Instead of giving my talk at the Nordic Winter Meeting two days ago, I was in bed with a cold, fever and almost no voice … that’s how it goes with all these winter viruses swirling around! Pretty disappointing though, given all the nice talks by colleagues I had planned to listen to, and especially those that focused on the subject whether Sweden and/or Fennoscandia were ice free or ice covered between 60000 and 30000 years ago (sometimes called Marine Isotope Stage 3 or MIS 3). For some of you who read my blog, this topic might seem rather boring; but for those who work with this time period, it is really interesting, because it teaches us about ice sheet dynamics and environmental conditions that were pretty different from present day. So, as promised a few days ago, here is what I had thought I would talk about.
We are still faced with contrasting ideas when it comes to the build up and dynamics of the Scandinavian ice sheet. On one side we have the picture shown in the Geological Atlas of Sweden and in its online version: an ice sheet that advanced and attained its maximum position at around 60000-50000 years ago, and stayed there more or less until it started to melt some 20000-18000 years ago. On the other side we have some evidence from Norway, Denmark, Sweden and Finland, that the ice margin was more dynamic, that ice free intervals alternated with ice advances along the Norwegian coast and that larger parts of Sweden and Finland were not ice covered until about 30000-20000 years ago. Moreover, the famous Greenland ice core records show that this whole time interval between 60000 and 30000 years ago was characterised by very abrupt fluctuations between colder and warmer conditions – and the question is should these distinct cold/warm shifts and prolonged warmer periods seen in the Greenland record, not have had an influence on the dynamics of the Scandinavian ice sheet?
Long records that would extend as far back as 60000 years are very scarce in Scandinavia, because much of what had once been there has been eroded by the last ice advance. But, over the years researchers have found a number of sites, where such deposits may still exist. Unfortunately they did not believe their radiocarbon dates and did thus argue for a much older age of these sediments and as such for an ice cover during much of the past 50000 years (lesson one: don’t trust what everyone else thinks is true!). To see if there could be any clues that these radiocarbon dates could be correct, I compiled a few years ago data from all sites (see the yellow and red diamonds in the figure), evaluated the dates carefully and finally chose only those radiocarbon dates that could possibly be accepted. It should be noted here that many of the sites that had been studied were investigated some 20-50 years ago when radiocarbon dating was not as advanced as it is today. Therefore the dates really had to be checked carefully. Still rather many red diamonds remained and these suggest that at least some of the radiocarbon dates may be correct.
In a next step I only focused on Sweden and plotted the acceptable dates from different sites against age and latitude. The left hand plot shows calibrated radiocarbon ages on organic matter from different sites and the right hand plot those derived from mammoth remains (molars, tusks). It is quite obvious that the radiocarbon dates from northern and central Sweden range between 35000 and 60000 years (35-60 cal kyr BP), while those from southern Sweden seem younger. As for the mammoth dates – these are all over the place and would suggest that there was enough of a rich herbal steppe between 65 and 55 degrees North to feed these beasts. Of course now I can already hear people saying “but radiocarbon dates on bones are difficult, and should not be trusted”. But in this case, most remains had been re-dated and the laboratory procedures were up to date (see for example Ukkonen et al. 2007).
A means of checking how valid the radiocarbon dates might be is to compare them with OSL dates on sediments from central and southern Sweden. Of course someone might say now “OSL is difficult, .. incomplete bleaching,…, quartz, feldspar … …”, yes of course it is, but so is everything else we do! In any case and despite all the arguments, there is a visual fit between OSL dates and radiocarbon dates on organic matter and to some extent also between OSL dates and mammoth dates.
Recently Ukkonen et al. (2011) compiled all good mammoth dates from northern Europe in an article in Quaternary Science Reviews and showed that mammoth ages for Sweden are between >40 and 30 ka (thousand years). Those for Sweden are similar to those I had compiled, although Ukkonen et al. (2011) included one more date that I had discarded. In any case it does show that mammoths seem to have roamed around in an area, where the Geological Atlas of Sweden places a thick ice sheet.
My little sketch in the next picture reduces this thick ice sheet pretty much and suggests that, given the dates on organic matter from sediments and on mammoth remains, an ice advance into northern Sweden did not occur until around 35 ka (thousand years) ago and that the ice sheet did not reach southern Sweden until 25 ka ago. This picture has however a problem: there was pretty much ice on the Norwegian side, and in Denmark two ice advances had been noted, and these coincide with the assumed ice free conditions …
Within the last years, the idea of ice-free conditions between 60000 and 30000 years ago has actually been revisited in a number of studies, and this has led to new and really interesting results. Sokli in northern Finland is one of these sites, where Karin Helmens and co-workers could reconstruct arctic tundra between 54-42 ka. Further to the south, Martina Hättestrand revised the pollen stratigraphy of Riipiharju and suggested ice-free conditions during the approximately same time interval. In central Sweden, a new date on muskox remains gave an age of 49-46 ka (Bennike & Liljegren), organic sediments at Pilgrimstad an age of 56-39 ka (Wohlfarth et al.), and sediments in Idre an age of 54-41 ka (Möller et al.). In the southern Baltic, Johanna Anjar and colleagues showed ice-free conditions between 42-26 ka ago. All in all, everything points to not much or no ice! Detailed studies of cave deposits along the Norwegian coast on the other hand, suggested that ice-free intervals and ice cover alternated (Mangerud et al. 2010).
So how can one merge all these new findings? When I plot the new and most recent results together and along side the Greenland oxygen isotope record, then ice remains in the Norwegian mountains (how much?), but large parts of Fennoscandia remain ice-free until 45-40 ka. Then the ice sheet may have gradually built up and advanced through central Sweden, reaching southern Sweden by 25 ka. This scenario leaves us with an ice sheet that gradually advances into the lowlands, but does not show an ice sheet that likely was much more dynamic. Maybe the till deposits of the Ristinge and Klintholm advance actually show that the ice sheet was more dynamic? The problem is just that these advances occur, when the other records assume ice-free conditions!
In my next sketch I added the mammoth dates to complicate the picture a little bit more. We still have the same set of advance and retreat phases along the Norwegian coast, but the ice sheet of Sweden has shrunk pretty much and apart from a slight advance into northern Sweden not much seems to have happened until 30-25 ka ago, when the ice sheet advanced over central and southern Sweden.
The take home message is that the large ice sheet needs to be reduced; that most or all of Sweden was ice free during the early part of Marine Isotope Stage 3; that the ice sheet may only have advanced into central and southern Sweden around 30 ka; that the ice sheet was probably much more dynamic, but that the current age control on organic/mineral deposits is too coarse to resolve stadial/interstadial fluctuations. Finally that even Mikael Houmark has revised the Klintholm and Ristinge advances and that they don’t sit anymore where I had placed them in my pictures. Things move quickly though and hopefully in a few years we will know so much more about this enigmatic Marine Isotope Stage 3!