Flash Flood, Scales and Societal Impacts Isabelle RUIN - isar@ucar.edu National Center for Atmospheric Research Boulder, CO 4 Corners Emergency Management Conference Branson, MO December, 2008
Flash flood are especially dangerous for drivers More than half of flash flood fatalities are vehiclerelated Other 20% to 30% are pedestrians 10 to 20 % of casualties happen inside buildings Indirect casualties as heart attack, electrocution Flash flooding is the #1 weather killer in the Ozarks
- Why drivers are especially vulnerable? - Why do they decide to travel in flash flood conditions? Spatial / temporal DAILY Physical Cultural CRISIS Professional Social Professional Social Activities Social Activities Vital Adaptation / Resistance Vital
A vulnerable area
September 2002 Flash Flood High water mark Dions
Pont du Gard September, 9 2002 5:30pm Pont du Gard maximum peak flow Photos Jean-Pierre MEGER
Tourists are also exposed
Pont Saint Nicolas de Campagnac Septembre, 9-6pm Maximum of the water: 2,3m (7ft) above the bridge 16,90 m / 55 ft Septembre, 10-2pm
What s left after the flood Pont Saint Nicolas de Campagnac
Power of Flash Floods Nîmes Rues Catinat - Richelieu October - 1988
Nîmes Boulevard G. Pompidou Before During October 1988
Spatio-temporal scales of Flash Floods Watersheds time responses (mn) 1 day 1 hour 1 Flash Flood domain 2 2- Meso-scale convective system 100 Km Several hours Scale of atmospheric objects (Orlanski, 1975) Scale of hydrological responses 1- Convective cell Source : Creutin, 2001 One hour 10 Km Ardèche (640km 2 ) time response = 5h
A different risk depending on catchment size Ruin et al., 2008. Journal of Hydrology, 199-213. 11 young individuals died in 9 watersheds smaller than 20 km 2 11 old individuals died in 5 watersheds bigger than 1000 km 2
Learning from loss of life circumstances Comparing the timing of the event with societal impacts Accumulated rainfall in mm during phase I of the storm (8:00UTC - 22:00 UTC) Ruin et al., 2008. Journal of Hydrology, 199-213. Simulated hydrological response at the outlet of Basin #11 where fatal accident occured
Learning from loss of life circumstances Warning timeliness Extreme speed of watershed responses Extremely short leadtime for warnings Ruin et al., 2008. Flood risk management: Research and Practice, 1005-1012
Learning from those who did survive Warnings timeliness and response 2- Workers constrained by professional activities 13 % (sample 2004) Those avoiding evacuation 3 % (Sample. 2004) 1- The cautious 58 % (sample 2004) Ruin and Lutoff, 2004. Houille Blanche, 114-119.
Large road network exposure ALES NIMES Administrative area boundary Main streams Tributaries Toll motorways Highways Secondary roads Road s sections regularly flooded
Motorists danger perception on daily itineraries 29% of the road sections used by our sample are prone to flooding The 2/3 are not considered as dangerous Main streams East rural zone South urban zone West rural zone North urban zone Road network Road sections prone to flooding ALÈS Comparison between often flooded road sections and risk NIMES perceptions Source : DDE30, Cognitive mapping survey, 2006. N = 200 Road sections used and perceived as non dangerous Road sections used and perceived as dangerous by: 76 to 100% of users 51 to 75% of users 26 to 50% of users 1 to 25% of users Ruin et al., 2007. Environmental Hazards, 235-244.
Flash flood risk perceptions Rapidity of watershed time response is mostly underestimated, especially for small catchments More than 60% underestimate the height of moving water that can sweep a car away. But they are more realistic about the water depth for a person to be knocked off their feet. Spatial risk perception is correlated with everyday experience of road networks Spatial risk perception is high when dealing with short itineraries within a perimeter close to the place of residence Traveling during a flash flood event is known to be dangerous, but thresholds of dangerousness are not accurately perceived.
Age influence perception and declared behaviors Youngest people tend to undervalue dangerous water depth They also tend to be more mobile in emergency situation
Tourists are more likely to do the right thing even though warnings hardly reach them Residents tend to act according to their own e n v i r o n m e n t knowledge Tourists behavior is rather based on standard safety recommendations
Influence of spatial and temporal settings 1000 km2 20 km2 Hydographic network Relief Catchment outlet Catchment People located at the confluence of watersheds of different sizes Succession or simultaneity of flood peaks due to differences in catchment sizes Vulnerability variations within the time of the day, week, season... 16
Social constraints Parental reaction to warnings when children are in schools In reaction to warnings, 50% of the parents would pick up their children from school. Immediatly pick up their children Nothing, you know they are safe in school You ask a relative to pick them up Others Responses to Météo-France watches (orange) and warnings (red) for heavy precipitations Cancel Travels Workers would hardly cancel their travels Search information Unchanged activities and/or travel patterns Response to orange alert Workers Non-workers Mean deviation Response to red alert Workers Non-workers
Research perspectives Need for Behavioral verification surveys to assess adaptative capacities of drivers in different weather conditions: 1. Observe behaviors in both normal daily life and extreme weather conditions 2. Organize extreme events post investigations
1- Observe drivers behavior Objective 1: Identify environmental factors influencing drivers behaviors at low water crossings Do different types of barriers prevent motorists from driving through flooded roads? How are behaviors influenced by environmental cues (visibility, rainfall, water depth, water flow ), weather forecast, watches and warnings? Does the type of car make a difference in terms of behavior? Quantitative assessment through car counting, video observations, cell phone activity
1- Observe drivers behavior Objective 2: Identify personal factors influencing drivers behavior at low water crossings How much local knowledge and cognitive mapping of neighborhoods residents influence driving behaviors? What personal traits may be correlated with risky or cautious behaviors? What travel purposes lead to most risky behaviors? Qualitative assessment thru GPS tracking, travel diaries, in-depth interviews, community collaborative network, YouTube videos
FFRAP GIS Risk Analysis www.weather.gov/springfield Flood Reports Per Basin Spatial Analysis Tools also allow us to correlate number of flood reports, fatalities, injuries and water rescues across warning basins
Improve post-flood investigations DELUGE: Eve Gruntfest new initiative Disasters Evolving Lessons Using Global Experience Focus on post-event field studies for floods to maximize interactions between social scientists, hydrologists and meteorologists New guidelines on postevent investigations for use by integrated teams of physical scientists, social scientists, and practitioners.
Thanks for your attention, questions? The city of Nîmes, 3 October 1988 Last Spring... Fall 1988 (500 years return period) Cadereaux River in Nîmes, 50 km 2
Observe drivers behavior Objective 1: Identify environmental factors influencing drivers behaviors at low water crossings in Texas Do different types of barriers prevent motorists from driving through flooded roads? How are behaviors influenced by environmental cues (visibility, rainfall, water depth, water flow ), weather forecast, watches and warnings? Does the type of car make a difference in terms of behavior? Quantitative assessment through car counting, video observations, cell phone activity
Observe drivers behavior Objective 2: Identify personal factors influencing drivers behavior at low water crossings in Texas How much local knowledge and cognitive mapping of neighborhoods residents influence driving behaviors? What personal traits may be correlated with risky or cautious behaviors? What travel purposes lead to most risky behaviors? Qualitative assessment thru GPS tracking, travel diaries, in-depth interviews, community collaborative network, YouTube videos