A flood will occur somewhere around the world this year. And when it does we know what the experts will say: they're getting worse because of climate change. How do they know floods are getting worse? The same way they know everything is getting worse: because with climate change everything is getting worse. And in some places they probably are. But for the world as a whole, the last IPCC Report said, heavy precipitation events have likely increased in some places since the mid-20th century but “confidence in an overall increase at the global scale remains very low” (which doesn’t mean they believe it but need more evidence, it means on the evidence they think it’s extremely unlikely to be happening) while streamflow records showed no global trend. And before you dismiss the IPCC as a bunch of industry-funded denier crackpots, there’s more evidence where that came from.

A 2023 study by a group of French and Australian climate scientists gathered up rainfall and streamflow records from around the world to try and get a long-term perspective on the topic. The problem is that in many places the data only go back to the 1950s, but a century is needed to really nail down trends. So the authors got 100-year records from as many places as they could, then got even longer records of temperature, air pressure and windspeed data and did some statistical modeling to determine how well those variables seemed to correlate with precipitation. Based on that model they projected rainfall and streamflow records 180 years into the past, and created an index of streamflow and rainfall data as a general flooding measure.

For the period where they had actual observations, they found that North American and European precipitation and streamflow activity tended to trend in opposite directions, with western Europe, for example, tending to have less streamflow activity over time. And for the world as a whole, the combined flood index had no trend.

When they used their statistical model to project far into the past, they identified several years where it predicted floods in the US Northwest, then checked those predictions against historical records. And they found the match was reasonably good, although there were some discrepancies. For instance the model says there should have been a big flood in 1876, but they couldn’t find any evidence of one.

The authors concluded: “Overall, the results in terms of trends (or lack thereof) are remarkably consistent with the literature.” They found that heavy precipitation events have increased since the 1950s and even farther back. But flood events have not: “This is also in line with the lack of globally-consistent flood trend reported in the literature, suggesting that this negative result is not due to the relative short period used in most flood analyses.”

So yes, floods happen, and one or more will happen somewhere this year. But despite the alarmist slogans, they aren’t getting worse.