Following the successful installation of three rainfall X-band radars in 2018, a rainfall forecast model and a web interface to disseminate information, the Public Utilities Board of Singapore (PUB – https://www.pub.gov.sg/) has awarded H2i a follow-up project to further enhance this system’s capabilities and scope. Awarded early February 2020, three more X-band radars (supplied by Furuno Singapore) will become operational over the next 12 months. The new radars will be strategically positioned to cover the parts of Singapore that currently have the less coverage. This will bring the total number of rainfall X-band radars to six and as such a dense network of radars will cover the whole country soon.
H2i will apply Machine Learning techniques to further improve the accuracy of the rainfall observations. These rainfall observations are input for a nowcast model that predicts flash floods; in meteorology, nowcast refers to a short term forecast localized rainfall forecast of up to 2 hours. The existing nowcast model, which can predict movement, growth and decay of rainclouds, will be enhanced with Deep Learning techniques to better forecast heavy rain events 30 minutes before they actually occur.
This project has been featured by PUB in several media such as Channel News Asia (CNA), The Straits Times, The New Paper, Facebook and YouTube as well as on PUB’s homepage; reference to some of the links below: