Located on the Chao Phraya river basin, Bangkok has long been prone to flooding events, impacting the local economy. Green spaces, which act as a retention area during flooding, have greatly decreased. While the city recognises the need for more water retention areas, development planning is yet to incorporate flood risk management. With climate change likely to increase flood risk due to higher intensity rainfall. Sea level rise will also increase the risk of tidal, riverine and storm water flooding due to the interactions between these three systems.

As part of the Global Future Cities programme, Mott MacDonald in partnership Bangkok Metropolitan Administration (BMA), team of engineers, analysts and computer scientists have been helping to enhance the city’s ability to prepare and respond to flooding caused by high-intensity rainfall and identify mitigation solutions.

Together, the team have developed a Decision Support System (DSS) for Flood Management to improve flood prediction, emergency preparation and response in one pilot catchment area of the city. Meeting the challenges presented by impacts of flooding, which have been exacerbated by rapid urbanisation, the DSS is one of three interventions being carried out in the city by the Global Future Cities programme. 

As a global first, asset optimisation and digital innovation has enabled the DSS to increase BMA’s ability to undertake short, medium and long-term planning for flood risk management and optimise flood mitigation strategies and related investments, while additionally reducing traffic disruptions and emissions.

The DSS integrates world-class rainfall estimates, gauge data and rainfall and flooding predictions to support flood management. Significant improvements in rainfall radar observations facilitated automatic generation of near real-time rainfall maps and operational rainfall forecasting. Machine learning is then applied to a hydraulic flood model to produce the storm water flood forecasting, which converts rainfall into flood maps in seconds.