Friday, 17 August 2012

How high was sea level in the last interglacial?

Let me start with a disclaimer: I will not give a definitive answer to the above question. That being out of the way I can now freely discuss the difficulties of trying to answer a question like that.

It is, in a way, an interesting question: the last interglacial was only ~125.000 years ago, so the continents were all practically in the same location as they are now, and the oceans were equally deep. We know from ice cores and fossil plants and such things how high CO2 levels and CH4 levels in the atmosphere were. We know from ice cores, microfossils, and many other sources how warm it approximately was. So you could say that if you know how high sea level was back then, you know how high sea level is at these CO2 levels and these temperatures, and that might be useful. Additionally; if you can reproduce that in a climate model, you have reason to believe your model resolves sea level well, and may be able to predict it too. Unfortunately, it's not that simple. 

Picture of the 1953 flood

One thing is that we had already passed the CO2 levels of the last interglacial when the Mauna Loa observatory started measuring in the fifties. We may not yet have reached reached globally averaged temperatures comparable to those of 125.000 years ago, but it's hard to pinpoint that; every place on Earth has its own temperature history, and it's not easy to compare two periods that are so similar. And ice does not melt instantaneous. Just suppose we reach, this very day, the very temperatures we had in the last interglacial, the ice caps won’t have had time to adjust to that. And we don’t give them any; temperatures keep rising. We can only get an equilibrium when the situation remains stable for a while. That won’t happen in our lifetime!

So apart from the issues associated with a comparison, we still have this sea level issue to deal with. If you talk about “sea level” in a given time, without further specification, it almost automatically means “globally averaged sea level”. And the problem with that is that you can perhaps measure it today; satellites scan the entire Earth surface, and you can calculate the average of all their data. And sea level famously isn’t level; temperature differences, the rotation of the Earth, wind, gravitational pull of things such as ice sheets and so on, all make the sea surface rather bumpy. So you can’t measure past global average sea levels (further back than the satellite era), as it’s not possible to make a sea level reconstruction for every location on Earth covered by sea. So what can we measure? Local sea level, evidently.

There are many ways of reconstructing local sea level (such as this, this, this, this, and this), and every method, of course, has its own caveats. They also might not represent the same aspect of sea level; some might e.g. reflect low tide, while others are more representative for mean sea level. And tide ranges don't stay constant over time.

And if you have produced a local sea level reconstruction, you're not done; you also have to take into account that not only sea level in itself, but also the Earth’s crust might have gone up and down. Locations affected by earthquakes, volcanism, or (occasional) ice cover are prone to do a certain amount of moving vertically, and fast enough to pose a problem over the time scales discussed here. You can, of course, only use stable regions, but that leaves you with quite a small number of data points. An iconic paper by Robert Kopp and co-workers in Nature tried to extract a global average from a limited amount of data from the Last Interglacial. They, for instance, had no data points along the eastern and western shores of the Pacific; the only data from that ocean came from islands in the middle, and from its polar boundaries. They come up with an average of ~7m higher, but given the data scarcity it is hardly surprising they give rather large uncertainties. And these 7m, being a global average, of course are only a mathematical reality; at any specific location the value may be drastically different. 

The difficulty with reconstructing sea level also makes its prediction difficult.We only have good data coverage over the last few decades, and only moderate coverage over the last few centuries. If your sea level model manages to hindcast the patterns observed in that time interval, that doesn't necessarily mean it can resolve anything novel happening, like, say, the collapse of the West Antarctic Ice sheet. And that's exactly the sort of things we would like to get a handle on.

So if you headlines in the newspaper, on Twitter or wherever, that say something along the line of “sea level in period X (say, the last Interglacial) was #m higher/lower than today”, do realise it could mean all sorts of things. If one person for instance claims sea level was likely to be 8m higher than today in the Last Interglacial, and you read somewhere else it was 8m lower, it doesn’t mean one of them has to be wrong, as much as climate skeptics would like that to be the case. In this case it concerns one global average, and a local record from the Netherlands. And it's many, very many of such results, which may seem contradictory, that altogether will paint a comprehensive picture of past sea level. And in the long run, this will hopefully give us a solid grasp on future sea levels.

No comments:

Post a Comment