Decision making on flood mitigation incorporating uncertainty, socio-economic factors and a changing future
Date of Issue2018
Interdisciplinary Graduate School (IGS)
Nanyang Environment and Water Research Institute
The decision-making process in flood mitigation typically involves many factors reflecting flood severity, flood vulnerability and the cost of the mitigation measures. In this PhD study, a Multi-Criteria Decision Analysis (MCDA) decision framework for optimal decision making in flood protection design, specifically levees are developed. This framework accounts for climate change, increasing urbanization, and evolving socio-economic features of the flood plain, and as well as for uncertainties in rainfall predictions. The changing climate and the growing urbanization alter the flood frequencies with large uncertainties that strongly influence future projections. These uncertainties make the flood mitigation decision more complex. Thus, uncertainties involved in both current and future conditions are quantified via computation of the expected values of change indices developed. The MCDA uses as its criteria the annual expected loss, graduality, a newly developed Socio-Economic Vulnerability Index (SEVI) and levee construction cost. It is further demonstrated for a central basin of Jakarta, Indonesia. The annual expected loss at current conditions is calculated with recorded rainfall data as fitted using the Log Pearson Type III distribution. The graduality represents the severity with change in discharge as reflected by the deviation from linearity between percentile discharge and percentile loss values. The SEVI reflects the social and economic impact on the flood affected population via scaling with the flood inundation area and depth while capturing uncertainty in the rainfall forecasting. Temporal change factors (CF) as calculated from a statistically downscaled global gridded rainfall projection, the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) are applied for future rainfall (specifically 2031-2070). The uncertainty across the 20 Global Climate Models (GCM’s) is quantified via an expected CF scaled across the GCMs via an inverse distance method. Landsat data over year 1989-2009 is classified into various land use, including urban, and is used to develop urbanization trends and projected to year 2050. The impact of climate change and urbanization are incorporated in the MCDA flood decision framework through the criteria being recalculated under the future conditions. The changing climate, growing urbanization and the socio-economic features, are shown to drive the best choice of levee protection plan towards higher protection levels under the future time period considered. Inclusion of socio-economic factors is also shown to change the best plan. The developed methodology and the results are expected to guide decision making in flood mitigation and can be extended to include decision-making on other flood mitigation measures as well as additional sources of uncertainties.