Risk-based determination of polder height against storm surge in the south- west region of Bangladesh

Abstract

Coastal areas of the world face a range of risks related to natural hazards. Cyclone and associated storm surge are the most destructive hazard that frequently occur in Bangladesh and cause enormous suffering to the people of the coastal area. Construction of coastal polder system is one of the common ways to protect the coastal area from storm surge flooding. Although Bangladesh was introduced with coastal polder system in 1960s, most of the existing polders have been being raised and strengthened through implementing traditional return period based method which is disputable as this method is not based on proper risk analysis. This study has been conducted through applying risk-based decision making to assess the optimal polder level in the context of storm surge hazard in Polder No. 33 of Dacope Upazila in Khulna district. Assessment of storm surge hazard and vulnerability were required to complete the study as risk-based methods generally consider the consequences of failure due to the occurrence of hazard. Frequency analysis of storm surge hazard was carried out using surge level data of 19 independent cyclones using Gumbel’s EV-1 distribution. The study was conducted following the principle of risk-based approach to determine optimal polder level for Polder No. 33, including overtopping as the polder failure mode due to storm surge hazard. Both the infrastructural and risk cost were investigated for this analysis. This study involved the application of some softwares for technical analysis like ERDAS IMAGINE 10.1 and ArcGIS 10.2. Primary and secondary information from different sources were analyzed to assess the losses in agricultural sector in the concerned area. Primary data were collected using Participatory Rural Appraisal (PRA) tools. Secondary data were collected from different government organizations and published scientific literatures.

The study found that around 50% of the total areas were inundated during the Cyclone Aila whereas another deadly cyclone named the Cyclone Sidr inundated approximately 17% of the total area. From the study it was also found that inundation pattern had a direct linkage with agricultural damage as net agricultural loss due to the immediate impacts of the Cyclone Aila was about 5 times higher compared to the agricultural loss during the Cyclone Sidr. The analysis was of use to generate vulnerability function of agriculture where depth-damage relationship was developed. This analysis shows that, agricultural products in the medium high land and high land areas were susceptible to storm surge flooding compared to that of the medium low land areas. This is because land areas adjacent to the river and inside the polder were mostly dominated by medium high land to high land. The optimal polder level was estimated as 4.29 mPWD with a design return period of 12.50 years and an approximate annual cost of BDT 150 million. However, this polder was already taken under rehabilitation program by Coastal Embankment Improvement Project (CEIP). According to the study by CEIP, the proposed average polder level was 4.5 mPWD which is 0.21m higher compared to the present study. Such over estimation ultimately increased the average annual cost due to polder heightening. Hence this is recommended to apply risk-based approach instead of applying traditional return period based approach in case of designing coastal polder system. Risk- based approach considers probability and consequences of hazard during analysis which ultimately results optimal project size.