DESPITE increased understanding of a number of different climate processes and their impact on a range of different timescales, this knowledge is not being used to inform planning and decision making. This is because long-term climate risk is often viewed only in statistical terms.
For instance, engineering techniques for estimating flood risk, where records exist, are largely based on simple statistics of their historic occurrence rather than on any real understanding of the processes that actually cause them.
In essence, if we have 100 years of flood record, then the largest flood measured represents, more or less, the hundred year flood level. This hundred year flood level is probably the most important of hydrological statistics in terms of its use in planning management. It is the yardstick by which decisions are made.
The problem with a simple statistical representation of risk is that it implies a static climate – the expected flood risk is equally likely in any year, irrespective of the actual climate processes that may or may not be dominating at that particular time. If this were the case, we would expect to see an equal spread of floods throughout our historical records. In Australia and in fact many other parts of the world, this is not so.
Updated below with three papers by Stewart Franks