Design of Water Storages, Etc (continued)

Neil Body recalls that during the period 1959 to 1962 the Hydrometeorological Section was involved in two main areas of investigation, the development of flood forecasting systems and flood estimation for the design of dam spillways . We shall examine the flood forecasting aspects in later paragraphs.

Neil remembers that the first flood estimates for dam design were for the SEC in Victoria and the HEC in Tasmania. The Victorian design required estimates of the one in twenty year flood, and the maximum possible. These estimates included not only the magnitude of the rainfall but the hydrological estimate of streamflow, Allan Rainbird working on rainfall and Neil Body on streamflow. The figures they produced were so large that the HEC people asked them to reduce their estimates which they refused to do. Later experience confirmed that the magnitudes of their estimates were realistic. The reason that HEC engineers were frightened by the magnitude of the estimates was that the cost of constructing dams to withstand such large values of rainfall and streamflow was likely to persuade the State Government to delay the projects.

In his notes Neil Body paints an interesting picture of Professor Crawford Munro of the University of NSW Engineering Department, who was a strong supporter of the Hydrometeorological Section. Neil describes Crawford Munro as not one of your normal run of academics. He got things done and developed one of the most successful schools of civil engineering in Australia. He was particularly impatient with those who, like Geoff Alexander of the SRWSC of Victoria, quibbled about the statistical approach to hydrology. Munro was a great supporter of the use of probability paper in giving a first look at a statistical frequency distribution and at that time, with the help of Uwe Radok, I had become persuaded that this was a useful tool. Neil recalls that when Gerry O'Mahony, fresh from his academic training in statistics at ANU, scoffed at the use of probability paper, Munro suggested that such an approach "was alright for you academic bastards but an engineer needed to get a feel for the data so he could make a proper interpretation". I believe that Munro had identified an important use for probability paper since it provided a very useful appraisal of the nature of the frequency distribution of rainfall and streamflow.

Bureau of Meteorology