Pearls and Irritations

Last Monday, a couple spoke to an ABC television reporter on the back steps of their home on the edge of Wollongong’s Lake Illawarra. They were confident that the flood they could see in front of them would not rise beyond the level it had reached. After all, they’d been living there for 19 years and no flood in that time had exceeded that level.

This reasoning is common and deeply flawed. It indicates how people are hard-wired to rely on their own experience and how little they understand, trust or rely on forecasts from authorities. It also shows how poorly people comprehend extremes in nature and how far beyond recently-experienced events nature can go in terms of severity. This goes for bush fires, floods, storms, tropical cyclones, storm surges, droughts, earthquakes, plagues and all the other perils of nature.

Not as infrequently as might be imagined, records in nature are not just beaten, they are beaten by large margins. A resident of Nyngan in 1990, having lived there for the hundred years of Nyngan’s existence to that time, could have been excused for thinking he’d seen all that floods could throw at the town. But the flood of that year peaked a metre higher than any flood previously experienced at Nyngan – and that in an area seemingly completely flat as far as the eye can see in almost all directions. The flood of 1990 brought four and a half times the amount of water through and past Nyngan than had been recorded in any previous flood there.

Or take Lismore at the end of February this year. The two highest floods of living memory, which struck in 1954 and 1974, had peaked at 12.1 and 12.2 metres respectively. This year’s event peaked at 14.37 metres – or more than two metres higher than the two biggest floods previously seen there. The record, established over the 150 years of Lismore’s history as a town, was smashed not by a little but by a lot.

In appreciating nature, we must think of the long term. This is not the period of one’s tenure in a location, or even the period of one’s life. We must think centuries, not decades, let alone years. And we must recognise how often, given the short period over which our records have been kept, records are broken. Indeed in 2007, a drought year in most of Australia, flood height or rainfall intensity records were set at some location or locations in every state. Sometimes, the new records beat the old ones by considerable margins.

Sooner or later, people in any location will see an extreme event: you only have to live long enough. But equally, you may not live long enough to see anything truly extreme. The other way of putting that is to say that some generations never see a genuinely extreme event whereas other generations might be unlucky enough to see one – or indeed more than one.

Take Maitland, on the floodplain of the Hunter River. Maitland was first settled by Europeans in 1818. In 1952, the highest flood recorded since that year struck – and it was eclipsed by a much larger one within three years. The 1955 flood peaked almost a metre higher than the flood of 1952 and was vastly more damaging.

The residents of Kempsey in 1949 saw the highest flood ever known there – and then the following year they saw a flood nearly as large. Thus the two biggest Macleay River floods of Kempsey’s history occurred within ten months of each other.

Grafton, on the Clarence River, tells an opposing story but one which also carries a cautionary message. Grafton has had many floods in its 180 years, but its biggest six have all been within half a metre of each other in terms of peak heights. When the authorities fifty years ago were contemplating constructing levees and calculating how high they should be built, they ‘deemed’ the then-record flood of 1890 as the ‘one-in-100-years’ event and the standard to which the levees should be constructed. That flood peaked at 7.9 metres on the local gauge, and the levees were built to that height plus 0.3 metres ‘freeboard’ to counter any inaccuracy in the estimate of the 100-years flood.

Unlike Maitland, Nyngan and Lismore, Grafton has not yet had an ‘outlier’ flood, a flood far bigger than any other seen there. It follows that Grafton’s biggest floods have all been of a magnitude that occurs quite frequently – perhaps of the order on average of only 20 years. The ‘on average’ is vitally important here: nature is not regular in these things. Grafton, in all likelihood, has not had a flood of genuinely extreme proportions.

This is shown by the fact that the ‘one-in-100-years’ event at Grafton is now thought likely to reach something like 8.35 metres. Today’s levees, designed to keep out what was originally considered to be a relatively rare event, might actually be expected to be overtopped quite frequently. Indeed Grafton has had four ‘close shaves’ in floods since the year 2000. The 7.9 metre flood, once thought to approximate the 100-year event, might in fact be more like the 20-year one.

The ‘one-in-100-years’ designation is dangerously flawed, though, and ideally the term should be discarded. It misleads people into thinking that having seen a flood of a size deemed to be a ‘one-in-100-years’ event they will not have to face another flood of similar proportions in their lifetimes. But it is not true that such a flood will not occur for another 100 years. A flood of that size or larger should more properly be regarded as having a 1% chance of occurring or being exceeded, each and every year, at any designated location (such as Maitland, Nyngan or Grafton).

The best way to refer to it is as a 1% AEP (Annual Exceedence Probability) event: this is the arithmetic reciprocal of the so-called 100-year event. In actuality, a flood of this size or larger could occur on a number of occasions in a 100-year period or not at all for several centuries. In these matters (unlike the seasons or the diurnal cycle of day and night), nature does not like regularity.

The so-called ’20-year flood’, accordingly, has a 5% chance of occurring or being exceeded at a location in any year. Again, this is a statement of average. It does not follow that three floods designated as (say) 5% AEP events occurring at a place in quick succession (as Windsor has experienced since early 2021) invalidates the measure or the expression. Such variability is quite normal: Windsor had not experienced a flood as big as any of these three since 1990. The concept of ‘average’ is central here and there is variability, not regularity, around it. Windsor has just had a short (16 month) period of flood richness after a long period (30 years) of flood poverty.

Don Bradman averaged, in round terms, 100 runs per dismissal in Test cricket. He did not score a neat 100 in each innings. In fact he made several ‘ducks’ and many scores between 0 and his best of 334. He had sequences of low scores as well as periods in which he made several big scores in a row.

There are policy implications in all this. School classes in maths could teach probabilities using examples from occurrences of events in nature. Likewise courses in geography or science could deal with the matter.

And the emergency services should take the moment when a bad flood has just occurred as a ‘teachable moment’ to get the point across and educate the wider population: such a time is likely to be one during which the media will be attentive (and looking for new angles of coverage) and audiences will be listening and receptive.

The expression ’100-year flood’ has caused untold problems because it is so badly misunderstood in the community. And that is not its only deficiency: it is the basis in Australia for the designation of the minimum floor levels for new dwellings in flood liable locations. This too is a matter where it causes problems.

But that is something for another piece in Pearls and Irritations.

Share and Enjoy !

SHARE TWEET SHARE COPY EMAIL PRINT