Post-flood assessment of emergency evacuation policies in the Red River basin, Southern Manitoba

A discrete choice experiment (DCE) was used as a part of a post-flood surve y among the residents of four small urban communities on the Red River - Em erson, Morris, Ste. Agathe, and St. Adolphe - to assess their preferences f or emergency evacuation policies. The respondents were asked to choose repe atedly among three hypothetical evacuation policies described in terms of ( a) the type of evacuation (mandatory, voluntary no evacuation), (b) notific ation time for evacuation (1, 2 4 days), (c) post-evacuation flood relief ( 75%, 80% current, 90%) and (d) an overall evaluation of the risk of hazardo us flooding, i.e. chances of over-dike flooding or breaching through dikes (99%, 75%, and 50%). The survey results indicated that, despite significant differences between Ste. Agathe and other communities in the magnitude of impact of the 1997 flood, overall, a voluntary evacuation seemed to be pref erred at the 50 percent risk of hazardous flooding; whereas the preference for a mandatory evacuation was related to a 99 percent risk. A notification time of 2 days for evacuation was preferred over that of I day but the pre ference for 2 days was not statistically different from that of 4 days. In both types of evacuation a 10 percent increase in the amount of flood relie f was preferred over the current amount of 80 percent and the marginal util ity of a 5 percent decrease was nearly equivalent to the suggested 10 perce nt increase. Because of such ability of the DCE to assess tradeoffs of alte rnative choices /policies, the paper suggests further applications of the m odel for eliciting floodplain residents' preferences for various flood mana gement issues.