Torqumada said:
You're right statistics doesn't require an exhaustive sampling of each possible data point to be complete. However, you must have enough data points to be statisically significant. 1000 data points in 4.55 billion is not statistically significant.
No, you are misunderstanding how "statistical significance" works. Only the absolute number of data points determines the statistical significance, the fraction (number of data points)/(number of all possible data points) is completely irrelevant. I already gave some examples showing why your interpretation makes no sense in my last post:
Jesse said:
Radioactive decay has been going on since the earth was formed, yet we have only begun to measure it in the last century; by your logic, physicists therefore have no justification for saying anything about the laws governing radioactive decay. Likewise, people have been smoking for thouasands of years, but we only have data on the relationship between smoking and lung cancer for the last century or so, yet this doesn't stop us from coming to some strong conclusions on the causal relationship between the two.
Likewise, to deal with your own example:
Torqumada said:
Its like polling one person in the United States, lets say Mike Collins, and asking them who will you vote for in the next election and then reporting the Bush will win by an unprecented land slide of 100% of the electoral votes (the ratio is roughly the same).
The reason that would be a bad poll is because the absolute number of people polled is too low. But again, the fraction (number of people polled)/(total number of voters) is irrelevant to the statistical significance. If I polled 100,000 people about the upcoming election, the statistical significance of this result would be the same regardless if I wanted to predict an election for governor of Kentucky, with about 2.5 million registered voters, or an election for Supreme Emperor of the Andromeda Galaxy, with 70 trillion registered voters.
Torqumada said:
In that case, the period of time is the deciding factor since you are measuring an event over a period of time. If your period of time is too short, then your data points won't be significantly accurate. In regards to smoking, you don't have to worry about time, since millions of people die or are effected by lung cancer from smoking every year. Those are your data points in that case, the number of poeple with lung cancer vs the number of people with lung cancer that smoke vs the number of people with lung cancer that don't smoke. As an example, if you have 1 million people that are getting lung cancer every year, and say 70% of them smoke and 30% that don't and you poll that for 10 years, that is 10 million data points. That is enough data points to say that something is going on.
I agree that time has to be considered in some cases, although not for the reason you give. Time is a factor when you believe the relationship between different variables is likely to be
changing over time. The relationship between lung cancer and tobacco smoking is a matter of biology, and probably has not changed much over the centuries, so it's not really necessary to have data points from all centuries. On the other hand, if you wanted to study the relationship between nutrition and growth rates, if all your data points came from three-year-olds it would not be legitimate to assume the relationships found would hold for people of all ages (for example, adults basically won't grow or shrink at all, regardless of how they eat).
So, it might not be legitimate to say the relationship between carbon dioxide and temperature observed for the last 400,000 years is certain to work the same way 10 million years from now. But it seems to me that generalizing results found over a certain time period to the
same time period is perfectly legitimate. For example, suppose I study the relationship between exercise and muscle growth for a group of men for a period of 4000 days, or about 11 years. Suppose all these men all have the same birthday, were all aged 25 at the start of the study, and the study ended right on their 36th birthday. Since I only studied a particular age group, it would clearly not be legitimate to generalize whatever results I found to 10-year-olds, or to 80-year-olds. However, I think it would be perfectly legitimate to generalize the result to men aged 36-and-one-day. Although it's possible men go through growth changes which alter the relationship between exercise and muscle growth overnight, if I didn't see any such changes in the dates of my study and I have no prior reason to think there's anything special about the dates I picked, it would be unreasonable to estimate the probability of such a change happening the day after the study ended as much higher than 1/4000. Similarly, although it might be possible to imagine the earth going through sudden phase transitions which alter the relationship between carbon dioxide and temperature, if you assume such phase transitions have nothing to do with human activities, the chances of one
coincidentally happening in the same century as our massive industrial buildup and altering the regular relationship seen in the last 4200 centuries should not be much greater than 1/4200. Of course one could argue that such a change might happen for non-coincidental reasons, ie that human activities would actually cause such a transition, but then we'd be back to the idea of humans causing serious changes to the climate, even if not in the way that climate scientists usually imagine.
In reality it's not actually true that climate scientists base their belief in the relationship between carbon dioxide and temperature on nothing but statistical analyses--as I understand it there are a lot of other reasons involving things like lab experiments on the properties of carbon dioxide, observations of the atmosphere of Venus which is rich in greenhouse gases (Venus has the hottest surface of any planet, despite being further from the Sun than Mercury), computer models, theoretical analyses, etc. But my point is that even if we rely on nothing but statistical relationships, your argument that a sample of data points from the last 420,000 years is not enough to draw any conclusions doesn't really make any sense.
Jesse said:
As I said, I don't know, but the main point is that I think we can be pretty certain that scientists have some arguments as to why this is compatible with the CFC theory (perhaps there's a time lag in terms of how long it takes for new ozone to be produced even after CFC levels are dropped), so any credible critique of the theory must at least take these arguments into account and show why they fall short. An uninformed critique that can't even anticipate how defenders of the theory would respond is of little value; maybe it'd be hard to explain the failure of the hole to shrink and maybe it'd be easy (just like you now apparently say it is easy to explain annual changes due to ambient weather patterns, although you presented this as a serious problem earlier), I don't know and neither do you.
Torqumada said:
Did you read the website I posted? That is where the information came from on how the "local" (local in quotes because we are talking about a weather pattern over the entire Antarctic) weather effects the ability of the Ozone hole to form. Did you think I just made this up of the top of my head? As i said, I was wrong on that theory.
I think you're missing my point, I never accused you of making it up off the top of your head, in fact that would go against my whole argument. My argument, again, is that it's a waste of time when people try to discredit established scientific theories with simple one-line arguments without even bothering to do any research into how scientists would respond to these arguments, since it is extremely unlikely that the scientists will have just completely failed to have noticed a devastating critique of their theory. You proved my point when you first made the simple argument about the size of the hole changing over the year, then actually did some
research and found that this observation is easy to explain, and does not actually discredit the CFC/ozone hole theory at all. What I'm saying is that the same is likely true of your argument "the hole hasn't shrunk even though CFC levels have gone down, therefore the hole cannot be caused by CFCs"--if you actually did some research on how scientists explain the failure of the hole to immediately shrink, I'd be willing to bet a large sum that they aren't completely stumped by this observation. I speculated on one possibility in my last post when I said "perhaps there's a time lag in terms of how long it takes for new ozone to be produced even after CFC levels are dropped". Another possibility might be that although CFC levels are dropping, they are dropping so slowly that we wouldn't necessarily expect to see a significant effect on the size of the hole--this might be suggested by a quote on p. 237 of
The Earth's Biosphere, where the author writes:
Although the long atmospheric lifetime of CFCs means that the stratospheric effect will be felt for decades to come, atmospheric concentration of these compounds have been falling since 1994 (Hall et al. 2001), and the stratosphere may return to its pre-CFC composition before 2050.
And I also found
this article which says that although the hole's size did drop to its lowest size in a decade in 2002, the size increased again last year, and that the size in any given year depends a lot on meteorological conditions that year:
A single molecule of chlorine can degrade more than 100,000 molecules of ozone. And there are large amounts of them in the air – largely stored within intermediate and inactive 'reservoir' compounds, but activated by particular meteorological conditions, notably those found above the South Pole in winter.
...
Small ozone holes have also been known to form over the Arctic, but meteorological conditions there prevent the stratospheric temperature dropping enough to cause thinning on an Antarctic scale. The small size of the 2002 Antarctic ozone hole was due to the same reason – unusual wind patterns and warm weather prevented more ice cloud formation. But this year’s measurements show there is little room for complacency on this issue.
“Since the end of the 1990s, there is experimental evidence that the total chlorine loading in the stratosphere is decreasing, but this is a slow process,†observed Dominique Fonteyn of BIRA-IASB. “Superimposed on this is the meteorological variability which this year still allows sufficient chlorine activation over a large area and for a longer period than last year.â€
(emphasis mine)
So again, I don't know enough about this issue to give much of an answer, but these quotes suggest to me that scientists do have some understanding of why the hole has a given size in any given year, and that it was never expected that curtailing CFC emissions would cause an immediate dramatic reduction in its size.
Jesse said:
Who were these "more rabid members", exactly? Were real scientists claiming with a high degree of confidence that it was definitely an effect of global warming, or are you just talking about environmental activists? As you said, it was real scientists who also happened to be global warming advocates who discovered the volcano in this case.
Torqumada said:
I remember quite clearly on several television and radio news programs that Scientists (or at least people with PhD after their name) were talking about the collapse of the Ice shelf as a result of Global warming. I am sorry that I didn't record those for posterity and get their names. I can't be sure the following people were not one of those experts giving the interviews on said programs, but they were advocating Global warming as the culprit: Tim Naish of the Institute of Geological and Nuclear sciences, Ted Scambos, of the National Snow and Ice Data Center at the University of Colorado, Christina Hulbe of Portland State University, Mark Fahnestock (University of Maryland, College Park, MD) are some of the ones that have reported ont he event.
Fair enough. I'd say that what's important is not who exactly said it, but what degree of confidence they put in the idea that it was due to global warming, and how wide the consensus was that this was definitely the explanation. There's nothing wrong with putting forth theories that turn out to be falsified by later observation, that's just how science works after all. And as theories go, surely it is fairly reasonable to suggest that this was an effect of global warming--you yourself agree that the earth has gotten significantly warmer in this century, even though you disagree about the explanation.
Jesse said:
The graph shows that for the last 420,000 years CO2 levels never got higher than about 300 ppm, which I would not call "almost as high" as 365 ppm--that's about 20% higher than the highest levels its ever been over the last half a million years or so, and from the looks of things about 50% higher than the average levels. What's more, just based on the known amount of CO2 released by human activities, I don't think it's disputed even by most global warming skeptics that the main cause of this sudden upswing in CO2 is human-caused (if nothing else, it would be a pretty huge coincidence that this upswing coincided so precisely with industrialization, if it was a natural phenomenon that was just as likely to occur in any of the 420 centuries covered by the graph). Are you indeed arguing that there's a significant chance the change in carbon dioxide is primarily due to some other cause?
Torqumada said:
The picture is a bit blurry for me, maybe the resolution dropped. I went looking for the original and cannot find it at the site you indicated. Which figure is it please? Some of the data points appear to go above 300 ppm on the CO2, but I want to be sure on that.
Are you sure you were looking at the right site? The one I posted was
http://www.daviesand.com/Choices/Precautionary_Planning/New_Data/ and it's the only graph that appears in that article. Meanwhile, the book
The Earth's Biosphere which I referred to earlier gives some exact numbers:
As noted in chapter 4, only approximate reconstructions of CO2 levels are possible for more distant periods, but air bubbles from Antarctic and Greenland ice cores (the largest one is 3,623 m deep) make it possible to trace CO2 levels with high accuracy for the past 420,000 years (Petit et al. 1999). During that time CO2 levels have stayed between 180 and 300 ppm, and during the 5,000 years preceding 1850, they fluctuated only between 250-290 ppm (Petit et al. 1999; fig 4.10). Their post-1850 rise will be described in chapter 9.
(p. 135)
Then in chapter 9 he discusses measurements of current CO2 levels (measurements made far away from civilization):
The first systematic measurements of background CO2 levels (i.e., far away from major anthropogenic sources of the gas as well as from extensively vegetated areas) began in 1958 at Mauna Loa and at the South Pole, with stations in north Alaska and American Samoa added later (Keeling 1998). Mauna Loa's 1958 average CO2 was 320 ppm, whereas in 2000 the mean observation surpassed 370 ppm (fig. 5.8). This increase of more than 30% in 150 years is of concern because CO2 is a major greenhouse gas whose main absorption band coincides with the Earth's peak thermal emission (see chapter 4).
(p. 234)
So, that figure of 370 ppm in 2000 would be about 28% larger than it's been at any time in the 5000 years before 1850, and about 23% larger than it's been at any time before 1850 in the last 420,000 years.
Torqumada said:
As for other sources of CO2 on this planet and the reason for its rise, I can think of a couple of natural sources, that provide the majority of the CO2 in the atmosphere anyway.
1) Volcanic gases. There have been several large volcanic releases and many small low key eruptions (more like leaks in terms of volcanoes I guess) in the last 200 years or so: St Helens, Pinatubo, Krakatoa, Vesuvius, the one in Iceland I can't remember, downtown LA (if you can believe Hollywood.
) as well as others like those in Hawaii that are low key. I am not sure if there has been a relative upswing in activity to account for this but its a possibility.
2)Earthquales also can release lots of gases trapped in the Earth crust. If these events take place in an area that can't be easily observed, like the ocean floor, and CO2 can be relased without us seeing it.
3) Inhibtion of the ocean from absorping CO2 like its supposed to. Maybe there is a threshhold to how much the ocean can absorp. Once it reaches that limit CO2 begins to build up in the atmosphere, until the temperature rises, causing an explosion in the growth of plants, both terrestrial and aquatic that can once again absorp the excess CO2 and start the cycle over again. Maybe there are cycles to the algal and protozoan populations that last 125,000 years or so to account for this.
4) Everyone breathing my air! (thats a joke. this is a firendly debate after all.
)
5) Earth passing through an interstellar cloud of gases, composed primarily of CO2 and metthane, causing it to seed the atmosphere with excess that the eco-system then has to deal with. (Least likely, but plausible)
6) This one might be a bit circular one or at least a positive reinforcement system. If the amount of solar radiation has increased, causing temps to increase, then the melting ice of the Arctic and Antarctic of the Earth, could release trapped CO2 during the melting process, increasing the amount of CO2 in the atmosphere.
Well again, your arguments seem to be based on virtually no research into what led scientists to conclude the increase in CO2 was primarily due to human sources--you seem to be under the impression that scientists are pretty much in the dark about the contribution from other sources like volcanic gases, but from what basic research I've done that doesn't seem to be the case. For example, according to
this wikipedia article on CO2:
Volcanic activity now releases about 130-230 million metric tons (145-255 million short tons) of carbon dioxide each year. Volcanic releases are about 1% the amount which is released by human activities.
Since the annual gain in carbon in the atmosphere from increased CO2 concentration is measured to be about 3.2
billion tons (see
this page), scientist's estimates of the contribution from volcanic activity would have to be off by more than a thousand in order for it volcanic activity to play any significant role in the increase in CO2.
The Earth's Biosphere also has a detailed section on the
carbon cycle in which the author mentions that it is "the best studies of all elemental cycles". This lists all the various carbon sources and sinks that contribute to the annual flux of carbon in the atmosphere, like 100 Mt of carbon being buried each year by plankton sinking to the bottom of the ocean; it seems as though all the various elements of the carbon cycle are fairly well-measured. The
page I linked to earlier also gives figures for the various sources and sinks of carbon, summed up in this diagram:
(this is obviously simplified, since many of these categories can be broken up into further subcategories)
I can't vouch for the accuracy of this page, but very similar numbers are also given on
this page. Also, I don't know what the error bars on the various numbers given are, but they're good enough that when you sum the various fluxes from the atmosphere into carbon sinks with the fluxes from carbon sources into the atmosphere, you get 107 Gt - 102 Gt = 5 Gt, only 1.8 Gt off from the measured flux of 3.2 Gt. It's possible that this is just luck, but again, I think it's foolish to argue the estimates may be badly off without even bothering to learn how scientists arrive at these numbers in the first place.
This page gives some other reasons why it's thought to be fairly certain that the main source of increased CO2 is human activities:
http://www.radix.net/~bobg/faqs/scq.CO2rise.html
For one thing, they point out that the total mass of the extra CO2 in the atmosphere each year is actually
less then the mass of CO2 emitted by human activities:
The rise of airborne CO2 falls short of the human-made CO2 emissions.
** Taken together, the ocean and the terrestrial vegetation and soils
** must currently be a net sink of CO2 rather than a source [Melillo,
** p 454] [Schimel 94, p 47, 55] [Schimel 95, p 79] [Siegenthaler].
The geographical distribution of increased CO2 also matches human activities:
Most "new" CO2 comes from the Northern Hemisphere.* Measurements
** in Antarctica show that Southern Hemisphere CO2 level lags behind
** by 1 to 2 years, which reflects the interhemispheric mixing time.
** The ppmv-amount of the lag at a given time has increased according
** to increasing anthropogenic CO2 emissions. [Schimel 94, p 43]
** [Siegenthaler]
Finally, the ratio of carbon isotopes in the air matches that of human emissions, but not that of CO2 emitted by the biosphere or the ocean:
Fossil fuels contain practically no carbon 14 (14C) and less carbon
** 13 (13C) than air.* CO2 coming from fossil fuels should show up in
** the trends of 13C and 14C.* Indeed, the observed isotopic trends
** fit CO2 emissions from fossil fuels.* The trends are not compatible
** with a dominant CO2 source in the terrestrial biosphere or in the
** ocean.* If you shun details, please skip the next two paragraphs.
** The unstable carbon isotope 14C or radiocarbon makes up for roughly
** 1 in 10**12 carbon atoms in earth's atmosphere.* 14C has a half-life
** of about 5700 years. The stock is replenished in the upper atmosphere
** by a nuclear reaction involving cosmic rays and 14N* [Butcher,
** p 240-241].* Fossil fuels contain no 14C, as it decayed long ago.
** Burning fossil fuels should lower the atmospheric 14C fraction (the
** `Suess effect').* Indeed, atmospheric 14C, measured on tree rings,
** dropped by 2 to 2.5 % from about 1850 to 1954, when nuclear bomb
** tests started to inject 14C into the atmosphere [Butcher, p 256-257]
** [Schimel 95, p 82].* This 14C decline cannot be explained by a CO2
** source in the terrestrial vegetation or soils.
** The stable isotope 13C amounts to a bit over 1 % of earth's carbon,
** almost 99 % is ordinary 12C [Butcher, p 240].* Fossil fuels contain
** less 13C than air, because plants, which once produced the precursors
** of the fossilized organic carbon compounds, prefer 12C over 13C in
** photosynthesis (rather, they prefer CO2 which contains a 12C atom)
** [Butcher, p 86].* Indeed, the 13C fractions in the atmosphere and
** ocean surface waters declined over the past decades [Butcher, p 257]
** [C.Keeling] [Quay] [Schimel 94, p 42].* This fits a fossil fuel CO2
** source and argues against a dominant oceanic CO2 source.* Oceanic
** carbon has a trifle more 13C than atmospheric carbon, but 13CO2 is
** heavier and less volatile than 12CO2, thus CO2 degassed from the
** ocean has a 13C fraction close to that of atmospheric CO2 [Butcher,
** p 86] [Heimann].* How then should an oceanic CO2 source cause
** a simultaneous drop of 13C in both the atmosphere and ocean ?
Torqumada said:
As I have said before, I have no doubts that humans have added to the CO2 levels, I jsut don't think its a primary cause of Global warming. I think its a natural part of Earth's climatologicla cycle and we just happen to have the technology and knowledge to witness it at this time.
There are two separate questions here though:
1. Do you agree that the increase in CO2 concentration in the atmosphere is most likely due to human causes?
2. Do you agree that the increase in CO2 concentration is a primary cause of global warming?
My guess is that the minority of climate scientists who are "global warming skeptics" would cast doubt on #2, but virtually none would have significant doubts about #1. I'd be curious as to whether you could find even a single real scientist who thinks #1 is likely to be false, and whether they have responses to the various lines of evidence mentioned in the website above.
