Modelling future flood risks in the Bangkok Metropolitan Region
2017-07-12T08:19:05Z (GMT) by
Due to rapidly changing climate and socio-economic conditions, many coastal areas are becoming increasingly vulnerable to internal and external risks of flooding. Low-lying coastal mega-cities in Southeast Asia are widely recognized as hotspots of flood risk. The Bangkok Metropolitan Region is one of the largest coastal megacities in Southeast Asia that is challenged by the potential impacts of climate change and human activities expected over coming decades. The overarching aim of this research is to evaluate present and future flood risks due to the combined impacts of climate (sea-level, rainfall regime and storm surge) and human (land subsidence and drainage capacity) factors in Bangkok Metropolitan region, Thailand. To design plausible future scenarios, flow and precipitation records were examined using the Log Pearson Type III frequency analysis approach. Land subsidence (LS) and sea level rise (SLR) scenarios were derived from historical records and published studies. Future flood risks (fluvial, surface water, and coastal) were modelled under various combinations of key drivers (SLR, storm surge, LS and increased river flow). The October 2011 flood in Thailand was used as a baseline event for coastal and fluvial flood modelling. Scenarios were designed with projections of LS and SLR to 2050, 2080, and 2100. A two-dimensional flood inundation model (FloodMap) was used to derive coastal inundation depth, velocity and extent associated with each scenario. Coupled modelling of one-dimensional river flow (HEC-RAS) and two-dimensional flood inundation (FloodMap) was undertaken. Surface water flood modelling simulated the 2015 event in model calibration. A two-hour rainfall event that occurred in 2011 was used as the baseline to derive future scenarios with increased precipitation of various return periods and topographies accounting for land subsidence. For each type of flood modelling, sensitivity analysis was first conducted to investigate the effects of mesh resolution and roughness parameters on model predictions. Results indicate that the model is sensitive to both resolution and roughness, but to various degrees, depending on the metrics used in the evaluation. Spatial metrics such as the Root Mean Standard Error, F and point depth are able to distinguish between model predictions and reveal the spatial and temporal derivations between simulations. The impacts of flood risk on critical infrastructure nodes (e.g. power supply, transportation network, rescue centres, hospitals, schools and key government buildings) were then evaluated under various scenarios. Overall, results suggest progressively increased risks of coastal, surface water, and fluvial flooding to critical infrastructures over time from 2050, 2080 to 2100. Flood modelling of coastal and fluvial inundation processes suggests that the combined impacts of individual risk drivers is, in most cases, far greater than any of the individual factors alone. This study demonstrates that flood risks in coastal mega-cities like Bangkok must be evaluated in a holistic manner, taking into account multiple key risk drivers and considering the potential joint-occurrence of various types of flooding. Moreover, where numerical modelling was undertaken and infrastructure data are available, local hotspots of flood risks under various scenarios can be identified, allowing potential adaptation measures to be evaluated within the modelling framework developed. This research is the first to consider multiple flood risk drivers and interacting flood risks within a single modelling framework in the Bangkok Metropolitan Region. It will have long lasting legacy for flood risk management in the region and beyond, enabling more effective adaptation in a changing climate through: (i) raised awareness of multiple risk drivers and interacting flood risks for both the public and policy makers; (ii) further and more complete assembly of various data sets when they become available based on the template demonstrated in this study; and (iii) identification of hotspots of critical infrastructure and communities at risk using refined and alternative modelling approaches within the modelling framework developed in this study.