Climate variability of tropical cyclone impacts in the North Indian Ocean and exploration of risk reduction strategies for Bangladesh

Tropical cyclone (TC) is a well-known natural disaster that can devastate much of a society, environment, economy and result in people’s deaths. The North Indian Ocean (NIO) is one ocean basin that is very prone to TC. TCs often cause huge human casualties in densely populated communities like Bangladesh, India and Myanmar around the Bay of Bengal (BoB) region of the NIO. Unlike those in the other oceans, the TC frequency over the NIO has a unique bimodal annual distribution, and reaches it peak during the pre-monsoon (April-June) and post-monsoon (October-December) seasons, respectively. Past studies focused on climatology of TC in terms of large-scale environmental parameters as well as interannual climate variability, for example El Niño Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD) for the NIO. More studies have been conducted in the BoB than the Arabian Sea (AS). This dissertation has adopted an interdisciplinary and integrated approach to examine TC impacts, which include climate variability analysis, future impact assessment and study on TC risk perception. The specific aims of the dissertation are to: (i) analyse the teleconnection patterns of different types of El Niño events on TC formation, development and landfall over NIO; (ii) identify the climate factors accounting for the asymmetric active or inactive TC seasons over AS and BoB; (iii) estimate potential impacts of TCs from storm surges and inundation for the rim countries of NIO for past and future climates; and (iv) explore cyclone risk reduction strategies for coastal Bangladesh based on a field survey on risk perception. For aim (i), a new type of ENSO the central Pacific (CP) El Niño, or known as El Niño Modoki, has been identified recently and its impact on TCs in the NIO is still not well known. Under this CP-type ENSO, the oceanic warming pattern is very different from that in the canonical (EP)-type ENSO. This study reveals that although there is no significant change in TC number and tracks during the major CP and EP years, there is a general reduction in the number of TCs during the post-monsoon season in the CP years. Genesis Potential Index (GPI) analysis shows that the mid-level humidity and vertical wind shear contribute to the suppression of TC formation during CP years. These changes in humidity and vertical wind shear are due to the large-scale circulation pattern originating from the Pacific in response to the CP event. TC activity in AS have become more active in recent years. During 1983-2015, AS was very active in a few years (but with quiet BoB season) and the opposite occurred in some other years. Aim (ii) seeks to identify the climate factors for this asymmetry. Results reveal that no single climate mode (such as ENSO and IOD) can well explain the TC development concentrating on the AS or the BoB only. Instead, it is found that variability of the northeast (post) monsoon is an important factor responsible for the difference between the two basins. Excess moisture is available over the AS due to anomalous low-level flow from the equatorial Indian Ocean in the years when there are more TCs in that basin, and the BoB is subjected to drier conditions. There may be climate change signal when TC activity increases over the AS. For aim (iii) of this study, storm surges and inundation at the coastal areas due to TC landfall for the past (1990-2010) and future (2075-2099) climate are analysed via a storm surge model and Resilience-Increasing Strategies for Coasts Toolkit (RISC-KIT). While almost all intense TCs had historically made landfall around the BoB region, climate model projection simulated more intense TCs over the AS. The storm surge model and RISC-KIT estimate an average of 3-5-meter inland flooding depending on TC intensity and topography, and more high-impact cases will be shifting to the AS. An impact assessment tool, InaSAFE, is also applied to estimate the social and economic impacts of TC for the Bangladesh region. To accomplish aim (iv), an extensive survey that consisted of face-to-face questionnaire surveys and site observation was undertaken in the southern coastal areas of Bangladesh to measure perceived and actual risk for TC impact. Results indicate the average level of perceived risk is high both at the household and expert level. Female participants perceive slightly higher risk than males in coastal Bangladesh. Moreover, income, occupation, farm size and educational attainment/qualification influence risk perception to some extent. This survey advocates some cyclone risk reduction (CRR) measures, as follows: build more cyclone shelters; build cyclone resilient private houses; build or improve embankments and polders; build or improve roads; preserve dry food; supply adequate food and pure drinking water; and improve the quality of early warning messages or alarms. This study concludes that risk perception assessment is a prerequisite for implementing any risk reduction plan or strategy, and it is critical to integrate bottom-up and top-down approaches for CRR plans and actions.