In 1987, however, the Bureau was facing a crisis in respect of its equipment and, more acutely from BMRCs point of view, its computing power. With the strong backing of the new Minister for Science, the Honourable Barry Jones, and the Meteorology Policy Committee, the Bureau finally gained the funds necessary to allow for an adequate level of staffing and administrative support of the BMRC, including the purchase of a Cray X-MP/14 supercomputer in 1988. This support has continued, albeit at a lower level than previously, permitting the purchase of another Cray supercomputer in 1992, which quadrupled the Bureau's numerical modelling capability. And whilst the Bureau has been required to make staff cuts in response to the government's efficiency drive, BMRC has not, as yet, felt the full effect of these. This is a result of the present Director of Meteorology's strong commitment to Bureau-based research. As in Bill Gibbs' case John Zillman came to the directorship via the research arm of the organisation and views research as an integral component of the Bureau's activities, in both a national and international context (Zillman in a personal communication).

Whilst BMRC is the Bureau's major research group, its emphasis has changed slightly recently to include some responsibility for ensuring that its research can be properly utilised by the day-to-day forecasters. In doing so, it liaises with the relevant Head or Regional Office section which then tailors the new techniques to suit local practices and climatology. This approach overcomes the demarcation problems which so plagued CMRC and ANMRC mentioned earlier in this paper. At the same time, the outlying sections still have the expertise to undertake their own small-scale research and development work without having to involve BMRC. No doubt this is how Sir George would have wanted the system to function, with BMRC operating as a mainly autonomous entity within the larger utilitarian department responsible for daily forecasting duties.

Bureau of Meteorology