Forecasting of flash flood impacts integrating the space-time distribution of social vulnerability

Galateia Terti, LTHE, Janv. 2014 - Janv. 2017
Bilan Thèse

Doctorant : Galateia Terti, LTHE, Janv. 2014 - Janv. 2017
Direction : Isabelle Ruin – Sandrine Anquetin (LTHE) ; Jonathan J. Gourley (NOAA-NSSL, Norman, Oklahoma)
Financement : 50% LabEx, 50% NOAA (USA)
Ecole Doctorale : Terre Univers Environnement

Sujet de thèse

If you have ever attempted to hand contour a map of damage from flash flooding, you will quickly discover the influence of surface controls, which create discontinuous and untidy contours. Low-water crossings, infrastructure built in close proximity to streams, normally dry riverbeds in populated zones, and roads in low-lying areas are all examples of anthropogenic factors that increase societal susceptibility to flash flooding.
In addition, impacts in terms of loss of life are very much related to the space-time distribution of the everyday life social activity. In fact, in such fast evolving events impacts depend not just on such compositional variables as the magnitude of the flood event and the vulnerability of those affected, but also on such contextual factors as its location and timing. Those contextual factors can alter the scale and social distribution of impacts and vulnerability to them. In the case of flooding fatalities, for instance, the elderly are often said to be the most vulnerable, but when fatalities are mapped against basin size and response time, it has been shown that in fact it is young motorist who are most likely to be killed in flash flooding of small catchments, whereas the elderly most frequently perish in their homes from large scale fluvial flooding (Ruin et al., 2008).
Unfortunately, tools commonly used to monitor and predict flash flooding, including sophisticated distributed hydrologic models, rarely incorporate any information about anthropogenic vulnerability factors and never envisage the dynamic nature of the exposition of the population to such fast evolving events.
This research will rely on impact data analysis coming from a detailed impact classification and analysis of flash flood reports from the National Weather Service (NWS) Storm Data and Severe Hazards Analysis and Verification Experiment (SHAVE) conducted at the National Severe Storms Laboratory (NSSL) in Norman, Oklahoma.

Plan et déroulement du travail de thèse

 1) The US Flash Flood impact Database (Gourley et al., 2013) will be analyzed to test two main hypotheses :

  • a. The flooding of small catchments (< 20 km2) is mostly associated with punctual inundation of the road network, vehicle-related and “open-air” fatalities of young to middle-aged adults. On the other hand, large hydrological scales (> 1000 km2) are responsible for building damages, evacuations and inside drowning of elderly or physically vulnerable populations.
  • b. The spatial distribution and loss-of-life circumstances are related to the flood timing and the diurnal rhythms of the social activities. For instance, vehicle-related fatalities or rescues are expected to be more frequent when a flood event happens during rush hours when everyone is driving than when it happens during working hours. On the contrary, during night hours, less people are usually exposed to road flooding but the poor visibility makes people more vulnerable to hazardous circumstances.

 2) A small number of representative flash flood events (in terms of timing, hydro-meteorological and socio-demographical contexts) will be extracted from the database to be further analyzed. Those events will be used as case studies to :

  • a. Implement two types of distributed hydrological models : i) the French CVN model described by Braud et al. (2010) and developed at the LTHE (Grenoble University) ; the Ensemble Framework For Flash Flood Forecasting (EF5) model developed by the University of Oklahoma and the NSSL at the National Weather Center ;
  • b. Identify space-time flash flood vulnerability patterns using GIS layers of road networks, population density, stream locations, elevation, degree of imperviousness (urbanization), land use, etc. (Calianno et al, 2012b) ;
  • c. Combine the space-time vulnerability indices with the distributed hydrologic models to test the efficacy of such integrated models in reproducing the impacts of the selected past catastrophic events.

This interdisciplinary research will take place in the context of a collaboration between Jonathan J. Gourley, NOAA research hydrologist of the National Severe Storms Laboratory (NSSL) at the National Weather Center (NWC) in Norman (Oklahoma, USA), and CNRS researchers’ Sandrine Anquetin, hydro-meteorologist and Isabelle Ruin, human geographer both from the Laboratoire d’étude des Transferts en Hydrologie et Environnement (LTHE) at Grenoble University (France). Until now this collaboration has taken the form of Master students exchanges and teaching experiences concluded by the publication of three peer-reviewed papers (Ruin et al., 2012 ; Calianno et al., 2012 ; Gourley et al., 2013).

The “delivrables” of this PhD work are twofold :
 Building and validating flash flood impacts forecasting tools based on existing US databases and envisage the transferability of the methodology in a the French Mediterranean context ;
 Intensifying and strengthening interdisciplinary research in an international environment.

Mis à jour le 11 avril 2018