Abstract:
This study introduces a stochastic programming (SP) approach for modelling sustainable development issues under uncertainty. The aim is to point out how to apply SP for high risk with low probability events and how to form the best combination of mitigation alternatives by using SP logic. Two different problem areas are selected in this study, namely earthquake and forest fire hazards in Istanbul and its vicinity. For the earthquake risk mitigation problem, event based scenario approach is developed, and an SP model is proposed. First, we define mitigation alternatives, and the problem of choosing an alternative out of eight alternatives is described to minimize earthquake risk at sustainable level. Then, we develop an SP model including the cost of building damages, loss of lives, infrastructure damage, and the benefit of insurance return for each scenario. It is shown that "relocation" and "rebuild" options decrease the effect of big earthquakes at high marginal levels, and buying insurance is more useful especially in case of medium intensity levels of earthquake risk. In the second part, a different SP approach which is the time based scenario approach, is applied for the forest fire problem. This model searches effective controlling for the forest-level under the risk of uncertain fire losses. Harvest and enhancement are the decisions to be made before the fire whereas regeneration and rehabilitation are made after fire. The important results are that buffer stock area should be set in order to reduce fire loss effect on harvest quantity, and the application of mitigation techniques is effective to reduce fire loss. Furthermore, it is shown that the stochastic programming approach is a useful method for solving real life risk mitigation problems.
