The independent review of Easter's flooding and the Environment Agency's performance was published to schedule on 1 October. Even after a miserable summer, the exceptionally heavy and prolonged rainfall which flooded thousands of homes, mainly in Warwickshire, Northamptonshire and Oxfordshire, still sticks in the mind. Five people died.
Two men set out to draw lessons from the disaster: Peter Bye, until last year chief executive of Suffolk County Council and the county's emergency planning convenor, and technical expert Dr Mike Horner, managing director of Bullen Consultants. They have completed a formidable task in a fraction of the time required by a public inquiry.
True to form, most newspapers last Friday were full of dramatic flood photographs, and all reported calls by local MPs for the resignation of EA chairman, Lord de Ramsey. He has yet to make a public visit to the affected areas or venture any opinion on what happened.
I have no views on whether anyone should resign. It is far simpler to report that somebody must be made to pay for what went wrong rather than to detail those technical findings which could be applied with benefit elsewhere. Why? Risk management is not exact. Flood defence is not flood prevention. People do not like to be told that their house might be flooded or that a bridge could collapse.
There is one fundamental lesson from the review, and that is how easily computers can lead everyone astray.
The Environment Agency is required under the Water Resources Act 1991 to undertake surveys to define flood risk areas. Although lots of work has been done, the exercise is far from complete and progress differs widely between regions.
As far as non-tidal rivers are concerned, the Agency wants to issue maps showing the flood plain limits for a 1% annual probability - more widely known by tradition as a return period of 100 years. To have any hope of getting the job done computer models must be used to show how far the water spreads, and the models need to be calibrated with actual water levels and flow measurements from previous records.
When it comes to predicting what is likely to happen across long stretches of countryside, it turns out that computer models are reasonably accurate. But when applied to towns, flow behaviour around bridges, weirs and low level ground characteristics are too complex for simulation. The programs still use formulae which older engineers used with slide rules and logarithms, not many decades ago!
Computer models are also vulnerable to error when extrapolating from lower, more frequent, flooding events to the rare big flood which leads to tragedy and damage.
The independent review finds that EA flood limits are often no more than crude estimates, 'particularly where hydraulically complex urban reaches have been computationally modelled, whereas the assignment of a specific return period implies a spurious notion of precision'.
As one regional flood committee chairman put it this week: 'I keep telling them to mark the boundary with a wide felt-tip pen - not a pencil.'
Ask yourself on your current projects: How might blind reliance on computers be a trap waiting to catch you out? Who will cast the first stone?
This year's Easter floods will lead to a more rapid and accurate mapping of likely flood levels. Planners will be required to pay more heed to the EA when it objects to future developments on flood plains. Developers who do get permission to build will be required to pay substantially more to cope with the subsequent protection against extreme floods. The Agency will use its powers more rigorously to compel owners of water courses and flood defence structures to maintain them properly.
The Bye-Horner report is worth a read, even for those not directly involved. The authors show how much can be learned from a detached examination of routine procedures. It is only a pity that tragedy had to precede the review.