If I understand it correctly, research suggests that Arctic warming is slowing the eastward progression of the northern hemisphere jet stream, so that weather patterns tend to be more persistent over any given location. If that’s combined with an intensification of the hydrological cycle then presumably we can expect both factors to contribute to more prolonged and intense droughts and flooding events. Does that make sense?
Is a study like this closely relatable to the other studies that document/predict and increased “acidification” of the oceans?
Does an increase in saltwater salinity increase the pH as well? Are there any perceived trends in salinity to compare to the trends in pH? I’m wondering how/if a general and downward trend of pH in the oceans may also be manifested in these salinity models to perhaps show that, overall, though there is a ramping up of the hydro-cycle such that we go from flood-to-drought more quickly… to which side does the globe itself trend toward?
The primary driver of change in ocean pH is carbon dioxide dissolving in the oceans. This is a function of atmospheric carbon dioxide concentration and a few other factors. As long as we keep burning fossil fuels at high rate, the ocean pH is going to dropping. In fact, even if we stopped using fossil fuels right now the pH would keep dropping for a very long time until a new equilibrium could be reached.
Um, that’s not Argo measurements over a period of 50 years… Argo began in 2000… it’s the more difficult comparison of recent Argo data with other data gathered by other means over the past half century.
By the way, the sky is blue not because it reflects the ocean, but because nitrogen and oxygen selectively scatter blue wavelengths. This is a common myth.
[Response: You are correct – but this isn’t what Rasmus said. He was explaining why the ocean was blue (presumably away from the English Channel seaside resorts of my youth where it was only ever a dull grey). However, it is isn’t quite as simple as simply reflecting the blue from the sky – in the ocean as well, reds and yellows are very strongly absorbed (as anyone who has gone diving notices), and so light from the ocean is strongly shifted towards blues. – gavin]
Having lived (previously) in the upper midwest for decades, and thus experienced more than one “Siberian Express”; which are what the local Weather People call the masses of super cold air that start in Siberia, move over the pole, and then come crashing down – without a tree to slow it – through the midweatern corridor, triggering snow storms as they hit the warm, moist air masses moving northward from the Gulf of Mexico; I’ve been wondering if, what with the Poles slated to become much warmer, and the Arctic Ocean possibly going to become Ice Free, what the models predict will happen in the Upper Midwest, during the Wintertime.
It would seem to me that a warmer, ice free Arctic Ocean might just ADD both warmth AND moisture to the Siberian Express, and that that might have the effect of increasing wintertime precipitation – perhaps even to the degree that it will offset at least some of the summertime drying that the models are predicting that the Upper Midwest will experience, as the Rain Belts migrate further northward.
It’s not surprising that the hydrological cycle has intensified (if what we mean by that is an increase in global mean precipitation or evaporation), however I’m very skeptical as to the alleged magnitude. It doesn’t make a whole lot of sense and the authors don’t do any justice in contrasting their results (that scale with Clausius-Clapeyron) with the vast literature on this subject.
In fact, given statements like “Given the above broad-scale model responses and the CC relationship, an intensification of ~4% in the global water cycle (E-P) is expected to already have occurred in response to the observed 0.5°C warming of Earth’s surface over the past 50 years (11)” I’m not really sure they know what they are talking about (the citation is simply to the IPCC report).
Global precipitation/evaporation emphatically does not scale with CC, as has been discussed by countless papers (e.g., Allen and Ingram, 2002; Held and Soden, 2006; Schneider et al 20010; O’Gorman et al 2011). The overall intensity is primarily constrained by the availability of energy, not moisture, and even the response is strongly forcing dependent…showing more sensitivity to shortwave perturbations at the surface than to longwave perturbations in the atmospheric column. Because of the forcing-dependence, changes in surface temperature are not sufficient to specify the equilibrium response of precipitation.
Energetically, the key is the ability of the troposphere to radiate away latent heat released from precipitation. Changes in precipitation are directly related to changes in the radiative cooling of the atmosphere (and also the surface sensible heat flux, which could adjust to accommodate a given radiative cooling, for example by change in the air-sea temperature difference). When the math is worked out you get an expected (and modeled) result in the neighborhood of ~2% per K, a rather robust result in the literature.
Precipitation changes scaling with CC would require unrealistic changes in parameters such as near surface relative humidity or air-sea temperature difference. In models, these are usually indistinguishable from zero trend. There is no discussion in the Science paper as to what other changes would produce the observed changes of this magnitude.
9 James S pondered, “perhaps even to the degree that it will offset at least some of the summertime drying that the models are predicting that the Upper Midwest will experience,”
I don’t think so. “Ice-free” generally refers to late summer, not winter, so the Siberian Express will still flow over an iced up Arctic Ocean. Then, the snow will melt sooner and faster each spring, leading to furious flooding with the moisture long gone by summer.
If the world warms so extremely that the Arctic Ocean becomes ice-free year round, then that begs the question as to where it will be cold enough for snow to last through winter. Tops of mountains, I suppose.
Nice post, and much appreciated from someone who does research on this issue. For the more technical readers, however, you may want to make it clear that the 8%/degree number refers to the amplification of the surface precip-minus-evap field, and not to the increase in global-total precipitation (which is largely orthogonal, and thought to be much lower.) Confusingly, different authors use “global hydrologic cycle” to mean either of those two things, and this has led to misunderstandings within the scientific community. Isaac Held has a good post about this at: http://www.gfdl.noaa.gov/blog/isaac-held/2011/06/29/13-the-strength-of-the-hydrological-cycle/#more-2629 See, e.g., his first sentence.
Grad student, Atmospheric Sciences, Univ. of Washington
[Response:Jack, thanks for stopping by, and thanks for those additions.–eric]
(Thanks Jack, the statements in my comment should be modified accordingly…it seems I didn’t know what I was talking about. I’ve rarely seen “strength of the hydrological cycle” expressed in this way, but my comment still stands for the precipitation field)
Do we have any data yet on the prediction of increased intensity of extreme rain and drought? I know that we should not make claims like this whenever there’s an extended drought or a major flood. It would be useful though if there is a measurable effect to have a source to refer to. One of the persistent memes in the denialosphere is the claim that climate scientists keep predicting droughts but in the area concerned there’s just been a flood (or vice-versa) and to some extent you can counter this by the “it’s weather you’re talking about not climate”.
If there is not yet a measurable trend, do any of the models indicate how long we need to wait for a detectable trend?
Do we have any data yet on the prediction of increased intensity of extreme rain and drought? …
For extreme precipitation events, I would suggest the fingerprint study:
Our results provide to our knowledge the first formal identification of a human contribution to the observed intensification of extreme precipitation. We used probability-based indices of precipitation extremes that facilitate the comparison of observations with models. Our results also show that the global climate models we used may have underestimated the observed trend, which implies that extreme precipitation events may strengthen more quickly in the future than projected and that they may have more severe impacts than estimated. There are, however, uncertainties related to observational limitations, missing or uncertain external forcings and model performance.
Seung-Ki Min, Xuebin Zhang, Francis W. Zwiers, Gabriele C. Hegerl(2011 Feb 17) Human contribution to more-intense precipitation extremes, Nature, 470(7334):378-81
As for drought, a good place to begin would be:
[from the abstract:]All the four forms of the PDSI show widespread drying over Africa, East and South Asia, and other areas from 1950 to 2008, and most of this drying is due to recent warming. The global percentage of dry areas has increased by about 1.74% (of global land area) per decade from 1950 to 2008.
Aiguo Dai (2011) Characteristics and trends in various forms of the Palmer Drought Severity Index during 1900-2008. Journal of Geophysical Research-Atmospheres, 116, D12115.
I think you’re asking whether the world warms _the_same_way_ from an increase in CO2, or from the Sun getting brighter?
Those models would be assuming the Sun stays about the same, and greenhouse gases and sulfates and whatnot change.
I think they would not see “polar amplification” if it were the Sun for example, or cooling of the stratosphere. But I don’t know what difference it would make — early on — eliminating both of those, and having instead more heat along the tropics under with the sun high overhead for 12-hour days.
In the long run, if the Sun got a bit warmer, we’d be seeing the kind of feedbacks — including in increase in greenhouse gases.
Lather rinse repeat.
Dunno if any of the models can be run with the Sun changing. Some of the astronomers who study how stars like our Sun can vary and even flare brightly might have looked at what that would do to us.
Variable-brightness star and climate change?
Global precipitation changes depend strongly on the climate change mechanism and how it perturbs the atmospheric and surface energy budgets (as opposed to TOA energy budgets for temperature, for example). Simple relationships such as global radiative forcing that work well for mean temperature change are not useful for the hydrologic response.
The precipitation response can also be decomposed into several components, depending on the nature of the forcing. For CO2 increase, precipitation actually goes down at fixed temperature. However, precipitation ultimately goes up due to temperature-dependent responses. These dependencies introduce variability in the response between shortwave perturbations felt predominately at the surface or longwave perturbations which modify the atmospheric absorption properties.
Precipitation needn’t increase for all warming agents however. Consider black carbon which absorbs shortwave radiation in the troposphere and has a positive TOA forcing, but reduces the shortwave radiation available at the surface. This atmospheric absorption can reduce the lapse rate and suppress precipitation even with a warming column, though it also depends on the vertical structure of the shortwave absorber. (see e.g., Ming et al., 2010, GRL)
For very warm climates, precipitation hits an upper bound that is determined by the incoming absorbed solar radiation (divided by the latent heat of vaporization). This is true even if temperature increases due to CO2 at fixed solar irradiance. There are some subtle effects that can modify that argument somewhat, but it’s pretty robust and the argument becomes rather simple in that limit where the atmosphere is optically thick in the longwave. One wouldn’t be able to sustain a planet for example with constant downpour (the star wars clones planet in Episode 2?) on purely energetic grounds.
I get a bit frustrated with the constant drum beat of hot’s hotter, cold’s colder, wet’s wetter, dry’s drier mantra that has been going on for the past 5 years.
Not only is it almost unfalsifiable (still waiting for the big news story on that big run of really normal weather we have been having), it is trivial to say that the climate is going to change from whatever is considered “normal” over time in different areas of the world. Cynical people see this as forward looking political opportunism for linkage to any weather related disasters. Over at the the Yale Climate Forum they even tried to tie in volcanoes and earthquakes again recently.
Rarely does it rise to the level of printable what these expected magnitude changes are going to be. The verbage leads one to imagine changes of 50% or more everywhere, not isolated. When the reality is likely changes on the order of 5% or less.
Color me unimpressed by what is best described as very detailed technical speculation with poor performing computer models. When a Russian heat wave hits and a post-facto analysis shows a computer model actually predicted it, but 35 others didn’t, what does this say? Nothing useful.
Want to impress me?
When and where are the droughts going to be? Where are the floods going to be? Is Russia going to have a huge heat wave this summer? Are the Europeans going to freeze? Make some high profile public predictions that you have confidence in and lets see what happens. The type of statements we get in articles like this have zero “actionable intelligence”.
[Response: Want to impress me? Tell me the biggest gaining stocks from now until May 2013. Or the next three winners of the Kentucky Derby. Can’t be done? Then by your logic nothing you say is worth assessing. More seriously, you are asking for the impossible (because of the stochastic nature of short term variations). Even if you assumed the models were perfect, this kind of information would not be accessible. The best that can be done for the longer term is to make predictions for the change in the statistics, and estimate impacts accordingly – if this is not actionable for you, there are plenty of resource managers and policy analysts who beg to differ. Setting impossible goals and then criticising scientists for not doing the impossible is empty rhetoric. – gavin]
Hurricane season prediction made high profile predictions over the last decade, and we now know they aren’t very good at it yet. Don’t confuse this as being bad info, or even a bad idea, as knowing the season is very unpredictable is useful in many aspects of planning.
Pretending you can predict things you can’t is not useful to the public, and can be detrimental. Regional climate prediction has been shown to be quite unreliable to date. I wonder if these type of articles shouldn’t have some sort of NOAA warning sticker “Climate Models Have Proven To Be Unreliable”.
Articles on metrics for the reliability of climate models vs. observations would be interesting. How close are their predictions vs. observations for the drier drys and wetter wets for different regions? How much divergence is there between models? Are there ANY strong areas of agreement in the models for distinct regional changes coming up?
I’d like to see a snapshot of the performance of the model science summarized here by the people who know it best. Strengths, weaknesses, areas of focus, performance vs. observations, etc. I have read almost every article on this site for the past 3 years, and I really don’t have much insight into this yet, and
Not to veer completely off topic but rather just as quick parenthetical reminder, the House of Representatives of the United States as presently constituted is absolutely determined to gut funding for research activities of the exact type discussed here, to the very greatest extent possible. In fact, dominant forces in the House wish to end the U.S. Department of Energy’s participation in climate research, notable in this case because the lead author of the very paper being discussed here is able to do this work thanks to an organization funded by DoE:
PCMDI was established in 1989 at the Lawrence Livermore National Laboratory (LLNL), located in the San Francisco Bay area, in California. Our staff includes research scientists, computer scientists, and diverse support personnel. We are primarily funded by the Regional and Global Climate Modeling (RGCM) Program and the Atmospheric System Research (ASR) Program of the Climate and Environmental Sciences Division of the U.S. Department of Energy’s Office of Science, Biological and Environmental Research (BER) program.
Those of us who prefer not to have our metaphorical eyes poked out shouldn’t forget that seemingly crazy people with sharp sticks are on the loose and in charge of House committees, shaping budgets that help determine what we’ll know or not know a few years from now.
Tom Scharf wrote: “I’d like to see a snapshot of the performance of the model science summarized here by the people who know it best. Strengths, weaknesses, areas of focus, performance vs. observations, etc. I have read almost every article on this site for the past 3 years …”
You have been reading this site for three years, and yet you managed somehow to miss the “model-data comparison” articles — in other words, exactly what you are asking for — that the hosts have posted each of the last three years?
Perhaps you might reflect on whether the rather belligerently stated erroneous comments in the rest of your post are similarly ill-informed?
Comment by SecularAnimist — 21 May 2012 @ 10:34 AM
Just two short comments; First, Chris, to your radiant pathway please also consider advection, this ties into Hanks polar amplification… Second, in relation to the mechanism of the manifestation of the thermal/hydrological correlation, I can only suggest a long review of the NCEP/NOMAD, SRRS, NH Analysis, of the 250mb isotach station, N. Jetstream flow…
Comment by l. david cooke — 21 May 2012 @ 10:43 AM
I have a bit of a naive question here. By a fingerprint, I take this indication of changes in the hydrological cycle to mean a unique if-and-only-if type connection to warming due to the increase in atmospheric carbon dioxide. How would this be contrasted to, say, changes to the water cycle due to other forcings?
It’s a response to the radiation flows in the troposphere and can include upper atmospheric temperature changes, while the surface remains unchanged, as in fixed SST experiments. Cheifly, CO2 induces a higher radiative forcing at TOA (~4 W/m2 per doubling) than it does at the surface (~1 W/m2 per doubling). This difference cts to stabilize the troposphere and suppress convection/precipitation. Because the troposphere has very little heat capacity, the differences in forcing between the surface and TOA will be eliminated on timescales of months via non-radiative adjustments, including reductions in surface evaporation and hence precipitation.
More generally, the precipitation changes can be decomposed into “fast” and “slow” responses (e.g., Andrews et al., 2010, GRL), or into a temperature dependent part (call it kδT where k is some constant) and a temperature-independent part, C. A linearized response would let us write. Lδp = kδT + C, where L is the latent heat and δp is the change in precipitation. C would include the direct influence of the radiative forcing.
A number of studies have found only weak dependence on the kδT term, while the dependence of precipitation change on the specific forcing agent seems to arise primarily from the second term on the right hand side. This introduces strong dependence on the vertical profile of the absorber and the spectral regions that are most affected by a climate change process.
I’m so sorry life isn’t as simple as you’d like it. It must be a terrible burden, being unable to accept any sort of trend analysis or probabilistic prediction. It precludes you from investing in any sort of 401K or making plans based on a weather forecast or riding in a plane or driving a car over a bridge. I suppose one answer to your dilemma would be to learn something about probabilistic evidence. Naah. That’s just crazy talk.
Gavin, the second link, 2011/01/2010, that secularanimist put up in reply 28 shows a correction dated this month on the OHC graph. Was this correction pointed out/discussed anywhere on this site? And should a correction also be made to the OHC graph in link 2012/02/2011?
[Response: Yes. I have now posted an errata. Sorry for the confusion. – gavin]
No wonder Hansen felt comfortable penning his recent NYT op-ed.
Further to Tom Scharf’s complaint, while it would be wonderful to be able to look into a numerical crystal ball and know if it’s time to begin filling sandbags or stocking up on box fans, it’s hardly necessary to have that level of granularity at hand when the statistical drumbeat of change is becoming so obvious. Composite numbers from all sorts of incidental instrumentation all point in the same direction.
Presumably drier dry regions (like the Sahara) would mean more dust put into the atmosphere. In addition to obvious reflective properties of dust, I have also seen a number of things about dust affecting plankton and thunderstorms.
I dont understand your response on #25, In order to “estimate impacts” , you would have to make a prediction first.
If you were a “resource manager” in the Alps, and you have to decide whether to invest in snowplows or snow cannons, that will last for next decade, based on your models which one would you choose ?
[Response: For a ski resort, the risks of not enough snow probably outweigh the problems of too much snow. So it is probably worth hedging against less snow if there are indications from recent trends and model projections that the chances of insufficient snow and/or warm winters are likely to increase. However I would also bear in mind that winter-to-winter temperatures and snowfall can be erratic, so I wouldn’t sell the snowplows just yet. Of course, I know nothing about the ski business, nor local projections in the Alps, but I make these points to demonstrate that actions that anyone takes are not predicated solely on what models predict, but rather they are an element in the decision making process – despite the lack of perfect predictive power. – gavin]
Comment by Tietjan berelul — 21 May 2012 @ 10:18 PM
If you were a “resource manager” in the Alps, and you have to decide whether to invest in snowplows or snow cannons, that will last for next decade, based on your models which one would you choose ?
The best part of Gavin’s response is this:
Of course, I know nothing about the ski business, nor local projections in the Alps
Not to put down the rest of his inline response, but really, to answer your question, go talk to those who are studying this precise problem. They’re the experts, and they’re not burying their heads and pretending that global warming isn’t happening.
Gavin works on NASA GISS Model E, which is focused on global climate change, and doesn’t pretend to give answers to questions such as you put forth.
Ask your questions to those who study your very specific question ….
Comment by Pete Dunkelberg — 22 May 2012 @ 6:51 AM
Tietjan Berelul and Tom Scharf,
The thing you seem to fail to comprehend is that a statistical prediction is in fact a prediction–it is merely one that requires multiple observations to answer. In a noisy system like climate, that means that it takes time. We already have 30+ years that are consistent with what the climate models predict. You would have us believe those 30+ years are a fluke, and sieze on every glitch or fluctuation as a turning point. You are concentrating on noise (weather) rather than signal (climate).
Mr. Ladbury, Dhogaza i dont want to make you believe anything. I do not have the education to argue with you. The thing you seem to fail to comprehend is that you can not insult people into compliance.
Why would your own line of thinking not apply to you; it must be easy living in a world where you are a scientist in a field that can predict everything from collapsing gingerbread houses to the rate of African circumcions. That would be just as ridculous.
To me, Gavin is ‘just’ a climate scientist, and a very good one. His answer answered my question, and I very much appreciate a scientist on his level to take the time to answer my question. I would understand if he not spent time on my post.
Comment by Tietjan Berelul — 23 May 2012 @ 6:04 PM