The Meghna Basin in the North Eastern (NE) region of Bangladesh is prone to frequent flash floods [15]. The flow in this basin is influenced by various transboundary sub-basins; however, the primary trigger of flashiness in this region is the orographic precipitations in the Meghalaya region [16]. The occurrence of flash floods during the pre-monsoon period is a notable aspect of this region. In June 2022, an unprecedented amount of rainfall in Meghalaya caused a severe flood in the northeastern portion of Bangladesh. Approximately 70 to 80% of the Sylhet and Sunamganj areas were submerged in water [17]. Owing to the effects of the rapid climate change on meteorological characteristics, the frequency of occurring hazards of comparable magnitude will be escalated not in the distant future. Additionally, the capricious behavior of flash floods, and scant response time, made it difficult to provide accurate and useful early warning with necessary lead time resulting in severe damage to the local agriculture and households. The Flood Forecasting and Warning Center (FFWC) [18] in Bangladesh provides observed and forecasted water levels – deterministic and probabilistic – on a daily basis with 5 to 15 days lead time[19]. Here they utilize coupled Hydrologic-Hydrodynamic modeling where rainfall estimates from the Weather Research and Forecasting (WRF) model are utilized to predict the discharge using the hydrologic model under the MIKE 11 NAM package. Later, the estimated discharge is used in the Hydrodynamic model of MIKE 11 HD packages for the BWDB station points to forecast the water level with 1 to 5 days of lead time. Additionally, they provide structured-based flash flood forecasting of the water level with respect to the height of the submersible embankments situated in the basin regions locally known as Haors. However, the rapid response of rainfall-runoff of flash floods hinders the capability of the coupled models to accurately provide forecasts more that 24 hours which consequently shifted the focus towards rainfall-based early warning of flash floods that depends on the rainfall forecasts of IMD in Meghalaya. Hence, it has become essential to focus on alternative early warning mechanisms, for example, rainfall-based flash flood forecasting, for alleviating these damages.
Atmospheric Rivers (AR) – a climatic driver – have a substantial impact on extreme precipitation events and flash floods. When a moisture-laden air mass reaches land and encounters elevated topography such as mountains, it is compelled to rise, resulting in concentrated precipitation in the form of heavy rainfall or snowfall across a relatively limited region [20-24]. An AR is a narrow and extended pathway of highly concentrated moisture within the Earth's atmosphere, typically that extends vast distances of thousands of kilometers across oceans [25-27]. Browning and Pardoe’s [28] study delineate certain physical attributes of Low-Level Jets (LLJ) that have since become linked to AR. Their schematic illustration of the LLJ emphasizes the frontal boundary at higher altitudes, the presence of warm and humid air preceding the front, and the occurrence of convection and vertical –movements related to frontal dynamics. Carlson [29] showed that warm and cold conveyor belt (CCB) wind flows are connected to tropical cyclones. Heini Wernli, in his thesis based on the Cold Conveyer Belt [30-32], precisely defined the WCB as the area of upward motion. Newell et al. [33] defined "tropospheric rivers" as thin streams of concentrated water vapor transport that can be observed moving in a meandering pattern across the atmosphere. Nonetheless, several years later until Zhu and Newell [34] named the phenomenon "Atmospheric River" based on their previous findings. In addition, they illustrated that atmospheric rivers were situated beside the cold fronts connected to the cyclones in the mid-latitude, indicating a dynamic connection to track of the extratropical storm. These outcomes yielded visual representations of the global-scale transportation process as well as gathered statistical data on the worldwide movement of water vapor. Additionally, around 95% of poleward water vapor flow towards the mid-latitude regions is attributed to ARs. However, ARs only cover around 10% of the globe's circumference, highlighting the slender geometry and fast movement of moistures along these atmospheric rivers.