Post‐quake decision analysis using dynamic programming

We introduce a general decision analysis procedure based on stochastic dynamic programming in the post‐quake aftershock environment. The damage sustained by the building due to the mainsheet, the time‐varying aftershock rates and the potential for further damage progression in the post‐quake environment are all factors taken into consideration in the proposed methodology. This procedure enables the optimal decision after the mainshock to be selected based on the minimization of expected financial losses, subject to a constraint on a minimal level of individual life‐safety, using a consistent probabilistic framework to explicitly quantify the uncertainties in the variables. Copyright © 2008 John Wiley & Sons, Ltd.