Deciding whether to grab a brolly before stepping out of your home could become easier if a new forecasting and prediction system takes off here.
Currently, forecasts indicate the weather for only a general area and time of the day, such as whether there will be showers in the western part of Singapore. The forecasts are updated on a three-hourly basis.
But a new computer system is in the works that could predict weather conditions more precisely over a 1 km by 1 km area, on a shorter two-hourly basis.
For example, a weather forecast for the specific Orchard Road area could be available, as opposed to a forecast for the central area, for a two-hour period.
The National Environment Agency (NEA) and technology giant IBM said they are working on such a system, which can also forecast air quality, predict hot spots for dengue outbreaks, as well as detect hot spots for food-poisoning outbreaks.
The three-year agreement to build this system, which will cost some $13 million, was inked by the two organisations at Marina Bay Sands yesterday.
The system will harness real-time data from NEA's environmental sensors and social- media feeds to help NEA and other public agencies in their operations.
For instance, if someone falls ill after patronising a restaurant and complains about it on social media, NEA officers will be able "take more timely action" against food-preparation malpractices, said NEA's chief executive, Mr Andrew Tan.
Predictive information about air quality could also be made available six to 24 hours in advance.
At the moment, readings based on the Pollutant Standards Index - a measure of air quality - are refreshed every 24 hours and do not give a forecast for the days ahead.
Mr Tan said the purpose of developing the system is to "manage environmental pollution and public-health risks so as to provide the public with useful environment-related lifestyle information to plan their daily activities".
The data will also be made available to the public through NEA's myEnv mobile app, Facebook page and website as quickly as possible, depending on the situation, he added.
IBM managing director Janet Ang noted that while IBM researchers can "test solutions in real-world settings" and use "Singapore as a living lab", the NEA can leverage on IBM's research expertise and experience gleaned from more than 2,000 city projects it has worked on around the world.
A similar system in Rio de Janeiro in Brazil, developed by the local authorities and IBM, has been able to help manage crises, such as predicting floods.
But Dr Jayant Kalagnanam, the research director of IBM's Singapore Research Collaboratory, told my paper that this floodprediction model is not being planned for use in Singapore at the moment, as the tropical meteorology of Singapore is different from that of Rio de Janeiro.
gerrardl@sph.com.sg
For more my paper stories click here.
