MPs have backed calls for boosted supercomputing capacity at the Met Office, saying that the benefits of improved severe weather warnings and better long term forecasts could provide "as much as a 10-to-one return on investment".
In evidence submitted to the House of Commons’ Science and Technology Committee, the Met Office said it would cost £42 million over five years to build a supercomputer with twice the processing power of an "almost one Petaflop" system that is currently being implemented. This would allow it toprovide more accurate short term extreme weather forecasts, as well as long term forecasts on a scale of months or years.
The Met Office said that forecasting improves at a rate of about one day per decade, meaning that a three day forecast today is as accurate as a one day forecast 20 years ago. The latest proposed improvements have been demonstrated in research, the Met Office said, but cannot be implemented across the weather service due to lack of supercomputing capacity.
The committee noted that the Met Office’s move from the Ministry of Defence to the Department of Business, Innovation and Skills had created the potential for the Met Office to become more collaborative and expand its commercial activities.
"While we recommend that the Met Office continue to expand activities that generate commercial income, this must not put core services for the public sector or the Met Office’s international reputation at risk," the committee said.
The committee also touched on data sharing in its report, which comes after the Met Office opened its data to the public via the Public Data Corporation. Asked by committee chair Andrew Miller MP whether there were private sector data sets that the Met Office couldn’t get its hands on, Professor Ed Hill of the National Oceanography Centre said that the Met Office’s job would not be possible without data from the private sector.
"The private sector certainly contributes to the collection of weather data. For example, out in the oceans, ships of regularly collect Met data. It is an important part of the data stream. It would not be possible without them."