A new, longer, hockey stick

There is a new hockey stick in town, one with a shaft extending back all the way back to the end start of the holocene about 12,000 years ago when the last ice age ended:

In the article in science A Reconstruction of Regional and Global Temperature for the Past 11,300 Years the authors took 73 proxy records of temperature, including isotope ratios and pollen records in order to construct a global temperature record for the entire Holocene period. The reconstruction shows warming as we exit the last ice age, followed by a stable temperature plateau, followed by about 5000 years of gradual cooling ending around the beginning of the last century.

The decade of 1900-1909 was colder than about 95% of all other decades in the Holocene. But since then there has been rapid, unprecedented, warming so that by the end of the most recent decade (2000-2009) temperatures were warmer than about 82% of all decades in the past 12,000 years.

In 100 years we have undone 5000 years of natural cooling!

This reconstruction fits nicely with what we already know (especially for the past 2000 years) and doesn’t really change the big picture, it just adds some details. This gives us confidence that these results will withstand the inevitable post publication scrutiny.

There are a few important takeaways from the paper. The first, is a concern raised by Robert Rohde:

In discussing their result, there is one important limitation that I feel deserves more attention. They rely on proxy data that is widely spaced in time (median sampling interval 120 years) and in many cases may also be subject to significant dating uncertainty. These effects will both tend to blur and obscure high frequency variability. They estimate (page 1, column 3) that only 50% of the variance is preserved at 1,000-year periods. This amount of variance suppression is roughly what you would expect if the underlying annual temperature time series had been smoothed with a 400-year moving average. In essence, their reconstruction appears to tell us about past changes in climate with a resolution of about 400 years. That is more than adequate for gathering insights about millennial scale changes during the last 10,000 years, but it will completely obscure any rapid fluctuations having durations less than a few hundred years. The only time such obscuring might not occur is during the very recent period when dating uncertainty is likely to be low and sample spacing may be very tight.

While Michael Mann disagrees:

Regarding the resolution issue, this was my main concern initially when I looked at the paper. But I’m less concerned now that I have read the paper over more carefully, because I think that Figure 1a and 1b give a pretty good sense of what features of higher resolution reconstructions (specifically, our ’08 global reconstruction which is shown) are potentially captured. Based on that comparison, I’m relatively convinced that they have the resolution to capture a century-long warming trend in the past were there one comparable to the recent trend.

I think there is something to Rohde’s concern. It is possible than a century long blip in temperature might not show up in this reconstruction so in some way we are limited in how we compare this reconstruction to the past 100 years of warming. However we know with a lot of certainty that our current warming trend is not just a blip. Since the CO2 we have emitted into the atmosphere is going to stay there for a very long time the current change in temperature will be much longer than a simple blip. And were something similar to have happened in the Holocene this reconstructions would have very likely picked it up.

The other thing to keep in mind, also brought up by Michael Mann (and echoed by Jeremy Shakun, one of the study authors) is that the variation in the ‘shaft’ of the hockey stick potentially has more variation than it should:

My only real concern is that their data and approach (e.g. the use of pollen records in the higher northern latitudes) seems to emphasize the higher latitudes of the Northern Hemisphere, during the summer season. This is an issue because we know there is a substantial long-term natural cooling trend for high-latitude summers because of Earth orbital effects, but the trend is nearly zero in the global annual average. One gets the sense from looking at their reconstruction that there is a very strong imprint of this orbital cooling trend — stronger than what one would expect for the global annual average.

The interesting thing about that, is that it suggests that the true conclusions might even be stronger than their already quite strong conclusions, regarding the unprecedented nature of recent warming. That is, it may be that you have to go even further back in time to find warmth comparable — at the global scale — to what we are seeing today. If you look at their tropical stack for example (Figure 2J) [a particular set of data], the modern warmth is unprecedented for the entire time period (i.e, the past 11,000 years). That’s why I said that there results suggests recent warmth unprecedented for at leaat the past 4,000. It’s possible, given the potential seasonality/latitudinal bias, that there is in fact no precedent over the past 11,000 years (and likely longer, since the preceding glacial period was almost certainly globally cooler than the Holocene) for the warmth we are seeing today. In that case, we likely have to go back to the last interglacial, i.e. the Eemian period (125,000 years ago) for warmth potentially rivaling that of today.

Here is Jeremy Shakun being interviewed by Andy Revkin:

So while this reconstruction indicates that current temperatures are still slightly cooler than the Holocene climatic optimum (this isn’t anything new) there is reason to think that this might represent a bias in the data and not reality. It would not be surprising to learn that current temperatures are higher than anytime in the past 125,000 years.

Regardless, if we aren’t yet warmer now than in the past 125,000 years then our current temperature trajectory means that we will get there before too long.

Lastly one of the most important features in the reconstruction is the difference between the warming at the end of the last ice age (left side of the graph) and the current warming (right side of the graph). It demonstrates how truly abnormal the current change in climate really is.

What matters more than how much climate changes, is how fast it changes. If it changes slowly, then human society and natural ecosystems (and everything in between) will have a much easier time adapting. If, on the other hand, the change is quick (as the reconstruction shows that it has been) then adaptation becomes much more difficult and suffering becomes inevitable.

But quick change seems to be where we are headed, and we are doing little to mitigate that change. We might be able to adapt to it, but we wont be able to avoid suffering. That is the path our inaction is leading us towards.

It isn’t the path I want to be on.