Even if global warming isn't real (which it is) there are still many other forms of environmental destruction.

As the title suggests, this thread is specifically about global warming and climate change.  

If you want you can talk about environmental effects related to it in this thread, but if you want to talk about the environment in general then make a new thread.

[youtube]r7aZ6vqCk2E[/youtube]

Well researched and well presented, particularly the segment about change through Earth history.  This video is much more interesting and watchable than the title would suggest.

The video is right in addressing half truths and out of context presentations such as that of the video targeted, but is not free of the same faults.  In particular, nailing the 21st century sea level rise to 3 feet instead of 1 foot is equally stretching the IPCC summary, both numbers are outside the median projections but barely within the likely range.  So it's strictly a draw in honesty on that topic, but the rebuttal video does correctly touch on the complexity of the question.  The video also fails to defend the hockey stick.  It's of course silly to use the GISP2 data and point out that there is no 20th century uptick, but it's frankly just as silly to attach an instrument record to proxy records, because the temporal resolutions of proxies and instrument records are significantly different.  For instance, ice core data has a sealing time of nearly a century.  The second half of the video, about political abuse of science, is spot on.

The video also fails to defend the hockey stick.  It's of course silly to use the GISP2 data and point out that there is no 20th century uptick, but it's frankly just as silly to attach an instrument record to proxy records, because the temporal resolutions of proxies and instrument records are significantly different.  For instance, ice core data has a sealing time of nearly a century.

Temporal resolution of gas abundance and isotopic fractionation are indeed limited by diffusion through the unconsolidated snowpack, but other factors like precipitation (from annual layer depth) and summer temperature (melt layers) are not.  Tree rings and varves also provide climate information at the annual scale.

Paleoclimate research is a pretty complex field of science and these sorts of misunderstandings are unfortunately quite common in the popular discussion.  A good rule of thumb is that if you think a very well established and verified result in a field of science is wrong, and your reasoning for why is simple, then most likely it is your reasoning that is in error.  I think the video did a good job of explaining the error with trying to reject the hockey stick by using ice core data.  I do agree with you that his presentation of a "corrected" figure could be better though.  He did show that it is robust, with many independent studies using different methods arriving at essentially the same thing, but I think it is also important to show what the hockey stick graph does and does not constrain in terms of global temperature change in that period.  For example there are variations in the record that are larger, but they are brief (volcanoes).  But it takes a lot more time to explain, and I think is beyond the scope of a response/corrections video.

 In particular, nailing the 21st century sea level rise to 3 feet instead of 1 foot is equally stretching the IPCC summary, both numbers are outside the median projections but barely within the likely range

I don't think he is so much trying to nail what it is as showing what it isn't.  His argument is that the prediction of 1 foot based on linear extrapolation is too low, because the rate of rise will accelerate, and he presented research justifying the expectation of acceleration.  IPCC AR5 WGI says that sea level rise by 2100 under RCP 8.5 is 0.74 +/- 0.23 meters, or between 1.7 and 3.2 feet.  So in fact a claim of 1 foot by simple extrapolation is too low by 1.9 sigma, while 3 feet is within the 1 sigma uncertainty range.

He could have made a point of quoting the IPCC ranges more concisely, and showed that the result of linear extrapolation is outside that range unless we follow a lower RCP.

A good rule of thumb is that if you think a very well established and verified result in a field of science is wrong,

I didn't say that the data was wrong, but you must be very careful about what the data is used to prove.  You can't simply stick very different ways of measuring things to each other and highlight an abrupt change starting at the transition.

So in fact a claim of 1 foot by simple extrapolation is too low by 1.9 sigma, while 3 feet is within the 1 sigma uncertainty range.  

I don't think the emission scenario was specified.  1 foot is within RCP2.6.  In any case, to give a figure in the upper range of the worst case scenario without stressing that this is what you do isn't a good rebuttal in a video targeting half truths.

I didn't say that the data was wrong, but you must be very careful about what the data is used to prove.  You can't simply stick very different ways of measuring things to each other and highlight an abrupt change starting at the transition.

Proxy data extend into the instrumental record, and they show an abrupt transition.  It is not simply an artifact of slapping different kinds of measurements together, nor is it an artifact of calibrating proxies to instrumental temperatures.

There are issues of scientific integrity with presenting a smoothed reconstruction and slapping an unsmoothed historical record at the end of it, because this masks a lot of true variations in global temperature.  Personally I prefer the unsmoothed ones for this reason.  But those faster, larger variations are basically just volcanoes and solar cycle -- nothing else perturbs the system that dramatically that quickly.  The most recent change comparable in magnitude and timescale to what is occurring now was ~55 million years ago, and it left a record in the rocks and fossils.

I don't think the emission scenario was specified.  1 foot is within RCP2.6.

Not explicitly, but I think it is pretty clear that both are in the context of business as usual.

RCP8.5 is business as usual with proven fossil fuel reserves.  RCP2.6 is an enormous concerted global effort to mitigate climate change, with emissions peaking in 2020 and then declining to zero, and CO2 concentration peaking at ~450ppm in 2050.

So the most accurate statement is that 1 foot of sea level rise is in the potential futures, but not in the potential futures where there is not immediate and dramatic action to mitigate climate change.

Proxy data extend into the instrumental record, and they show an abrupt transition.  It is not simply an artifact of slapping different kinds of measurements together, nor is it an artifact of calibrating proxies to instrumental temperatures.

I searched a bit, and I found one paper by Ljungqvist using proxies extending into the instrumental record, going 2000 years back.  It's only for the extratropical NH and not all the proxies used cover the entire period, so it's not quite apples to apples, but the study seems pretty comprehensive.  Yes, the proxies show the modern warming, but they fail to produce the hockey stick that you get by splicing proxies and instrumental observations.  Let's look at Ljungqvist's data(temperature relative to the 1961-90 average): It doesn't mean that the instrumental records exaggerate the warming.  I think the proxies are just sluggish.  But it also means that any previous abrupt changes over the last 2000 years will also appear dampened in the proxy data.  We don't really know if there were such abrupt changes.

I searched a bit, and I found one paper by Ljungqvist using proxies extending into the instrumental record, going 2000 years back.  It's only for the extratropical NH and not all the proxies used cover the entire period, so it's not quite apples to apples, but the study seems pretty comprehensive.  Yes, the proxies show the modern warming, but they fail to produce the hockey stick that you get by splicing proxies and instrumental observations

Good approach and I understand your confusion.  I think you come close to recognizing the misconception, and the solution is to think carefully about the implications of the underlined.

The region covered by these proxies (30N to 90N) is 25% of the surface area of the Earth.  So this dataset will emphasize regional climate changes in that area more than a global dataset.  In particular, this region will emphasize the Medieval Warm Period.  

Try comparing the form of this graph to more spatially complete reconstructions.  You'll see it smooths out a lot even just by including the northern tropics.

So that is the key insight.  You need to think about regional vs. global changes, and how they will appear in regional vs. global datasets.  Regional temperature is subject to internally driven climate dynamics, like what the ocean or atmospheric circulation does.  Global temperature is subject to changes in the energy balance on a planetary scale -- such as solar output, volcanoes, albedo, and greenhouse effect.  When you understand what is capable of driving global change, the form of the Hockey Stick will make sense.  The record shows all of these influences, and their magnitudes correspond to the magnitude of the change in radiative forcing.  (The volcanoes don't show up in the above graph because the effect is too brief -- just a few years -- while the data have decadal resolution.  You can see them in higher resolution datasets).

Added:  That many proxies don't cover the whole period is also important to your conclusion.  For example it means the current warmth is slightly under-represented here, since some proxies don't cover the last few decades.  It may produce some biases in previous climate changes as well, depending on what was going on in areas where these proxies are dropped.

Accounting for dropout of proxies in global reconstructions is an important part of the analysis of uncertainty.  It's a big part of the reason why the uncertainty increases farther back in time. 

Well, I looked for what you asked and this was the closest I could find which extended into the instrumental record without splicing instrumental observations and also uses an extensive set of proxies.  I hadn't seen it before, but it seems to have attracted some attention (it turned up frequently in searches).  I did state its limitations.  If you know of a compound, globally well distributed proxy study, please let us know.  Whilst a 30-90N proxy will emphasise the MWP, it should also emphasise the modern warming, since the modern warming is also somewhat stronger in the extratropical NH.  By the way, in this plot the LIA seems more striking to me than the MWP.

I'm not dismissing the study or your presentation of it. The study plays an integral part of reconstructing Earth's temperature history (this author's work is even included in the IPCC reports).  Reviewing it here is useful for answering important questions -- what proxy data do and do not show, and how we arrive at an understanding of how Earth's temperature has changed over time.  I am glad that you shared it.

I think what you are struggling with is the understanding of what adding in instrumental data does, why they do it, and what this means as far as forming conclusions from the Hockey Stick graph.  That's what I am trying to help you with.

The use of instrumental data is in telling us what happened in recent times.  The use of proxy data is in telling us what happened in ancient times.  However, you see that many proxies do extend into the instrumental period, and they show the form of the hockey stick.  So this is not simply an artifact of mixing two different datasets as you think it is.  Nor is the smoothness of the preindustrial age an artifact.  Another helpful clue that these are not simply artifacts of methodology is that the proxy data show what you would expect them to show from volcanic eruptions and other known changes in radiative forcings.

I think that if you think about this carefully, it will make a lot of sense.

If you know of a compound, globally well distributed proxy study, please let us know.

Sure.  Your best reference for this and anything else regarding our knowledge of Holocene climate history is Chapter 5 of IPCC WGI AR5.  The following figures will be the most relevant:

Whilst a 30-90N proxy will emphasise the MWP, it should also emphasise the modern warming, since the modern warming is also somewhat stronger in the extratropical NH.

Correct, but it is also under-represented due to proxy dropout at the end of this interval.

One obvious problem is that the data for the SH is much thinner than for the NH.  What do you think about this commentary?

Correct, but it is also under-represented due to proxy dropout at the end of this interval.

Right, as I mentioned, the proxies don't all cover the entire period.  Probably for different reasons, and it will be interesting to see how they track the instrumental record in the coming decades.  The "divergence problem" (aka "hide the decline") is an interesting case.

EDIT: This paper sounds interesting, but is paywalled.  (Judith Curry post)

One obvious problem is that the data for the SH is much thinner than for the NH.

True!  But counter-intuitively, not as big of a problem as you might think.  I have a fondness for counter-intuitive things.  Here we encounter another good example of it.

Filling in as many points in space as possible would of course be nice, but it is not necessary to get a good idea of what is happening on a global scale.  Proxies are locations, but they reveal information about dynamics regionally, and often far removed from the location of the proxy itself.  Do you remember on the old forum when I showed how Antarctic ice cores reveal information about the atmospheric dynamics very far away?

A good analogy to this situation is with the production of an image of the Sagittarius A* event horizon using the Event Horizon Telescope.  Each pair of telescopes in the array samples a particular Fourier component of the image.  "Obvious" intuition might lead you to think we would need a very large number of telescopes in as many locations as possible to fill in enough space to produce an adequate image.  Sure, more telescopes would be great!  But it actually takes a surprisingly small number to get an image and do good science.  In particular, the rotation of the Earth changes the components sampled by each pair over time, and the network and data analysis are designed to take advantage of that.  There are also a number of symmetries that we can use to "fill in" missing information without compromising the ability to test important predictions.

Similarly as to how we don't have telescopes wherever we want, we also don't have proxies wherever we want. But we can take advantage of particular climate dynamics and proxy behavior to get a lot more information from them then you would think just based on their location.

So, the basis of the commentary at climateaudit is correct, but can easily generate wrong conclusions because of this counter-intuition.  In any field of science, it is easy to be surprised, confused, and skeptical about how things work or how conclusions are reached/justified if you're not deeply involved in that field. 

That's not to say there are not problems with hemispheric reconstructions, or reconstructions in general.  There are problems and they are discussed in the literature.  Judith Curry correctly points out one of them.  But do these problems imply the Hockey Stick should be rejected?  No.  You have a lot of sources of confidence for concluding that it is essentially correct.

True!  But counter-intuitively, not as big of a problem as you might think.

The size of that problem varies.  The basic problem is that you add layers of assumptions.  Without sending probes we determine the masses of distant binary stars, and even figure out what kind of molecules they have in their atmospheres, which, when you think about it, is magic.  Because we can be fairly confident about the assumptions made.  How solid are the assumptions needed to measure how SH temperatures differ from NH temperatures in Finnish sediments?  It sounds like a fair question to ask, if climateaudit got the underlying data right.

Do you remember on the old forum when I showed how Antarctic ice cores reveal information about the atmospheric dynamics very far away?

The Earth is a single system, so events in one hemisphere will affect the other.  The attribution, however, is hard, and this indirect methodology is prone to confirmation bias, ad hoc reasoning and will easily miss the unexpected.  And as I've said many times, climate science suffers from the compulsive thought that if observations are missing, there simply has to be a way to derive them otherwise.  I think this is to work things out from the wrong end.

But it actually takes a surprisingly small number to get an image and do good science.

You can do the maths and figure out the upper resolution.  The assumptions aren't that many.  But a temperature proxy serving as another temperature proxy?  Is it "settled science"?

But do these problems imply the Hockey Stick should be rejected?  No.  You have a lot of sources of confidence for concluding that it is essentially correct.

I have the impression that the hockey stick has been significantly downplayed by the science (or IPCC) in recent years.  Steve McIntyre apparently calls it spaghetti instead, but if that better shows differences, it's a more honest picture.  How much warmer or colder was the SH in, say, the 6th century than the past century, and what was its variability?  Well, we don't really know for sure.  That seems to be the primary answer.