Jun 18, 2013
Predicting Rhine floods
Storms and floods: over the last two decades Europe has received more than its fair share of these extreme weather events. Climate models suggest that the continent is going to have to brace itself for more as climate change kicks in. In particular, communities along the Rhine basin may be hit particularly hard. But how much can we trust the climate model predictions? A new study shows that climate models are failing to reproduce real observations when it comes to winter rainfall in the Rhine Basin.
Late winter is the time when people living along the lower Rhine look nervously to the skies. Westerly winds bring in moist air from the Atlantic, snowmelt bumps up the river levels in the upper Rhine and soils are frozen or saturated, creating the optimum conditions for the Rhine to burst its banks. The floods of February 1995 are a classic example, when heavy rain and snowmelt caused the Rhine to spill over in France, Belgium, Germany and the Netherlands, submerging vast areas of many cities and causing more than 250,000 people to leave their homes.
Winter rainfall is highly variable in the Rhine Basin but even among all of the noise it is possible to see an increasing trend. Current climate models simulate an increase in winter precipitation of between 12 to 23% per century for this region. Ronald van Haren of the Royal Netherlands Meteorological Institute and colleagues decided to compare how well these simulated figures matched the real data. Using winter-rainfall figures dating from 1950, they found that winter precipitation had increased by 14% (equivalent to 23% per century). “We found that the observed trend was outside of the modelled range,” says van Haren.
So, which aspects of European climate were the climate models failing to capture? “We looked at the individual events in the models and they look like the corresponding events in the observations: a series of low-pressure zones hitting Europe with lots of rain on the frontal systems,” explains van Haren. “The problem is just that the trend in intensity of these is weaker in the models than in the observations.”
Delving deeper into their data, the researchers split the precipitation trends into the part linearly related to changes in atmospheric circulation from the remainder (not related to circulation changes). Observations show an increase in westerlies, bringing in moist air from the Atlantic Ocean, but van Haren and his colleagues found that climate models were underestimating these changes in atmospheric circulation over the past century.
“We are still unsure about the nature of the observed circulation change; it could be a missing multi-decadal natural variation in climate models or it could be from anthropogenic causes,” says van Haren, whose findings are published in Environmental Research Letters (ERL). If it is natural variation then climate models may still provide us with reliable estimates for the future. But if the circulation changes are linked to increasing carbon dioxide and man-made climate change, then climate models are missing a vital piece.
Now, van Haren and his colleagues are taking an even closer look at the data, to try to establish precisely what is causing the mismatch between observations and model results. Ultimately, they aim to make sure that climate models do produce trustworthy predictions for the future.
- Evaluation of modelled changes in extreme precipitation in Europe and the Rhine basin Ronald van Haren, Geert Jan van Oldenborgh, Geert Lenderink and Wilco Hazeleger 2013 Environ. Res. Lett. 8 014053
- Ronald van Haren on regional uncertainty (PDF)
About the author
Kate Ravilious is a contributing editor for environmentalresearchweb.