Specifically how reliable is the science? It seems to me the empirical evidence is nowhere near what is required for cause & effect making this a statistical study of trends and potential causes.
I would be nice to know what cars scientists or science professors use to drive at work, and what is their driving style. Do scientists that work on climate change drive other types of car than scientists from other research fields (Physics, Chemistry, Biology, Theology, Literature, .....)? Do people that are more or less exposed to potential consequences of climate change do not use the same types of car?
Obviously you never googled or else you would have found the scientific IPCC report on climate Chaneg(by the United Nations/World Meteorological Organisation). Recent and fully for free. I strongly suggest you first read a scientific report and then have your own opinion
@John, Its a fact now and we individually seeing that climate is changing. anthrogenic Carbon Emissions have surely contributed to Climate Change. May be Global Warming is also part of this Changing Climate. We are also experiencing colder winters than before and seasons have also shifted to around 15days to one month in India.
@Kenneth Towe: Sir, My research topic is not on global warming but its a common trend which as a keen observer, anyone can observe. Even the wheat sowing timing is delayed by around 15 days, due to shifting of winter season forward. And yes there is global warming too but experienced only in summers. Means summers are getting warmer and winters colder and at the same time all the seasons are shifting around 15 days forward.
My opinion is that asking around for opinions on a scientific issue is not a scientific approach. First of all you have to inform yourself as good as possible of the science behind the issue. For that the IPCC report is the most appropriate starting point by far.
As in other scientific areas models are used to interpret data and make projections; this is the way of progress. like to remind you that the 5-day weather forecasting at present is as good as that for the next day 50 years ago and the projection is that in the near future the 10-day forecast will be as certain. Thanks to models meteo/climte-science has progressed. What wouls you suggest instead?simple linear extrapolation?
In response to the attached word document and the comment that measuring CO2 emissions from the ocean, really air-sea fluxes, on a global scale is a statistical exercise I would like to point out Takahashi et al. (2009) "Climatological mean and decadal change in surface ocean pCO2, and net sea–air CO2 flux over the global oceans". This paper collates *3 million* measurements from over the global ocean to calculate the net air-sea fluxes of CO2, not quite the 'scaling factor based on a known datum point' you suggest!
Hard Science should not be based on opinions: then anything is allowed. "IPCC" does not present opinions but does a good job in gathering and critically analysing info on all aspects of climate change. The people doing this are climatologists, not politicians. The AR5 report for instance provides the best update on the changes (this is not limited to mere temperature data). When you are that critical then please advice whom of the authors and contributors of the report you do not trust as scientist.
It is not an opinion but a fact that the global temperature did not increase in the past 14 years; however it is also a fact, not an ,that you are not aware how weather or climate models work. They are not predictions but guidance; that is exactly how they are called. The Earth system, whether as old as Kelvin thought or older, has to be schematised to be described; this implies approximations and continuous improvement of this approach. Climate Change is projected because of the increase in terrestrial up-radiation and "back radiation" that increase the temperature of the atmosphere close to the surface and the sea. This increase is caused by the increase of the -anthropogenic- greenhouse gases. The models describe this increase using weather averaged over a given period.
All climatologists agree on this . See for instance the sites of a Clive Best / Roy Spencer, the better known skeptical climatologists.
The use of atmosphere/ocean models is not confined to climate. In the understanding of air pollution, ozone hole, models play the key role. A question in return to you: how could such phenomena including trends be understood without models?
Thanks Ken, that's led to an interesting half hour of reading up on the US temperature dataset, particularly given the original question of stats/trends vs. empirical data.
So you're right that the NCDC has changed the way it processes data in March, which is not the only time it's been done. What's interesting is one the features of the older dataset (the one prior to March 2014):
All temperature data before 1931were reported as state averages. Since the dataset is divided into areas smaller than states, the averages for these areas were estimated using linear regression models based on relationships between each state and associated regions in the dataset from 1931 to 1982. In the new dataset from March, this has been removed and the actual data has been used with the same methodology applied to the rest of the dataset.
This seems like an improvement to me, 30% of the timeseries that was based on statistical interpolation is now based on real data treated in a consistent way to the rest of the dataset! To be honest I was quite surprised these changes haven't been made long ago given that these climatological approaches have been used for the global temperature data for a while now.
Just for clarity...
the old dataset description: http://journals.ametsoc.org/doi/pdf/10.1175/1520-0477(1996)077%3C0293%3AAHPOUC%3E2.0.CO%3B2
the new dataset description: ftp://ftp.ncdc.noaa.gov/pub/data/cmb/GrDD-Transition.pdf
Hi Stuart, your question is really interesting. They can be analyzed from a scientific paradigm, political, sociological and many more. Some time ago I began to wonder about the real contribution of humans on global warming, and would like to contribute to the discussion of a Danish scientific studies (Henry Svenmark) that attracted my attention.
http://en.wikipedia.org/wiki/Henrik_Svensmark
(Citations to his published articles are at the end of the page)
I appreciate that, it's important that there is a cross-check against the older data. I guess one issue is that there are no error bars on any of these stats, so it's difficult to know, for example, how representative the mean from the US Weather Bureau data is and respectively for the NCDC data. The NCDC would do better to have some error bars on their data.
Taking your figure from above, I was interested to see what the two time-series would look like given the differences between the datasets you show (Assuming I've understood correctly, your graph shows the difference between the annual mean US temperatures from the new NCDC temperature dataset and its predecessor? ...if not I can redraw!). What strikes me is that the two aren't really that different overall and it doesn't make much difference to the overall trend. There might be bigger differences when looking at specific parts of the time-series and when breaking it down regionally, but for the purpose of annual average US temperature change it appears fairly robust.
(The NCDC temperature is the annual mean US temperature from www.ncdc.noaa.gov/ and the corrected is that plus the data in Ken's graph above digitised using http://arohatgi.info/WebPlotDigitizer/index.html )
It is quite illustrating what you have found out about the US data and it certainly gets me worried. As I mentioned in other questions in this topic. Climate Change is about precision not accuracy, meaning that the absolute values matter less than the precision in the change of the respective parameter, be it temperature or precipitation
Question to: where in ResearchGate did you show the failure of the models on the Ozone hole?
Increasing the amount of carbon dioxide and methane implies increase in temperature. This has been known now for over a century. This can thus not be a matter of opinion but of scientific fact.
See for instance Wikipedia: "Arrhenius". Also there is the site of CliveBest, who is skeptic about the value of the temperature increase by the increased CO2. HOWEVER, this skepticism is not about the direct warming by the CO2 but the warming by the increased water vapour.
Also, I notice a complete absence of you in the discussion and the unscientific remark "it seems to me.....". In science you have to back up personal remarks with facts/data/references. So I ask you to ad those to your question in a revised question.
@Ken You wrote "The early changes are consistently lowered, which seems to be the opposite of what one might expect if the role of urban heat from a growing population is to be taken into account." So what are we looking at here? Data or conjecture? We expect to see a heat island effect, so we discount the existing data because it is not there? You have (correctly) criticized others for this flaw of logic in other posts.
I understand that it is interesting (even important) to spend one's academic career looking at fine detail, but at the moment it is critical that we look at the big picture of how we use science to guide government policy. And the big picture in this discussion remains unchanged. I challenge you (again) on the "fact" that there is a hiatus in warming. On a decadal scale (the only reasonable scale to use) warming clearly continues with each decade consistently warmer than the previous. If we focus alone on the temperature of the atmosphere over periods of less than a decade, we miss completely the impact of complex interactions that occur between the oceans/atmosphere over this timescale.
As I have written to you many times before, (so my comment is for the others who read here) we must look at what is happening to the total heat/energy content of the climate system (oceans and atmosphere) to consider how climate will change in the future. There is strong evidence to suggest that we still have an overall energy imbalance due to GHG emissions. Where is that energy going if not into the oceans? If you want to suggest it is reflected by clouds (the only reasonable suggestion) then please show me the evidence.
Re the IPCC.... I understand that the US Supreme Court has decided that cooperations are "people", but the IPCC is not such an entity. It is by regulation an organization that only collects and summarizes climate research. But I can see that it would certainly make your overall argument easier if we could all discount- by magical wave of hand- the data from NOAA/NASA/IPCC/NSIDC/ and most of the Universities in the country... and overseas. They are, after all, only climate "watermelons". Green on the outside but deep RED in the middle :-)
Oops the perils of late night programming, I didn't notice the sign on Ken's graph! I've attached a corrected version under the original, which shows the new gridded temperature as lower than the older one! My point that small differences in the averaging algorithm don't result in a different interpretation of the large scale trend still stands.
To clarify Ken, you posted a graph of the changes in the US annual temperature record resulting from the changes the NCDC have recently made. The old record is no longer available, but given the new record available on NCDC's website and your plot, I combined the two (i.e. new - old = difference old = new - difference).
You can still find snapshots of the old record on old webpages via google like here: http://www.noaanews.noaa.gov/stories2009/20090108_decemberstats.html , it seems to tally up pretty closely with what I did.
Remains my request to you how to project changes in atmospheric parameters, when the key constituents are known to appreciably change, without the recourse of models?
Yes, amounts can be easily projected, but my question was, in response to your negative appreciation of climate models, to project the resulting effect(s) on climate parameters. For this I refer you to the site of CliveBest, a skepticist, explaining with the simplest possible model an increase in temperature of some 1.3 C per doubling of "CO2". http://clivebest.com/blog/. He thinks that the so-called water-feedback, or rather the water-accelerator, does not exist
Also, you can find a more elaborate discussion at Scienceofdoom, including a discussion on water and cloud.
What is the origin of this graph? Which SRES scenarios where chosen to illustrate this? This is critical information to make an informed judgement, but it is not at all clear from the graph. Perhaps a more comprehensive analysis could help?
I have reviewed Va warming record. I agree that is certainly not warming as fast as some other states, but there is a recent trend of changing climate... one that is intense enough to concern State horticulturalists. it depends on where you set the base line, but I agree that it is clearly not the State that will suffer most over the next couple of decades...
But then we jump from this reasonable argument to "destroying world economies" and the old and long discredited idea that "all scientists were promoting cooling in the 1970s." A good source to impale that particular zombie can be found at http://www.skepticalscience.com/ice-age-predictions-in-1970s-intermediate.htm
Thanks re the graph. The trends observed still conform to most projections from the IPCC but do (I agree) track to the lower end over the last 10 years. No argument. I also agree that there was mixed opinion about warming/cooling in the 70s. As geologists we understand why... there was good reason to consider the possibility, but in researching my book, I discovered that by that time the UG Govt. was already taking the threat of warming from CO2 seriously. I have no reason either to doubt your data on record temperatures in the USA. Please understand that I do not mean to be rude when I say that I am not interested. It is, of course, in a general sense interesting to discuss, but when we observe the actual (tangible) impact of climate change in so many places (whatever the cause) and then consider the cost to the nation ranging from disaster relief/recovery to the price of food, the argument feels too academic to me? I must say I am relieved to have landed (once more) in a place where we generally agree. If I sounded disrespectful at times, it was not my intention. I am passionate about my work because I think this discussion really matters. I look forward to more discussions...
There are models and models. Feynman's models are easy, they describe very localized phenomena mostly in the realm of the lab with very defined EXPERIMENTS.
A climate model is not of that kind (rain-making the exception?!). Again, climate models describe changes not the absolute values. This is another difference with a "Feynman" model.
@Ken... I agree that any one instance of climate extreme we observe today is not that different from others we have observed in any one location in the past. Some do appear to be record breaking, but that does not prove much statistically at the moment. The main point for me here is that these events are now occurring in many different parts of the USA at the same time- and most importantly, all over the world in different places at the same time. In a random world of climate, taken globally, one might expect as many cold records to be broken as warm records, but that is not the case. Then add precipitation, evaporation (wildfires/drought) and sea level/temperature (for a partial list). This soon becomes statistically significant and points to climate change and globally... and thus to the search for a reason.
I am confused by your point about temperature change. Don't we use anomalies rather than exact temperature to define this change, exactly because we acknowledge uncertainty in the accuracy of the data that are nevertheless (in most cases) consistently precise?
Pick the one you prefer to use to illustrate your point and I am happy to engage in a data discussion. I see no evidence of NH cooling over 35 years.
BTW since you made me curious to look more at Richmond in general, I think Richard Muller's graph is interesting as is shows the onset of post LIA warming a little earlier than in many other places. Interesting...
As this is likely to take some back and forth, I am happy (if it works for you) to do this more directly? [email protected] if not, I am happy to keep posting here.
Climate models by definition do not operate at a year to year basis but for averages over a period long enough to smooth out year to year variations, for instance caused by known spikes like large volcano eruptions and el Nina / Nino's.
The confidence in model descriptions of changes for a short period are not warranted.
I disagree with you on absolute temperatures because climate change is concerned with changes / anomalies. Model (projections) are based on historic trends.
My answer was in reference to models. Those cannot work with absolute values; even weather prediction models start with the situation as is and then project the change.
Thus as long as data are consistent and precision is warranted absolute values are less of a concern for analyzing changes in the past and projections of future changes. These are also driven by first principle physical "laws", like the very accurately known absorption of radiation of greenhouse gases.
The models do not care about the absolute values, just the change per, at least, 30 yr period, and that in relation to the increase in "CO2". This increase is only significant, in terms of to be expected zero-order temp change over the last 60 years. Any change before 1960 is mostly natural. Cooling may have been induced by aerosol in the fifties/sixties.
The only answer I can give is that for climatological differences one needs a period in which the year-to-year variations are smoothed out and spikes, like volcano eruptions and el nino's, are duly accounted for. As I mentioned before a human signal can only significant since the time substantial human greenhouse components (also aerosols) were emitted and/or accumulated in the atmosphere.
did you check whether 1938 was an el-nino year like 1998?
as for anomalies: in climate it is the anomaly of a period not of the annual temperature.
Studied, again chapter 10 of AR5 and see that yous grapoh does not stand out as different. However, the chapter is called attribution of the change: here volcano eruptions and el-nino etc are taken into account to assess the trend. This has to be done because temperature data as such have no meaning without "correcting" for natural spikes.
This does match your graph quite well up to 2000, but of course extends beyond that for a further 15 years. Change the data to annual average and to 1970 and it suports what your graph shows? (attachment)
These might help clarify our discussion. I can't remember a time before when I referenced Antohny Watts site... there is always a first.
reviews these early temperature estimates. One must ask what data we had to estimate global temperature at that time? And when we compare the data from then with today, how exactly are we comparimg apples with apples? I believe the figure or around 15 degrees works for today, but back then? So much data was missing from the anaysis. I really do not see what advantage we gain from focusing on an accurate absolute temperature at one time. We need to set a base temperature based on data that has reliable precision and then work with the anomalies. The process of estimating global temperature has been iterative, working with new data and upgrading old data as we understand more about underlying errors and oversights. I agree that there is a theoretial risk of data manipulation to try and get the "desired" answer, and you do find flawed scientists who might do that... but with so many to check and balance? And you have to include outright skeptics like Richard Muller. Just for fun I am going to try to send you the chapter from my book that deals with global temperature (may be too big).
I am surprised to read that Jones mentions anthropogenic as one of the causes of the change as early as the 1930's. At that time accumulated manmade CO2 was negligible?
Ken, you have brought up some interesting facts and data and I will be sure to go over it in more detail. I've known of a few times where I myself have been annoyed at a lack of consistency through papers and references due to date tables being parsed and sampled rather than looking at a complete trend.
Harry, I understand that I have not created a correctly phrased Thesis or Research publication question outlining the facts and aims of my inquiry, if I had I may have written a paper on it instead. What I meant by this question was a thought experiment highlighting the flaws (some of which Ken has touched on) of attributing grandiose climate change to Industrial Mankind.
Sorry to everyone about not following my own question and the responses I've been working and moved 4 times since I finished university so had forgotten until I was running round my online presence.
The main crux of my opinion on the topic is that a lot of the "Research" I see has anomalies like the ones Ken has pointed out. The Earth is a hugely complex fluid system and the idea that we have so vastly changed the temperature of the world seems in congruous to me. Moreover the timescales for measurements have "World Atmosphere Temperature" measurements as far back as 1880 in some cases which cannot possibly equate to the standards or accuracy held today. Moreover as a Physicist I know that there are almost innumerable ways to measure ambient, absolute and relative temperatures of fluid and solid state matter, even 2 measurements made by the scientific community are unable to be equated if not the same methodology.
On top of that you have the inherent inaccuracy of inferring micro-scale chemical interactions with macroscopic scale fluid dynamics. Furthermore cosmic ray nucleon bombardment is a high energy consistent event in the exosphere and high energy electron bombardment in the ionosphere. These events are difficult to quantify or analyse even now considering the shifting and in-homogeneous nature of the Physical processes at our outer atmosphere the exchanges within the carbon cycle due to those processes as far as I know are not measured at all and I've not been able to find publications attempting.
All-in-all I am of the opinion (by which I mean I refer to known "facts" and logical premises from my own and others' experience to draw conclusion) that the idea that we can comment on the state of the Climate as a process we critically influence is dubious. Doubtless we have influence but I believe that the renewable and future technologies will become perfected in the next 200 years or so, we've only had over 200 years since the apparent source of our "Climate Changing" Industrial Revolution and gram for gram these technologies using renewable energy and recaptured EM spectrum etc will have no chemical effect on climate once made and if made from plastic will in fact re-absorb and lock in some of the available carbon via acid processed in their formation.
Eventually economy may become "Cloud" based in that as little physical media (books, cd, dvd etc) as possible will exist. By extension company models may shift towards the permanent renting of devices and objects that once made can power themselves via high capacitance solar power. If Solar, Wind, Tidal & Geothermal power were only providers (such as is already so in Iceland, Costa Rica etc) then all processes that produce a renewable product will inevitably pay off it's Carbon footprint.
THe basic climatology science is rock hard and supported by every denier also: more CO2 means higher temeperatures
With this I mean go for instance to the site of the "denier" Clive Best. He shows with simple math how the relation is between CO2 and warming
The discussion is on the feed back that is what happens with the amount of water vapour in the atmosphere (a very strong greenhouse gas) in a warmer climate and the increase in clouds
I refer to the basic physics known since Arrhenius early 20th century: for the temperature as a whole there is a log relation between CO2 and temperature at the surface
All meteorologists and climatolosts agree upon this. Then comes the extra: water in a warming climate more water vapour evaporation form the oceans
I was not referrring to RealClimate for its own sake
It is Marcott explaining in more detail his Nature publication for FREE; so everyone can read that he does not say that the climate is warming everywhere, as you claim. He states that one needs a large data set to see the warming.Talking about reliability.
PS the Original Nature publication is behind a pay-wall and thus not accessible for everyone to study
I reacted to your remark that the analysed warming was evenly spread over the globe. Marcott does NOT say that at all in his contribution to RealClimate. Also the thrust of the Marcott paper is the milennia before the previous century
I have mentioned before that I am not a data specilaist like you are.
In addition I understand limate change not the absolute values as valid today. Models do not calculate these they can project changes in hindsight and future with base case parameters:temps precip and some cloud cover CHANGES over time. As you rightly indicate climate is very complex and a model is always a simplification. But this is the case in all branches of science. Models are there for guidance and insight in the reason of changes. However one does not need complex models to know that extra CO2 means warming of the lowest level of the atmosphere/troposphere. Here the basics are spectroscopy and thermodynamics
You say the models have failed: then qunatify this failure. Do you mean US temps?
Again climate models are for a climate period not for a period of mere years. So the only scientifc change is that for a change of a climate period of decennia to another period of decennia; and then periods for whihc there is significant CO2 and CO2 increase. The latter is only the case for the last 60 years thus 2 climate periods
Climate models are simialr to weather models.
Weather models do not PREDICT the weather. They are a guidance and a very necessary one: a preview of coming storms, flooding and other disasters. But also for planning trips in the near future we make use of the model projections. Also in case of weather models the likelihood of the projections are given; in case of the European Model a set of 60 slightly different runs are made. The differences are generated in the starting conditions. This provides an inherent probability scala. This is shown in the KNMI so-called spaghetti-plots of temperature and precipitation
Indeed IPCC uses the word PREDICTION and that cannot be done but in the most simple systems.
Having data-sets without an idea where temperature changews can come from is in any case NO science
Here meteorology and climatology come in and these only can be understood with the help of models. Projections are only possible with an idea of what extra greenhouse gases CAN do and paleoclimate analysis als needs model guidance.
Or do you think one can do with a mere (linear) extrapolation of data or for that matter current weather for the next days?
Long time no writing... but lets highlight this difference between predicting what will happen and projecting what might happen given a future scenario. IPCC climate projections (and lets go to the most resent report) depend not only on the vagaries of the climate system, but perhaps even more so on the uncertainties of economic development, social cohesiveness and technology. Given these uncertainties, all that the IPCC projections can reveal is the probability of a future climate scenario- and on a "business as usual" model they project a world where there will be significant climate change with an overall negative impact on society and the environment. I see no significant departure or observation from projection, especially as we head for another hottest year on record with the additional impact of El Niño and +ve PDO ( not or weakly present last year). Botton line Ken: I agree that all models have inherent weaknesses born of simplification and incomplete understanding. I nevertheless consider the IPCC projections more likely to be correct than those who project the future on the basis of limited data to suggest that climate sensitivity will be at the lower end of the scale.
The questioner himself restarted the discussion 9 days ago. He doubted the science but did not provide any references for his doubt except for his distrust in thermometers and an entry on bombardment. Both of these items he can find answered in the scientific part of the IPCC-report. There are also more general answers that can be given:
1. For a change in temeprature the thermometers do not have to be accurate just precise meaning they should be able to indicate the change
2. The increase in accumulated CO2 is only significant since 1950/1960, not sicne the start of the Industrial revolution
3. The galactic influence: in the IPCC report: chapter 7
7.4.6 Impact of Cosmic Rays on Aerosols and Clouds ......................................................................... 45
You describe the accuracy of the data; so/but what does it mean for the precision of the trend over the decades as this is what climate chnage is about
Furthermore I'd verry much like to see that the questioner responds to my remarks on his new statements, especially whether he looked at the references given: this in reply to your remark what the status of this discussion was
I have studied your convincing graphs and it is what it is. Howevr, climate change is about global trends becasue regional trneds are very much governed by changes in "weather patterns / synoptic situations". That is also why models cannot (yet) describe regional trends.
My question thus comes down to how reliable in the sense of precision are the rest of the temp data outside the US, say for instance in Canada presenting quite an airmass in comparison
I am puzzled but then I am not so accustomed to your situation. Reminds me of the ttime I worked in air pollution over the NE USA and made aircraft meausurements. My perspective of the vertical layering was too much biased by my European knowledge.Similarly I refused to evaluate our more recent measurements in Atlanta because I am not familiar with the local situation
To make your own judgement on specific data it is nice to be accustomed to the local situation. For instance the local adjustments for temp in the Netherlands and also that of solar irradiation I can rather well follow. I kno the scientists who work(ed) on these items and their possible biases and views.
These I miss re US data in all aspects also those of air pollution / aerosols, even though I am more familiar with the US research than most of my colleagues in Europe