I have just returned from visiting family in Germany - I drove through western Belgium and Luxembourg on Tuesday 13th and it rained all of the way - but apart from being wet I did not realise the scale of the problem to the east. On arrival in Germany, (very close to the border with France and Luxembourg), the river Saar was at almost normal level but when we went to the Mosel valley at Trier the river level was 6.5m above normal. It was then that we started to get news of what had happened in the Eifel to the north where whole communities have been flooded, bridges destroyed, and sadly, many lives lost. In one part of the Hohe Eifel (High Eifel) they had 200 litres of rain over 1 sq m, (8 inches per square foot for those who use imperial measures), in 48 hours: no wonder they were reporting 10m flood waves on some rivers. The debris in the Mosel was frightening - whole trees being trapped on bridge parapets, and tragically parts of people's homes were floating down the river. Seeing that was awful. The standing waves in the Mosel below the weir at Trier were in the order of 1.5m high - that is extrordinary given that at normal flows there are hardly any waves at all. Along the Mosel down river from Trier people were pumping out basements and cellars, and all of the small streams and drains had flood debris and evidence of former very high flows.
An important lesson that we all need to learn from the fires in North America and the floods in Europe is that we have to adapt to changing circumstances: events like these are much more common than the experience of our short lives suggests. I looked at the photos of the damaged and destroyed buildings in Germany and was reminded of similar damage suffered by buildings during floods here in the UK. All of these buildings were very close to rivers which in the recent past have not had high flood levels. Flood prediction, and the prediction of most natural events is calculated by using probability models. Using probability data to calculate the frequency and magnitude of floods is all very well, but as with any model, it is only as good as the assumptions and measurements used to make it. Almost every flood model, (indeed every frequency/magnitude model), that I have seen, and I have seen many over the years, is based at best on data gathered in the 20th century, sometimes for periods which are even shorter. As a geologist who is used to studying long term change, I am appalled that anyone should claim that any event has a frequency of x or y years. Given that climate is an unstable phenomenon at all time scales, any predictive model based on a few decades of data can have little or no value beyond predicting the frequency of small to moderate scale events. Further, given that the climate is warming, larger scale events are going to become more common - and the historical data is there to show that this was so in the relatively recent past, (200-300 years), and that it was much lager in the geologically recent past too. The problem is that historical data is difficult to analyse without specialist training which very few people, including almost all climate scientists and engineers, have. Further the data, such as it is, is very patchy and is unusable in mathematical modelling.
Conclusion: we should not be building on low lying ground next to rivers or streams which are liable to flood unless we compensate by allowing large areas to flood in other places. We should be allowing small and large rivers to flow much more slowly and accept that if we dredge and straighten channels to allow for improved navigation, or build industrial sites on river banks and protect them with high banks, flood water will move as higher waves down the channel and cause destruction to lives and property elsewhere. If we clear upland areas of trees and then use heavy machinery or overgraze the land, water will run off very quickly indeed instead of infiltrating into the ground. All of this and more besides is well known, and some measures are being taken to mitigate some of the effects. However more difficult measures are not: they require that we accept that we cannot engineer nature to meet our needs and desires without there being consequences, some of them being potentially fatal to ourselves. In the past people often, (but not always), used their local knowledge and the experience of their forbears to avoid building in places that were potentially dangerous, even at intervals of over 100 years. If they did not they took risks. The same applies to the fires in North America - they have burned in the past and will burn again in the future - long after we and this forum are but a speck in history. That does not offer a solution to the immediate problems of course, but it perhaps might just remind ourselves that we live on a dynamic planet and that we are not as omnipotent as we sometimes like to believe: rather we are part cause and part participators in those changes. Accurate education is part of the answer, but sadly that is not being given anything like the priority it needs. Instead we are being fed apocalyptic stories because that helps to keep the media machine happy. It also stimulates our innate interest in the spectacular and dramatic.
We cannot bring back the lives of those killed recently in Germany, but we can think carefully about what actions to take to try to reduce as much as possible similar tragedies from occurring in future, while all the while accepting that such tragedies will occur. We will not live to see the burned out sections of forest regrow to their full magnificence in North America, but left to their own devices, they will do so in time. The news cycle is nothing compared to the length of our lives. Our lives are nothing compared to even quite moderate changes in large ecosystems. That too is worth remembering and keeping in mind at all times. The apocalypse is always just around the corner for some - it always has been - but the frequency interval of real apocalyptic events is actually quite long - even in geological terms!
Stephen.