Design of Water Storages, Etc (continued)

There was considerable discussion in the Research Division of the best method of providing the information required. Statistics Section under J. V. Maher had a mass of daily rainfall data from some thousands of voluntary rainfall observers, some more frequent observations from Bureau offices and part-time observers and some pluviograph instruments which provided graphs of rainfall occurrence. The problem was that much of these data were not recorded on punch-cards at that stage and it was some years before all of the data were rendered into computer compatible form.

It was thought that if this could be achieved a statistical analysis could provide the answer. However some of the analysis required rainfall data of shorter duration than 24 hours and some other analysis required a network of observations denser than that available.

Another difficulty arose with the requirement for a statement of maximum possible rainfall. Even with an ideally dense network of hourly or three-hourly observations a data set extending for some hundred or more years would be required to obtain a sufficient number of rare events to make a reliable statistical estimate of the maximum possible rainfall. Having recently completed his year with Professor PAP Moran at the ANU Gerry was acutely aware of this but other 'statisticians' in the CSIRO and State and Commonwealth organisations could not be persuaded that a method of estimating maximum possible rainfall other than conventional statistical analysis was required. At this stage I must point out that the CSIRO had a highly efficient Statistical Division but some amateur statisticians preferred to ignore their advice. We shall encounter this phenomenon of human frailty under the heading of cloud seeding.

The alternative method of estimating maximum possible rainfall was by no means elegant but in my view was that likely to provide the best result. I believe that Colin Hounam might have proposed the idea originally but I remember much discussion with members of the Hydrometeorological Section and others in the Research Division.

Bureau of Meteorology