Kenneth H. Tiedemann
Floods, Drowning, Natural disasters, Regression modelling , Model identification
Among the various natural disasters, floods are one of the most important causes of mortality. Research on specific major flood events has shown that a variety of factors such as flood height, flood rise rate, and flood arrival warning time influence the flood mortality rate. In 2005, Hurricane Katrina created mass destruction in several American states bordering the Gulf of Mexico, with the devastation of the hurricane centred on New Orleans. This paper models the determinants of flood fatalities in the aftermath of Hurricane Katrina, using White’s estimator which corrects for heteroscedastic error terms. Key findings of this study are as follows. First, using bivariate regressions, a 1% increase in flood depth increases the mortality rate by 0.016%, a 1% increase in flood rise rate increases the mortality rate by 0.0041%, and a 1% increase in flood arrival time decreases the mortality rate by 0.0011%. Second, using multivariate regression, a 1% increase in flood depth increases the mortality rate by 0.0086%, a 1% increase in flood rise rate increases the mortality rate by 0.00079%, and a 1% increase in flood arrival time decreases the flood mortality rate by 0.0064%.
Important Links:
Go Back
