APOD: Global Warming Predictions (2009 April 21)

[kingernie]

By extrapolating from a carefully selected and limited data set one could project nearly any desired conclusion. Why wasn't the data for the period from 1990 to 2008 included? It is always bad methodology to extrapolate far into the future based on a small data sample. Those who are promulgating the global warming theory as fact are practicing bad science.

[neufer]

http://apod.nasa.gov/apod/ap090421.html

By extrapolating from a carefully selected and limited data set one could project nearly any desired conclusion. Why wasn't the data for the period from 1990 to 2008 included? It is always bad methodology to extrapolate far into the future based on a small data sample. Those who are promulgating the global warming theory as fact are practicing bad science.

I'm guessing that this is (HadCM3) model results rather than simple extrapolation from recent trends.

In any event, bad science would be to not make any hypothesis at all.

[Chris Peterson]

By extrapolating from a carefully selected and limited data set one could project nearly any desired conclusion. Why wasn't the data for the period from 1990 to 2008 included? It is always bad methodology to extrapolate far into the future based on a small data sample.

This projection is not extrapolated from data. It is the output of a well regarded, and somewhat conservative climate model. Climate models are based on our best understanding of numerous physical processes. That is, their output is the solution of a large number of equations that have physical bases. The output is tested against as much historical data as is available. This particular model uses the 30-year average from 1960 to 1990 as its baseline, presumably because that was the best and most recent data available when the model was first being developed. The choice of baseline is arbitrary, and doesn't impact the quality of the output. The model does include data from 1990 to 2008; you are confusing the data used as an input to the model with the baseline used as a reference for the temperature change predicted near the end of the century. They are completely different things. The model is developed against several hundred years of historical data; for a few parameters several thousand years, even.

Those who are promulgating the global warming theory as fact are practicing bad science.

People who say that don't ever seem to be able to point to much in the way of primary research supporting that position. In fact, the science is very good. There is no evidence at all of bad science, of fraud, of bias. The fact that there are a range of models giving different results (all show heating, however) is sometimes claimed to mean there is this mysterious "bad science". But all it really means is that the theories that are used to describe the climate are not yet fully developed. That doesn't mean there is anything at all wrong with the science.

That the global average temperature is rising rapidly is almost undisputed; it can pretty much be taken as fact. Global warming isn't a "theory", it is an "observation". The theory that the rise is anthropogenic is very well supported by the evidence, but no reputable scientist would call it a "fact". But most accept it as likely, and differ only in their opinions of its magnitude.

[StACase]

A future temperature increase like that shown on the above predictive map may cause sea levels to rise, precipitation patterns to change, and much pole ice to melt. The result could impact many local agricultures and the global economy. Geoengineering projects that might include artificial cloud creation might reduce the amount of warming sunlight that reaches the Earth's surface.

"May cause" and "could impact" So let's do some "geoengineering" that "might include" and "might reduce"

If any one wants to read about geoengineering, here's a link:

http://iis-db.stanford.edu/pubs/22456/T ... Option.pdf

My opinion?

Polluting the oceans with nutrients, the air with aerosols and interplanetary space with dust particles, plus releasing bioengineered mutant plants into the environment wouldn't pass muster with anyone especially any environmentalists that I know of.

[Voyager123]

Greetings,

" ... Few disagree that recent global warming is occurring ...."

"global warming" is highlighted and links to National Geographic.

Fine.

However, "Few disagree" is not highlighted with link(s) to any of a broad
number of scientists who so disagree with this media hyped conclusion.

Missing also was any 'link' to a transcript of a FULL debate in a public forum
between followers of Nobel prize selectee Gore versus luminaries in the
science community who dismiss his adventurous notions

[jluetjen]

It's amazing how paradigms can pause blind spots...

Compared to the past 100 million years, the Earth is currently enduring a relative cold spell, possibly about four degrees Celsius below average.

Over the past 100 years, however, data indicate the average global temperature of the Earth has increased by nearly one degree Celsius.

The choice of baseline is arbitrary, and doesn't impact the quality of the output. The model does include data from 1990 to 2008;

A rational person wouldn't be surprised by a trend returning towards average, but yet somehow Global Warming advocates extrapolate that this must indicate the hand of mankind on the climate. Basic statistics suggest that you can't extrapolate a meaningful trend based on a select 1/1,000,000th sample of the data. Why do people believe that a warming trend over the last 100 years is statistically meaningful???

For example, if you look at temperature over millions of years, you see this...


The variation appears to be increasing, but the current and predicted range of temperatures is certainly within the normal range.

If you change the scale to 5 million years, you see this... (Temperature is the scale on the right)


Now the world is definitely cooling, but still the variation appears to be increasing, and this trend definitely exceeds the period of the industrial age.

Now if CO2 really is the cause of man-made global warming, we would expect to see it on a graph such as this one.

(from this site: http://www.geocraft.com/WVFossils/Carbo ... imate.html )

Certainly in the past when CO2 levels were far higher then they are now, and the world was a warmer place, you would have expected to see some sort of coorelation between CO2 levels and temperature. For example CO2 levels have dropped by something on the order of 90%, while the temperature changes have not shown anything near as consistent of a change. So why are we now expected to believe that because (man made) CO2 has gone up by a relatively minor rate that it's suddenly going to have a drastic impact on the climate? That defies reason.

They're talking about quanties of CO2 of less then 400 PPM, or .04% (0.0004) of the entire atmosphere. The atmosphere on Venus for example is 97% CO2. The composition of the earth's atmosphere is 79% nitrogen, 20% oxygen, and 1% other gases. So the premise is that a change of about 0.0002 is going to have the affect that we're talking about right now, dwarfing other factors. Mars's atmosphere is 95.3% CO2, and raises it's surface temperature by about 5 degrees K, in spite of the fact that it has almost no ice cover, and no oceans. The chart above shows that the Earth's temperature varies by about 8 degrees.

Maybe we should stop wringing our hands about how to control the changes in climate and starting thinking about how to live within the climate variations that we have. I have yet to see anything that suggest that we'll be any more successful at definitively managing the world's climate than a child's sand-castle is at holding back the tide -- or New Orleans is at holding back the sea.

[zypldt]

According to your own rules, "This is a scientific forum which focuses on astronomy". The Global Warming hysteria is neither scientific, nor astronomic. It has been proven that the original "hockey stick" was a flawed model. Furthermore most scientists are jumping off of the global hysteria bandwagon for a more realistic and scientific approach to climate change (i.e., it is most likely caused by variations in the solar magnetic field and sun spots, among many other factors - none of which are man made).
I realize that many scientists are paid a great deal of money to promote this theory, just as they were paid to promote Eugenics in the early 1900s. A simple look at the fact that the "results" won't be seen for 100 years or more means these scientists can make wildly bold predictions now. Thus, when they are due to come to fruition, no one will be around to be taken to task for the bogus models - nor will anyone care.
Let's for a moment assume that the temperatures are rising and global warming is real. According to this site, http://www.gsfc.nasa.gov/scienceques2005/20060310.htm, "At the South Pole (Amundsen-Scott Station), the average temperature of the warmest month (January) is – 18º F (–28.2º C)". That means if the temperature rises according to this model (by about 4º C at the South Pole), it will only reach -24º. Hardly enough to melt the southern polar cap. Not only that, if the average world temperature rises, the ocean temperatures will rise, creating more evaporation. More evaporation means more rain. More rain means more snow and ice at the extreme altitudes and latitudes.

[neufer]

A rational person wouldn't be surprised by a trend returning towards average, but yet somehow Global Warming advocates extrapolate that this must indicate the hand of mankind on the climate. Basic statistics suggest that you can't extrapolate a meaningful trend based on a select 1/1,000,000th sample of the data. Why do people believe that a warming trend over the last 100 years is statistically meaningful???

For example, if you look at temperature over millions of years, you see this...

The variation appears to be increasing, but the current and predicted range of temperatures is certainly within the normal range.

The variation includes a half dozen or more MAJOR ice ages over the last million years;
we really need to figure out how to STOP the next ice age
by having a better grasp on climate and how to modify it in productive ways.

Allowing unlimited greenhouse gas emission is NOT the way to do that!

If you change the scale to 5 million years, you see this... (Temperature is the scale on the right)


Now the world is definitely cooling, but still the variation appears to be increasing,
and this trend definitely exceeds the period of the industrial age.

The TREND in degrees per decade is minuscule compared to the trend over the last 150 years.

(from this site: http://www.geocraft.com/WVFossils/Carbo ... imate.html )

Now if CO2 really is the cause of man-made global warming, we would expect to see it on a graph such as this one.

One DOES see the correlation in this graph!

Physics says that the correlation would be even better if CO2 amounts were done on a logarithmic scale... and it would.

This is actually quite surprising considering that there are many other factors modifying the climate over hundreds of millions of years.

[Peter FBIS]

Let's just put to one side for the moment the question of whether little or much of the predicted change is contributed to by man. There appears still to be a consensus that this warming is happening and its effects on mankind could be severe.

That would suggest that we should certainly be looking at ways to mitigate the effect.

Having said that; what is the impact of what appears to be a lower than normal sun-spot minimum that we are currently experiencing. And if that is likely to be significant, is it of anywhere near the same order as the Global Warming phenomenon?

[Pops]

I'm quite saddened that, yet again, APOD has shown its true colours by pumping out politically-motivated hype. What about using current data? For the truth, go to http://scienceandpublicpolicy.org/repri ... etter.html. I'll no longer bother visiting the APOD site. I can get all the skewed politics and outright lies I need by viewing any BBC or CNN web-site, or any NASA web-site now it seems. What a sad, cooling world we live in.

Bye.

[aristarchusinexile]

Those who are promulgating the global warming theory as fact are practicing bad science.

Like the ancient adage goes, "It ain't rocket science..." to know for certain without any doubt whatsoever that humanity is pouring heat into the atmosphere, and that heat is melting glaciers, and even rightwing fantatics are starting to sweat under the collar and other places.

[aristarchusinexile]

According to your own rules, "This is a scientific forum which focuses on astronomy". The Global Warming hysteria is neither scientific, nor astronomic. .

The sun does have a part in this .. and the sun is part of astronomy .. so zilch to you zypldt.

[Chris Peterson]

Having said that; what is the impact of what appears to be a lower than normal sun-spot minimum that we are currently experiencing. And if that is likely to be significant, is it of anywhere near the same order as the Global Warming phenomenon?

It is an insignificant effect. Climate models include the solar cycle, and its effects are very slightly detectable above the measurement noise level in terms of weather or decadal climate. The solar cycle imposes a variation on the average solar output of ±0.05%. Our current deeper than typical minimum really only means we've been sitting at the low end of that for a few extra months (out of eleven years). The only known example of solar output significantly affecting climate (not certain, but likely) was the Maunder Minimum in the 17th century, when solar activity was low for almost a century. Unless the Sun starts acting very differently than anybody expects, the current activity isn't going to have much impact, and certainly isn't offsetting global warming.

We don't have a good record of the long term solar irradiance. Data collected over the last few decades shows a very slight trend which would affect climate if it continued over centuries. I think that the models turn that effect on or off as part of their "best case/worst case" analysis.

[Bazzz]

Very disappointing to see APOD buy into unscientific trendy rubbish like global warming. The statement that "Few disagree that recent global warming is occurring" is an outright lie. The IPCC is an unscientific politically motivated panel of bureaucrats interested in taxing plant food (popularly referred to today as carbon dioxide). APOD owes its readers an apology for promoting an unscientific political agenda. Shame on you.

[geckzilla]

I almost think today's APOD has nothing to do with global warming and everything to do with a devious little experiment to see if we can attain a thread with a new highest reply count.

[Chris Peterson]

Those who are promulgating the global warming theory as fact are practicing bad science.

Like the ancient adage goes, "It ain't rocket science..." to know for certain without any doubt whatsoever that humanity is pouring heat into the atmosphere, and that heat is melting glaciers, and even rightwing fantatics are starting to sweat under the collar and other places.

We aren't putting enough heat into the atmosphere to even be detectable. The only significant energy source in global warming is the Sun.

[neufer]

Having said that; what is the impact of what appears to be a lower than normal sun-spot minimum that we are currently experiencing. And if that is likely to be significant, is it of anywhere near the same order as the Global Warming phenomenon?

It is an insignificant effect. Climate models include the solar cycle, and its effects are very slightly detectable above the measurement noise level in terms of weather or decadal climate. The solar cycle imposes a variation on the average solar output of ±0.05%. Our current deeper than typical minimum really only means we've been sitting at the low end of that for a few extra months (out of eleven years). The only known example of solar output significantly affecting climate (not certain, but likely) was the Maunder Minimum in the 17th century, when solar activity was low for almost a century. Unless the Sun starts acting very differently than anybody expects, the current activity isn't going to have much impact, and certainly isn't offsetting global warming.

We don't have a good record of the long term solar irradiance. Data collected over the last few decades shows a very slight trend which would affect climate if it continued over centuries. I think that the models turn that effect on or off as part of their "best case/worst case" analysis.

http://en.wikipedia.org/wiki/Little_ice_age

There is a case that can be made for about half of the current [0.8º C] global warming from 1850 to the present being attributable to a recovery of from 400 year long "The Little Ice Age" [from 1450 to 1850].

http://upload.wikimedia.org/wikipedia/c ... arison.png

Throughout the Little Ice Age, the world experienced heightened volcanic activity [as well as heightened human wood burning].

During the [much shorter 70 year] period 1645–1715, in the middle of the Little Ice Age, there was a period of low solar activity known as the Maunder Minimum. A growing body of scientific evidence indicates that there is a very low understanding of the correlation between low sunspot activity and cooling temperatures. The Spörer Minimum has also been identified with a significant cooling period near the beginning of the Little Ice Age. Other indicators of low solar activity during this period are levels of the isotopes carbon-14 and beryllium-10.

http://en.wikipedia.org/wiki/File:Carbo ... labels.png

[Chris Peterson]

I almost think today's APOD has nothing to do with global warming and everything to do with a devious little experiment to see if we can attain a thread with a new highest reply count.

As soon as I saw this APOD (which I think is quite nice), I knew the sparks would fly...

[mhervey]

It is hard to see how today's important topic (Global Warming) has much to do with main-stream astronomy. Surely global warming is not the main interest of professional astronomers. If there is a connection between this topic and astronomy, then the author(s) should point out the connection. Today's topic and the way it was presented made me wonder if someone has an agenda, and that is disturbing on a site that I expect to be neutral and objective in its presentations. Following the link "no one is sure" takes one to other links that are mostly unexplained: for example what are the family of curves in "global warming projections" of that link. If I presented a graph of this quality at work, I would be slammed by those reviewing it. Keep Global Warming in its proper forum

[jluetjen]

A rational person wouldn't be surprised by a trend returning towards average, but yet somehow Global Warming advocates extrapolate that this must indicate the hand of mankind on the climate. Basic statistics suggest that you can't extrapolate a meaningful trend based on a select 1/1,000,000th sample of the data. Why do people believe that a warming trend over the last 100 years is statistically meaningful???

For example, if you look at temperature over millions of years, you see this...

The variation appears to be increasing, but the current and predicted range of temperatures is certainly within the normal range.

The variation includes a half dozen or more MAJOR ice ages over the last million years;
we really need to figure out how to STOP the next ice age
by having a better grasp on climate and how to modify it in productive ways.

Allowing unlimited greenhouse gas emission is NOT the way to do that!

Huh??? I thought you were arguing to stop global warming, now you're advocating that we stop global cooling???

If you change the scale to 5 million years, you see this... (Temperature is the scale on the right)


Now the world is definitely cooling, but still the variation appears to be increasing,
and this trend definitely exceeds the period of the industrial age.

The TREND in degrees per decade is minuscule compared to the trend over the last 150 years.

(from this site: http://www.geocraft.com/WVFossils/Carbo ... imate.html )

Now if CO2 really is the cause of man-made global warming, we would expect to see it on a graph such as this one.

One DOES see the correlation in this graph!

Physics says that the correlation would be even better if CO2 amounts were done on a logarithmic scale... and it would.

This is actually quite surprising considering that there are many other factors modifying the climate over hundreds of millions of years.

Huh? (again). It's been a few years since I took stats, but I think that you're confusing Physics with Statistics and graphs with correlation. Correlation is a mathematical evaluation which is not dependent on scale. Using Excel I did the followinig:

1) I eyeballed the data and came up this table:
MYA CO2 Avg Temp
550 4750 22
500 4750 22
450 4500 12
400 3500 22
350 1200 20
300 400 12
250 1900 22
200 1300 22
150 2300 16
100 1500 22
0 200 12

2) I ran the correlation function against this data and came up with a correlation coefficient of ~0.303. This would suggest a weak correlation between CO2 and average Temperature. If I remember correctly, this means that about 1/3 of the variation in the temperature is related to the change in the CO2. Changing the scale will not change this. While a strong correlation between CO2 and Temperature would not prove causality, a weak correlation pretty much eliminates the possibility of changes in CO2 causing significant temperature changes within the context of the data.

[sid113]

There were many comments about good science, bad science, and what not. Let us not forget the scientific method. That method requires the following

• Make observations on a subject (in this case global temperature changes)
• Develop an hypothesis based on those observations
• Make predictions based on the hypothesis
• Define experiments or additional observations to test the hypothesis
• Execute the experiment or make the observations and compare against the prediction
• If the predictions are not observed, go back to the second step. If the predictions are observed go to the third step

Only after many, MANY iterations of this method can we say that the hypothesis is converging on something useful and sound.

So, there was an observation do in the '80's that showed a rapid increase in temperature. A hypothesis was created that stated that the increase was caused by a greenhouse gas effect, and CO2 being a greenhouse gas, the cause was manmade CO2 accumulation. The prediction of this hypothesis is that global temperatures will continue to rise until we can stop manmade CO2 accumulations.

since it is a little tough to run the experiments directly on the earth, computer models of the earth were created and the experiments were run on these computer models. The output of these simulated experiments met the predictions, and therefore the scientists indicated that hypothesis had some credence. Additional hypothesis were added (such as melting glaciers) and the models were enhanced to run the experiments to test the new predictions.

All of this is good science in that it is following the scientific method correctly. The problem, of course, is the experiments are not conducted on anything real, but only on models of a real thing. If the model is incomplete or biased, the data gathered from the model is incomplete or biased. The conclusions therefore can only apply to the model, not to reality. However, this subtle distinction is lost on the media and the government officials pushing an agenda. It is also not emphasized by the scientists looking for funding from the same government officials pushing an agenda.

So, how do we test the models? Well, we could run them backwards and see if they predict the past (they all fail). We could test to see if the predictions on data from 1900-1999 entered into the model will predict accurately the results of 2000-2009 (especially 2008-2009 timeframe). As far as I know, they all fail with this test as well.

You see, all of the models are based on limited time scale. They do not take into account the 26,000 year earth tilting cycle. They do not take into account solar activity cycles. They do not take into account non-cyclic volcanic and other seismic actions. They do not take into account any other action for which we have little or no data.

When we look at the data over ranges exceeding 200 years (the timeframe covered by large CO2 generations by humans), we see a much larger set of variations in climate than anything predicted by these models, and all of these variations have nothing to do with manmade anything.

So, that bad science comes from using a model and a data set that is so limited in scope that it cannot possibly be anything by an interesting game, and yet refusing to accept the scientific method once this has been pointed out. Worse, the politicians with an agenda are now using this “science” as the ultimate law of nature to force new policies on the public. These new policies will cost the world trillions of dollars in redirected resources and efforts to solve a problem that not only does not exist, but the solution will not have a measurable impact to climate change at all. (If pumping billions of tons of CO2 into the atmosphere is NOT the cause of climate change, then eliminating this CO2 generation will not reverse any observed climate change.)

I would like to make one last point before I leave this subject. Someone was looking at the temperature/CO2 level graph of the last 600 million and saw what they considered to be a correlation. First off, it is amazing what patterns the brain can see in random events (many pieces of modern art use that very feature). However, even if there was a complete correlation, the scale is such that it is impossible to tell which is causing what. It is highly possible, and in fact likely, that increases in temperature actually cause an increase in CO2, not the other way around.

As an example, I will point out that a tropical rain forest does not absorb any new CO2 – it has reached a steady state. While a tree may store a lot of CO2, when that tree dies and falls over, the bacteria, fungus, ants, termites, and what not eat that tree, releasing all of its carbon back into the atmosphere as CO2 and methane (and a few other things). None of it remains in the ground, sequestered from the atmosphere. On the other hand, a savanna actually does sequester carbon in the form of an increasingly thick layer of loam (where organic material stays in the dirt). As temperatures rise, tropical, sub tropical, and temperate forests expand into grasslands, resulting in releasing of the carbon stored in the dirt. This release could, in fact, cause CO2 levels to rise because temperatures rose. This is opposite of the assumption that it is CO2 levels rising that causes the temperatures to increase.

The bottom line: we know so little about how our climate works, and our models are missing so much information, any policy based on these hypotheses is way more likely to be wrong – completely wrong than they are in being right. It would be nice if this issue would be removed from public debate and put back into the scientific method.

[Chris Peterson]

So, how do we test the models? Well, we could run them backwards and see if they predict the past (they all fail).

Where do you get this. In fact, the models all do a reasonable job of predicting the past over the ranges they are designed around (that is, there are long range models- thousands of years or more, and short range models, centuries). The single most important component used to verify models is their success in predicting past climate.

We could test to see if the predictions on data from 1900-1999 entered into the model will predict accurately the results of 2000-2009 (especially 2008-2009 timeframe). As far as I know, they all fail with this test as well.

Again, I'd ask where you get this. On the whole, it makes little sense as the time frames are associated with weather, not climate. No model can be used to test if a particular year or two fits. Even a single decade is barely within measurable trends. Yes, the last two years were quite a bit warmer than the midline of the models suggests. But those sorts of fluctuations are completely normal, and meaningless within the context of climatic change.

You see, all of the models are based on limited time scale. They do not take into account the 26,000 year earth tilting cycle. They do not take into account solar activity cycles. They do not take into account non-cyclic volcanic and other seismic actions. They do not take into account any other action for which we have little or no data.

All of the major models include the Earth's dynamics- orbital variations, precession, etc. All of the major models include solar cycles. They do not include volcanic activity because it is unpredictable, and for short term climate- a few centuries- almost certainly unimportant. Naturally the models do not include effects that are not yet understood. A big part of climate research involves trying to better understand geophysical processes so they can be included or improved in models. That's the "good science" bit again.

I would add that we're mostly concerned with short term climate change. Stuff that happens over thousands or millions of years (the typical sort of change seen in the physical record) is not a problem for us. The concern is with the rapid change seen in the last century. To be useful in term of public policy, predictions of future climate really only need to be reasonable over decades or maybe a century. Over these short ranges, confidence in current models is fairly high.

[jluetjen]

So if large scale changes in CO2 (5000 PPM or more down to 400 PPM) do not correlate with changes in temperature, how you can you argue that relatively small scale changes in CO2 -- say a 50% increase from 400 PPM to 600 PPM, is causing a short term increase in the global temperature?

[aristarchusinexile]

I would add that we're mostly concerned with short term climate change. Stuff that happens over thousands or millions of years (the typical sort of change seen in the physical record) is not a problem for us. The concern is with the rapid change seen in the last century. To be useful in term of public policy, predictions of future climate really only need to be reasonable over decades or maybe a century. Over these short ranges, confidence in current models is fairly high.

Chris, We've done it again! We agree! I would modify though, "the concern is with the rapid change seen in the last 30 years."

[hydroresearch]

Please stop the false and irresponsible reporting of the climate change issue. Climate alarmist are causing a frenzy of bad science to be reported as facts. The post today is just one example. Perhaps APOD should stick to astronomy and not delve into other scientific disciplines. As a practicing climatologist for over 30 years I would suggest that you at least consider looking at some real science being done in the field of climatology at c02science.org.

Thanks for having a great site,

Richard Ziriax
Climatologist
Hydro Research