9

In public media the notion that climate change is more than global warming gets repeated frequently.

Are there strong predictions that show that the climate becomes more irregular? Would we expect the weather in any particular place to suffer greater extremes (of temperature, precipitation, wind, etc.)? Are extreme weather events likely to increase (e.g. hurricane frequency or intensity, major wildfires...)?

The claim is repeated often in the media. Here is the Guardian's recent take (decrying the fact that some US sources fail to make a link that is widely accepted elsewhere; highlights are mine):

The phrase "extreme weather" flashes across television screens from coast to coast, but its connection to climate change is consistently ignored, if not outright mocked. If our news media, including – or especially – the meteorologists, continue to ignore the essential link between extreme weather and climate change, then we as a nation, the greatest per capita polluters on the planet, may not act in time to avert even greater catastrophe.

An older archived link from the BBC also makes the point:

Increasing temperatures means the World is likely to see less frosty days and cold spells, but we are expected to experience an increase in heat waves and hot spells

Greater risk of drought in continental areas

Increase in extreme precipitation events

Hurricanes likely to be more intense in some parts of the World due to more rainfall and more intense winds

An intensification of the Asian summer monsoon is expected

Are these reliable predictions or global climate models or are they just the material that makes news a little more interesting thereby leading to coverage being dominated by the less solid parts of climate science?

On of the key things that separates a scientific prediction from the kind of predictions that you read in newspapers is that science isn't vague. A scientist should be able to give you a probability for whether the event he predicts happens.

NB the question is about whether weather variability changes, not just the average weather. That is, is the normal range of variation higher across a year as a result of climate change.

Community
  • 1
Christian
  • 33,271
  • 15
  • 112
  • 266
  • Less regular than what? Past years (annual precipitation, number of hurricanes, etc.)? Seasonal variation (longer summers, shorter winters, etc.)? – Beofett Jul 08 '11 at 16:05
  • I specified entropy and variance as metrics for regularity. – Christian Jul 08 '11 at 16:39
  • 7
    This question would be better if you referenced specific claims, I think. – Craig Stuntz Jul 08 '11 at 17:00
  • If we believe the scientists, it means whatever we observe right now plus doomsday heat eventually. – Captain Claptrap Jul 08 '11 at 20:42
  • 2
    IMO you're misunderstanding the claim. The claim is trivial: that climate change means that weather will become less 'normal', less 'like it used to be' (maybe that's what you mean by "less regular"). Depending on where you are, that might be warmer, on average: or it might mean cooler, or drier, or more rainy, or windier, etc. – ChrisW Jul 09 '11 at 00:59
  • Wouldn't it be better to ask the simpler question whether extreme weather events (e.g. storms, floods, droughts) will become more common? – matt_black Sep 29 '11 at 21:43
  • Are you asking whether the media is quoting mainstream scientists correctly, or whether the mainstream scientists are correct in stating that we'll undergo "global weirding"? – Andrew Grimm Feb 04 '12 at 01:24
  • By current community standards, this question doesn't make a notable claim. It is also unclear, as the comments above, and the misdirected answer reveal. Closing until it can be fixed. – Oddthinking Feb 15 '12 at 02:36
  • 1
    @Christian hope you don't mind the big edit: I wanted to see if this would get it reopened. – matt_black Jul 09 '12 at 14:31
  • @matt_black: I have no issue with your edit. – Christian Jul 09 '12 at 14:44

2 Answers2

2

Yes, climate change does mean that weather becomes more variable in addition to being warmer (Financial Times link, registration required).

Global warming “significantly” increased the odds of some of last year’s most unusual weather, including the brutal Texas drought and the freakishly warm November in Britain, according to findings released Tuesday alongside the latest “state of the climate” report in the Bulletin of the American Meteorological Society.

The paper is available online at DOI:10.1175/BAMS-D-12-00021.1 - Explaining Extreme Events of 2011 from a Climate Perspective

Dr Peter Stott and others have been using attribution science to look at these causes. See, for example, Human activity and anomalously warm seasons in Europe:

The attribution analysis described here employs temperature data from observations and experiments with two climate models and uses optimal fingerprinting to partition the climate response between its anthropogenic and natural components. These responses are subsequently combined with estimates of unforced climate variability to construct distributions of the annual values of seasonal mean temperatures with and without the effect of human activity. We find that in all seasons, anthropogenic forcings have shifted the temperature distributions towards higher values.

(my bold in each quote)

See also Tamino on Climate extremes and anthropogenic climate change and Pete Sinclair on "Welcome to the rest of our lives".

Finally, in the IPCC's 2012 special report on climate extremes (SREX), we find:

There is evidence that some extremes have changed as a result of anthropogenic influences, including increases in atmospheric concentrations of greenhouse gases. It is likely that anthropogenic influences have led to warming of extreme daily minimum and maximum temperatures at the global scale. There is medium confidence that anthropogenic influences have contributed to intensification of extreme precipitation at the global scale. It is likely that there has been an anthropogenic influence on increasing extreme coastal high water due to an increase in mean sea level

L2G
  • 317
  • 2
  • 9
410 gone
  • 4,701
  • 30
  • 45
  • 2
    What does "raise the odds" mean in term of numbers? Is it big enough that "extreme weather events are likely to increase". Whenever someone does real science there should be somewhere numbers that are more specific. – Christian Jul 11 '12 at 14:21
  • 3
    The Texas drought is part of a cyclic weather pattern at least 800 years old. It strikes the southwest approximately every 35-40 years. The last time it was in the 1970s in my childhood and the time before caused the dust bowl. When the pattern occurs, tornado alley moves northward and hurricanes are deflected to the east coast. I would point out that it was Stott and the Met Office that incorrectly predicted a warm winter last resulting in a serious misallocation of resources in the face of an unusual cold snap. – TechZen Jul 11 '12 at 22:34
-1

Entropy and Variance are different kind of quantities. Variance is a quantity to describe mathematically the fluctuation of a distribution of measurement values for e.g. Temperature. Temperature is thermodynamical linked to Entropy of a System (defined by the probabilistic model of a distinct system). Entropy is physcial prinicple and used in climate modeling to rule out distinct unphysical/unlikely scenarios. Here is (Paywall) article using a maximum entropy method to calculate local climate annual cycle. AFAIK this method plays no important role in several current IPCC climate models. But some do write papers & blogs about it, judge yourself if its a discussion or monolog ;)

But IMO your question basically is, will the variance of global average temperature development be higher than in the past.

http://en.wikipedia.org/wiki/File:Global_Warming_Predictions.png

This past-future prediction of global average temperature shows imo the answer. If you compare the fluctuations of the average temperature of the red lines in past and future, you cant make up a big difference despite it gets hotter. These models are based on calculations and assumptions how climate parameters will develop in future (economic growth, environmental changes,...). Also climate can become less predictable (doesnt have to mean less irregular!), which would mean, the variance between different climate model scenarious inreases, as models are based on physics of deterministic chaos (e.g. additional physical processes, new environment influencing technologies). The more unknown parameter, the less predictable these model calculations get of course. But for a single line (model) in the picture there is no much bigger fluctuation of the average global temperature (e.g. no sinus like cycles) despite the rising trend. Interesting would be how variance of global avg. temp. relates to 0-1900 but this is outside data and simulation capabilities to make senseful comparisons

Beside this you have to distinguish between climate and local weather. Weather will actually become more irregular in sense of more or stronger extreme weather. Here is a scientific consensus by a majority

Werner Schmitt
  • 951
  • 6
  • 15
  • 1
    There are other things than temperature that make up climate. Will hurricanes and tornadoes become greater dangers? Will rainfall become less predictable? Pile enough weather together and it becomes climate. – David Thornley Jul 09 '11 at 18:04
  • @david of course. But whats the point of your commment? He asked on global temperature variance. I will not argue here what climate model is most reasonable.....and still in Physics the Energy of a system (often linked to temperature) stays a defining parameter. Or you disagree here?! – Werner Schmitt Jul 09 '11 at 18:25
  • **its funny that this answer only quoting facts of wikipedia and physics gets downvoted without hint whats actually "wrong"** Did truth hurt some climate change skeptics and oil fanatics? ;) Good opportunity to see if the rating system actually works for the type of users here... – Werner Schmitt Jul 09 '11 at 22:26
  • Entropy is a valid statistical metric. Say I look at a year with 182 days with 15 C° and 183 days with 20 C° the variance is the same regardless whether the first 182 days have 15 C° and the last 183 days have 20 C° or whether there a 50% chance that a day with 15 C° follows after a 15C° day. The data sets however have a radically different Shannon entropy. | It's important to keep different meanings of "entropy" and "temperature" in mind because both are used in a specific way when you talk about computer modelling (Markov Chains etc.). – Christian Jul 09 '11 at 22:45
  • 1
    Without knowing about the error bars of the models in your graph you don't know whether those models make reliable predictions about the variance. – Christian Jul 09 '11 at 22:50
  • @chris classical case of inexact question-inexact answer.The time scales (year, days) you use have nothing to do with climate change models, thats out of their scope. There is no point in calculating variances on this time scale for a single model. You see the variance of the different models 2100??? Also mention shannon entropy if you mean it in your question, climate change- temperature-entropy in one sentence obviously means physical entropy. Thats kind of ugly post-reformulating of a question you didnt know what you were asking for. Next time do your homework before on such a complex topic – Werner Schmitt Jul 09 '11 at 23:17
  • @Werner: If you think that I meant physical entropy than your paragraph about the "maximum entropy method" is misleading. _By **analogy** with thermodynamics, two approaches are presented, involving **constraints** on the properties of individual climate models or of ensembles of climate models._ There are three broad classes of relevant entropies. Information entropy of temperature, information entropy of the contraints/cost functions of a model and thermodynamic entropy. I'm interested in the first class. You quoted a study that used the second and charge that I meant the third. – Christian Jul 10 '11 at 11:47
  • @WernerSchmitt let us [continue this discussion in chat](http://chat.stackexchange.com/rooms/768/discussion-between-christian-and-werner-schmitt) – Christian Jul 10 '11 at 11:53
  • @chris reformulate your question much more precisly (i answered everything on "irregular") or ask on physics.SE. I doubt you get much votes there as misleading and inexact it is now. The entropy in the papers is based on statistical physics/thermodynamics, not information theory. "information entropy of temp." is pseudo-science to me, what a wierd term. Look yourself how much google hits you get with this...i strongly doubt you get any other answer here. The diff. climate model scenarious above show, there is not more irregularity for prognosed climate by diff. models. – Werner Schmitt Jul 10 '11 at 17:07
  • @Werner: _Based_ is a tricky word. The entropy in those papers is based on thermodynamic entropy in the same sense that econophysics is based on Brownian motion. There a relation but the two aren't the same thing. Econophysics uses Brownian motion as an analogy. The paper you cite uses thermodynamic entropy as an analogy for it's own statistical model. – Christian Jul 10 '11 at 22:34
  • @chris my last comment on this site.Copypaste your question to physics.se.This still vague not specified question is outside the scope of the audience of this site, and yours probably too without back knowledge. Entropy is a hard to understand concept and surely no metric to measure temp. irregularity.Its used and defined as a hole system variable not to judge about irregularity, characteristics of single observable. It doesnt make any sense to compare entropy of one climate scenario with another, as different climate models lead to diff. entropy models, as, again, entropy is a system concept. – Werner Schmitt Jul 11 '11 at 11:23
  • The problem with this answer is that I think it misinterprets the intent of the question. I think the idea that the climate becomes more extreme (more unusual weather events such as storms or extreme droughts) is a reasonably clear one but one that is obfuscated by talking about the variance of model predictions. A better way to pose the question might be to ask whether extreme weather events will become more common. PS "maximum entropy" has nothing to do with weather variability but is better thought of as an optimisation technique for fitting models to data. – matt_black Sep 29 '11 at 21:41
  • 2
    Totally off the topic of these comments (Aside: I agree the question wasn't answered), but while I understand that the historical data would be highly fluctuating as displayed in the graph, I have to ask why the projected data fluctuates in the same way? How do you long term project short term fluctuations? Are the fluctuations mocked up to make the graph look consistent? Are they random, but based on the fluctuation rates in the historical data? Have they actually mapped the data to complex sinusoidal wave functions? – John Rhoades Feb 14 '12 at 19:57
  • In Maxwell–Boltzmann statistics a higher temperature implies that a greater fraction of possible states may be occupied: http://en.wikipedia.org/wiki/Maxwell%E2%80%93Boltzmann_statistics It's not unreasonable to suppose that extreme weather events follow this common statistical distribution with respect to temperature. –  Feb 14 '12 at 20:08
  • "consensus by a majority" sounds bit like "People's democratic dictatorship" :-P – vartec Feb 14 '12 at 20:22
  • The link to `springerlink.com` is broken. Is it supposed to point to the journal version of the arXiv article? –  Sep 17 '22 at 07:42