses • To account for the direct losses produced by floods • Direct economic damage is the tangible economic loss associated with a flood’s impact as determined after the event • These losses occur after floods as a result of the physical contact of the flood waters with damageable property • Indirect (i. e. secondary and tertiary) and intangible damages (i. e. loss of human life, Ill-health of floods victims), as well as longer-term macroeconomic effects are not considered

Flood damage models • Useful for assessing economic flood impact • These models produce estimates of flood damage • But the estimation of flood damage is a complex process involving a large number of hydrologic and socio-economic factors • Models: structure, inputs (data) and outputs. These depend on the purpose of the model

Flood damage models Baseline structure and data: Land cover Flood delineation Water depth CORINE data Copernicus – EMS Satellite imagery

Flood damage models Economic losses Flood damage functions € 400. 00 € 350. 00 € 300. 00 Land cover 111 Continuos urban fabric € 250. 00 Flood losses (Euro/m 2)€ 200. 00 € 150. 00 € 100. 00 € 50. 00 € 0. 00 0 1 2 3 4 Flood water depth (m) 5 6 Grid cell (100*100 m): Land cover class: 111 Water depth: 2. 0 m Economic losses: 2, 016, 500 Euro

Flood damage functions • Flood damage functions measure the susceptibility of assets to certain inundation characteristics, in this case specifically inundation depth • Flood water depth is the determining factor in almost all models in use today However other flood water parameters influence damage: • Flood duration: few minutes, days … • Water flow velocity: flash floods in mountain areas, narrow valleys • Sediment (and solid components) concentration in water • Pollutant content in water

Land cover limitations • CORINE land cover: Pan-European map • 44 land cover classes • Available at 100 and 250 m grid size • Minimum mapping unit: 25 ha* !!! • Linear (e. g. infrastructure) or small features are not well represented or not represented at all !!! * 4 cells (250 m); 25 cells (100 m)

Land cover limitations CORINE (100*100 m) 111 – Continuous urban fabric 18% transport 50% residential 5% commercial 17% other 5% unused 5% construction See: EEA 2006 The thematic accuracy of Corine land cover 2000 - Assessment using LUCAS (land use/cover area frame statistical survey), EEA Technical report No 7/2006.

Flood water delineation limitations Flood delineation and water depth • Copernicus Emergency Management Service: • Important support for flooded area mapping • Remote sensing limitations: lack of imagery (clouds), partial coverage, acquisition time of imagery (too early, too late) , … Water depth computation • This is a critical parameter that is usually not available: • Computation from EU-wide flood hazard map (100 -yr return period flood) • Assumptions based on ancillary information (minimum / maximum depth)

Comparing flood damage models

Comparing flood damage models We compared seven flood damage models: FLEMO (Germany) Damage Scanner (The Netherlands) 2002 2005 Rhine Atlas (Rhine basin) Flemish Model (Belgium) Multi-Coloured Manual (United Kingdom) HAZUS- MH (United States) JRC Model (European Union) Source: Jongman et al 2012. Comparative flood damage model assessment: towards a European Approach. NHESS, 12, 3733– 3752.

Results The models produced different results: • Modelled damage • Relative distribution: residential, industry & commerce and infrastructure Source: Jongman et al 2012. Comparative flood damage model assessment: towards a European Approach. NHESS, 12, 3733– 3752.

Results Uncertainty should be considered… Source: Jongman et al 2012. Comparative flood damage model assessment: towards a European Approach. NHESS, 12, 3733– 3752.

Summary of results • Model output are very sensitive to uncertainty in both vulnerability (damage functions) and exposure (asset values) • The first has large effect than the latter • Estimation of residential and commercial building damage is the best developed part of the models • Care needs to be taken when using aggregated land use data for flood risk/damage assessment • It is essential to adjust assets values to the regional economic situation and property characteristics Source: Jongman et al 2012. Comparative flood damage model assessment: towards a European Approach. NHESS, 12, 3733– 3752.

Conclusions • Results of models are estimations Check with reality e. g. claims from insurance companies customers • Land use data: • Varying density of assets at local level • Assets share inside each land use category • Flood damage functions: source, verification with reality, testing, sensitivity analysis • Infrastructure losses are usually underestimated • Regional/national variations in assets value

Thank you