Ni, Mengke; (2022) Decision making for short-term flood intervention planning under uncertainty. Doctoral thesis (Ph.D), UCL (University College London).

Text Ni_Thesis_for_Library.pdf - Other Access restricted to UCL open access staff until 1 July 2024. Download (3MB)

Abstract

Temporary flood defences are complementary to permanent engineered solutions in flood management. When investing in various deployment strategies, decision makers face two major issues. First, the risk of flooding is increasing due to unpredictable short-term weather conditions, and fixed strategies are vulnerable if the future becomes different. Second, resources are insufficient or underused, posing huge flood hazard or resulting in waste. This thesis proposes a novel technique to decision making approaches that address the aforementioned issues. The real options analysis (ROA) was applied by the use of multistage stochastic mathematical models. A scenario tree based short-term flood planning formulation is used to seek flexibility and adaptability given future weather condition uncertainties. The tree captures various deployment strategies that include numerous decision stages. The proposed method can optimally delay, accelerate or replace deployment options to flood-prone areas when future uncertainties are resolved. Furthermore, the suggested technique with alternate approaches were compared, including 1) deterministic 2) robust and 3) two-stage stochastic optimisation methods, to evaluate the influence on the proposed plan's performance. It is insufficient to consider simply direct economic losses when identifying the optimal strategies. In practice, few models are built considering intangible (indirect) flood losses such as fatalities, cultural loss and environmental influence; our thesis is innovative in that it develops models of intangible flood losses to identify effective defence deployment strategies. The proposed models are applied to nine flood-prone locations in Carlisle, Northwest England and the results show that there is a value achieved by building ROA approach in the design of temporary defence deployment under uncertainty. Additionally, it demonstrates that the proposed adaptive approach is more cost-effective than other optimisation approaches in this issue.

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item