Improved flood-related decision-making in case of urban flash floods in a metropolitan city of India

<p>Occurrences of short-duration extreme rainfall have significantly increased over India, leading to frequent flash floods. Growing incidences of urban floods pose a challenge to rainfall forecasting agencies and disaster mitigation authorities. Advancement in the numerical weather prediction (NWP) models has resulted in improved skills of rainfall forecast for longer lead times. However, in recent years, there is a growing emphasis on developing an impact-based approach to communicate the probable impacts of the forecast and reduce the socio-economic losses. In this study, we aim to generate Impact-Based Forecasts (IBFs) in response to the growing incidences of urban flash floods in metropolitan cities of India such as Mumbai. IBFs will provide warnings about the potential impacts as well as communicate protective responses based on the category of impact, i.e., high, moderate, and low. To this end, an inventory of several urban floods over the city of Mumbai during the past decades is prepared, and the relationship between past extreme hazards and related impacts is investigated. Various available Quantitative Precipitation Forecasts (QPFs) from the European Centre for Medium-range Weather Forecasts (ECMWF), Japan Meteorological Agency (JMA), UK Met Office (UKMO), and National Centre for Medium-Range Weather Forecasting (NCMRWF) will be used in the study. Moreover, several observation datasets, such as from the Indian Meteorological Department (IMD), and from Integrated Multi-satellitE Retrievals for GPM (IMERG), will be used to validate the forecast information. The raw precipitation forecasts will be post-processed using a Bayesian joint probability (BJP) model-based rainfall post-processing approach to improve reliability and accuracy. With this study, decision-makers are expected to gain crucial insights regarding the probable impacts arising due to multiple realistic flash floods in Mumbai scenarios. The analysis is underway, and the results will be presented at the conference.</p>