Why is it so hard to say what the future North European climate will look like? A recent review paper by Shaw et al, 2016 explains the reason in persuasive terms.
The weather over Northern Europe is strongly shaped by low and high barotropic pressure, associated with so-called mid-latitude cyclones. Most people refer to the most extreme cyclones as storms, although meteorologists like to use the term “synoptic storms”. The North European climate may be regarded as the weather statistics for over this region.
Shaw et al. (2016) provide a thorough assessment of the current knowledge about global warming and mid-latitude storms. They argue that there is a “tug of war” between various conditions which will affect the outcome for future mid-latitude storms as well as the jet stream. Such conditions include the south-north temperature gradient, the vertical temperature structure, ocean temperatures, and cloud cover, many of which are difficult to calculate:
The relative importance of these effects in model simulations is likely to depend on model resolution and the representation of small-scale physical processes that are not explicitly resolved by today’s global climate models.
In other words, the projection of future storm tracks is highly sensitive to aspects that are not so accurately quantified by the global climate models. This high sensitivity gives rise to a well-understood source of uncertainty connected to the way a global warming affects the mid-latitudes and the storm tracks. These effects are pronounced for the locations that already are affected by storm tracks or where future storm tracks will end up.
The implications of this was highly relevant for a recent conference on climate change adaptation, NordicAdaptation 2016 (NA16) in Bergen, Norway. Mid-latitude storms are associated with pronounced variations in precipitation, temperature, and wind speed. The question is where, how strong, and how frequent will these storms be in a warmer world.
It was unfortunate that Shaw et al. (2016) came on-line on the first day of the conference, and was too late for adapting the presentations and the discussions. The emphasis of NA16 was on climate research to action and transformation, which implies setting up and offering a means for providing climate information (“climate services”) to decision-makers.
The challenges with storm tracks and global climate models have implications for the downscaling of regional and local consequences associated with a climate change. Hence, it is relevant for the CORDEX community because the storm track is reproduced by many global climate models, but many of them do not place them in the observed locations.
Because of these difficulties, both CORDEX and NA16 should address the gap between the climate research community and decision-makers who need to take climate information into account. There has been a view that regional climate model results will be used if they are readily accessible through a data portal such as Copernicus in Europe.
The use of climate data is probably best done through consultations with experts, a bit like medical doctors. An analogy for data portals is the drug stores, which often require a prescription from a doctor before selling a drug to avoid misuse. Medicines also come with labels. There is plenty of examples where information and data have been mishandled (see past post on learning from mistakes).
What do we say to decision-makers who are thinking about climate change adaptation?
For one, mitigation is key. There is also more information than what we can derive from climate models. This includes empirical analysis from observations, from information derived statistical theory, and a “bottom-up” approach (Pielke Sr & Wilby, 2012) that may involve a sensitivity tests to identify key variables and effects.
Hence, the outlook for mid-latitude cyclones still is unclear according to well-understood reasons explained by Shaw et al (2016), but for decision-makers, it’s a question of risk management and there is some useful information to act on. Furthermore, it is not only the direct effects that matter; One observation made during NA16 was that there will be both direct as well as indirect effects in a globalised world. A factory shut-down due to flooding in e.g. Bangkok may affect an economic chain, and persistent drought may trigger migration.
References
- T.A. Shaw, M. Baldwin, E.A. Barnes, R. Caballero, C.I. Garfinkel, Y. Hwang, C. Li, P.A. O'Gorman, G. Rivière, I.R. Simpson, and A. Voigt, "Storm track processes and the opposing influences of climate change", Nature Geoscience, vol. 9, pp. 656-664, 2016. http://dx.doi.org/10.1038/NGEO2783
- R.A. Pielke, and R.L. Wilby, "Regional climate downscaling: What's the point?", Eos, Transactions American Geophysical Union, vol. 93, pp. 52-53, 2012. http://dx.doi.org/10.1029/2012EO050008
apparent typo paragraph 4:
These effects are pronounced for the locations that already are affected by [where] a storm tracks or where future storm tracks will end up.
Other meanings are possible. Please delete this post.
[Response: I don’t understand your concern, Michael. – rasmus]
This problem of understanding regional scale responses has been on my mind a lot lately. I’ve found similar sensitivities on my work in the Tropics. Some of the robust mechanisms that we see on a global scale are really hard to see on a regional scale, even though we know they have to be at play at a fundamental level. And when we think we have a robust conclusion, there’s no way to know if it’s just a weird byproduct of multiple remote biases in things like SST or cloud fraction. I think the complexity of this issue is going to be vexing us for decades.
Concerning the use of results: If such results are communicated in some way to decision-makers, it is probably necessary that they are accompanied by an explicit explanation, for what purposes they can be used und for what purposes or questions they cannot (giving a short reason if possible). This does not detract from the benefit of direct ‘doctor’s’ advise – and advise on questions that have not been thought of when preparing the accompanying explanations can be added to those later on. But many users will not take the effort to contact someone. If there is information directly in the data file (important things in bold and with exclamation marks), this does not guarantee reading, but at least some people might realize…
There is not enough conceptual distiction between weather and climate in this article. What is the precise conceptual understanding of a climate model?
[Response: There is a difference between climate and climate models. Climate may be considered to be weather statistics, eg the probability associated with a range of weather states. The statistical character is determined by physical processes involved. A climate model makes use of physical laws – often simplified or approximated – to compute various physical atmospheric variables at various locations in the atmosphere (a 3D mesh). A climate model is designed to compute these variables and take into account slow processes that influence the long time scales (eg oceans and land-surface), but the output mimics weather states. The climatic character is then derived from aggregation of these results. – rasmus ]
Perhaps this means:
… uncertainty about how global warming will affect mid-latitude storm tracks is pronounced. As global warming progresses, storm tracks may become frequent in areas rarely affected in the past, and areas commonly crossed by storm tracks may experience changes in [strength? duration? intensity?] beyond that in the historical [fifty-year or hundred-year] record.
That’s still too many complicated words, even if I’m getting the idea.
[Response: Thanks for pointing out this ambiguity. Basically, the storm tracks may move, but it depends on many details which the climate models provide only tentative numbers. They could move poleward, which means that those regions where they frequently affect at the present may get drier. The storms also affect the temperature. Of course, those places where the storm tracks move to will experience a change to wetter climate with milder conditions in winter and cooler in summer. -rasmus ]
Thanks Rasmus.
Can you say anything about the size and intensity of storms changing, along with the tracks?
(Some change from moving across different topography, but, anything more than that)?
I know a lot of places build on erosion fans that haven’t seen significant flooding in centuries, but I recall Pujalte et al. about the dramatic change in precipitation intensity that accompanied geologically rapid warming in the paleo record and wonder if we’ll see that happening.
[Response:Good question, which was not answered by the paper. I guess this is still an unknown. -rasmus ]
Hmmm…Speaking of recent, this paper by Osprey et al might be relevant.
http://science.sciencemag.org/content/early/2016/09/07/science.aah4156
“Unprecedented atmospheric behavior disrupts one of Earth’s most regular climate cycles”
>> anything about the size and intensity of storms changing, along with the tracks?… I recall Pujalte et al.
>
> [Response:Good question, which was not answered by the paper. I guess this is still an unknown. -rasmus ]
Any, well, hunch or guess? Pujalte et al. were looking at the rapid rate of change at the PETM from am already hot condition.
We’re in a faster rate of change but starting from a cooler base condition.
Do the models have any help about how much moisture is apt to become available for sudden intense local storms, given those two differences in the facts? Less bad than what Pujalte et al. studied, I hope. Because worse would be very scary rapid flood events.
I guess my worst case is — starting from a cooler climate and warming faster into the very hot range would put _more_ moisture into the atmosphere and — still guessing — could make the storm intensity get worse faster — which would suggest focusing attention early on building on those erosion fans that are currently dry and apparently stable.
In other words, while looking at the coastline, occasionally looking over one’s shoulder at the hills might be worth recommending.
Just handwaving here. It’s “what’s the worst that could happen and has there been anything done to model this possibility”
Base on what we know about decisionmakers’ actions: they listen much more to gallup + profit margins + their own wishful thinking, than to science. Especially they don’t listen to climate science, when they decide how to act. Politicians in our times seems to think that they can somehow negotiate away the actions of natural laws. The Paris conference 2015 is a striking example. They are not even beginning to understand the seriousness of the situation.
Which weather systems are projected to cause future changes in mean and extreme precipitation in CMIP5 simulations?
First published: 23 September 2016
DOI: 10.1002/2016JD024939