Flood damage and risk assessment model in the haor basin of Bangladesh

Abstract

The Sylhet haor basin is bounded by the hills of Meghalaya in India to the north, the hills of Tripura and Mizoram to the south and the high lands of Manipur to the east. The total area of Sunamganj district, some part of Habiganj, Moulavibazar, Kishoreganj and Netrakona districts are covered by major Haor Systems in the north east region. Flash flood is the most commonly occurring water related disaster in the haor area. The main cause of flash flood is the heavy rainfall that normally occurs in the hilly region and ultimately discharges with high flow velocity through downstream river network in Bangladesh. Flash flood usually occurs during the pre-monsoon season (March-April-May). During the flash flood it is very common that people lost their primary agricultural productions which are the only source of their livelihood. The main objective of this study is to assess flood damage and risk of agricultural boro crop due to pre-monsoon flash flood. Annual maximum and pre-monsoon maximum water level data were used for flood frequency analysis. Inundation information is extracted from an existing 2D hydrodynamic flood model (MIKE 21) for 2year, 10year, 20year and 100year recurrence intervals. Arc-GIS is used to generate the flood hazard map. Flood hazard map has been developed considering two combinations of hazard parameter. In the first combination, flood depth and flooding duration were considered, where a critical date ‘120 julian date of year’ is used to determine flooding duration. Hazard maps were prepared on the basis hazard magnitude. In the hazard magnitude depth and duration index were used corresponding to four different depth and duration ranges. From the hazard assessment it is found that flood area increased with flood intensity. Higher flood depth increased and lower flood depth decreased with an increase in flood intensity. Flooding of cultivable land indicated potential damages in boro crop production and negative effects on the livelihoods. Agricultural boro crop loss is estimated by developing a crop damage assessment model for haor basin. A typical depth damage curve is formed for the study area. Depth damage curve has been used to formulate damage function which is represented here by a stage damage curve. Also crop loss function is developed on the basis total production cost which is the secondary data of market price. Maximum damage occurs at the end of April which is the harvesting time of boro crop. For risk assessment, risks of 2Year, 10Year 20Year and 100Year return period flood events are calculated by developing a risk model. Risk map for the selected haor basin is developed that shows risk free, low risk, medium risk, high risk and very high risk zone based on different ranges of expected damage or risk. To reduce the expected damage some steps may be taken as, changing varieties of paddy, starting time of cultivation be changed, invest higher harvesting force before flooding, repairing damage location of submersible embankment.