Marine Frouin

The contribution of luminescence dating to the chronology of Pleistocene deposits in Turkana, Kenya

Past advances in radiometric and relative dating techniques have fundamentally changed our capacity to piece together our evolutionary past over millions of years. Marine Frouin's research is focused on the development and application of luminescence dating techniques for the study of Human evolution.

FULL TRANSCRIPT 

Hi everybody, my name is Marine Frouin. I believe I don't have to convince the room today on why chronology matters. So, let's go to the core of our issue. When we look at the Turkana and this particularly the Pleistocene, it's where all the main gaps we have in chronology. Yeah, so we have this large gap in the chronology and my mission is to fill these gaps using luminescence dating and to today I'm going to show you two examples. One that is a bit more conventional and traditional and the second one is much more cutting edge. These results haven't been published so far, so you have the pleasure to discover them today exclusively. Yeah. Before the start going into the field, I just want to give you a brief introduction of why luminescence and how it works. So, I'm sorry for those of you who already know how the method works, but I believe probably half of this room doesn't particularly maybe our students. 

So here it goes. Please bear with me. Okay, so when we look at the Pleistocene, we have a large amount of method that we can use for dating. Probably most of you know that radiocarbon dating, but those methods are very specific to the material that we found. Radiocarbon dating is specific to mostly organic matter and can go back in time around 50,000 years. You also have Argon/Argon dating techniques. All of you are familiar with this method, but it is very specific to a volcanic matter that doesn't always preserve in the landscape. Probably some of you are familiar with thermoluminescence dating technique, which is specific to burnt rock or flints and those are not that common in all sites. In comparison, quartz and feldspar are the two most common mineral on earth. So, luminescence dating and what the technique that we call optically stimulated luminescence dating techniques can be applied to a very large range of application and context. Here in this little cartoon, I just summarize, or I just show the range of application where luminescence can be applied. What we're dating is the last exposure to light or the burial time. 

My interest and the specialty of the lab here at Stony Brook is to use luminescence dating to archaeological sites. How it works, okay, bear with me here. That's the only technical slide that I have. Okay, so here we're in the environment these little circles here represent grains of quarts and feldspar when they get exposed to lights through transport, during transports, so by wind and water they get what we call resets, so they get bleached. It means that all the charges inside get ejected. When these grains get buried, they receive effective radiation coming from the decay of fragile elements such as potassium, uranium, thorium, contained in the soils, as well as the contribution from the cosmic radiation. So, as you can see on this little cartoon with depth and with time, you have more and more charges that get built inside these grains. Another way to think about it is to consider grains of sand (quartz and feldspar). Again, as they act like a battery, so with time, you have more and more charges that get built in this battery and the aim of the game for us is to determine how much charge has been accumulated inside the grains.

So, we have to measure two things in the lab, something that we call dose, which is the number of charges the grains have accumulated over time, and the dose rate, which is something related to the radiation of the soil. By simply divided the dose by the dose rate, we obtain the burial time, which is the date we're obtaining with luminescence dating. Okay, so the method is not new. It has been around since the 70s and it has been applied to many different environments. We can date quartz up to about 2 to 300,000 years from very different context. We can also date feldspar much further back in time around 500,000 years. We only have one date here at 700,000 years. I'm not going to explain you exactly the detail of all these different protocols, but yeah, now we have a lot of different signals that we can play with that are extremely useful for us to understand the speed of bleaching or the deposition of the sediment. In both cases for quartz and feldspar the age that we obtained have uncertainty that is around 5 to 10%. 

Okay, done with the technical part. Now let's go to the field. This project here is our first case study came from a discussion that I had with Fred Grine a lot who has been working on this skull. Some of you might know this skull as the Eliye Springs skull that was found in the western part of Turkana, actually 20 kilometers away from the village of Eliye Springs closer to the small camp called Lobolo. And this skull was discovered in 1985 by a tourist who was basically walking along the beach and stepped on the skull and find a skull. So, he was not finding any geological context also in this area, they are no dateable tuff. 

Based on the morphology, it has been described and hypothesized that the date of the skull may be somewhere around a 100 to 300,000 years. We also think that it might be coming from a nearby late Pleistocene deposits, but again, we didn't have any dating in the area, so it was difficult to link the skull to these deposits. In 2014, Fred Grine had the opportunity to collect sediments that were inside the skull, and he sent a sample for dating to the lab at Oxfords led by Richard Bailey who tried to date the sediments using the quartz. Unfortunately, all these quartz were saturated, so we on the dates that he obtained was not reliable. Another interesting study is done by a Professor Raspbury here at Stony Brook University who analyzed the sediment and determined that based on this radiogenic isotopes we could potentially trace these specific fingerprints of the skull back to the sequence. So that's why he motivated our field work. And in 2022 we obtained this TBI exploratory grant to do some field work near the site of Lobolo and collect samples. We went there with this first, this ground penetrating radar, which I would not recommend you guys to actually take in the field. It's a huge instrument we had. It was interesting to fly with that. 

So, we use this technique to determine, to try to understand the stratigraphy and the tectonic in the area. And this work has been currently interpreting, has been currently interpreted by Professor Dan Davis and Taylor Grandfield, one of my student in lab. The radiogenic isotope study will help us understand the ecology and the provenance of the sediment. That is the work done by Professor Raspbury and Katie Wooton. I'm not going to talk about this today unfortunately. I'm going to focus more on the luminescence dating parts. The mineralogy and luminiscence dating parts are part of the work done by one of my students Leland Resifield here in the audience. Here in this map, I'm just reporting the sites that we sample for luminescence dating. So, you can see where the skull was originally found. Now the shoreline of the lake is much more inland, so it was not possible to go back and collect samples directly from there, but we were lucky enough to find stratigraphy much further north. that is very promising. And we also described too the stratigraphies and collected samples from these two area. So, the first area is right by the shore of the lake here so you can see the elevation between the different sites compared with the current lake level. And so, you can see that all the beds are tilted toward the west and the succession of mudstone sand where you have some ripple beds here, some cross bed, there's a lot of things happening while you can really see it very clearly here, all the cross bedded sand. 

And on this stratigraphy here you can see where we took the luminescence samples and on the right side where we took the radiogenic isotope samples. Let's move on to the second site, which is further up here in the elevation. Again, we have a nice stratigraphy that took us a few hours to dig, but we found that it's nice stratigraphy that ripple succession of clay kind of sediment and then sandy, clay, sandy sediment. Again, you can see very clearly all these cross beds and ripple marks toward the top of the stratigraphy that indicates fast transgression and regression of the lake. Last one, the last site is the one that is event further up in the sequence up in elevation right here. And it records, again, more like horizontal beds and then followed by units where you have much more ripple marks. Okay, so our approach was to use both quartz and feldspar because they have different properties, especially we can use different signals from the feldspar. They're extremely helpful to understand how fast the deposition may have occurred in the past.

So, by example, if you're dating a sand dune where all the grains have been well exposed to lights during transportation, then you would expect the quartz age to correspond to be exactly the same as the feldspar ages with these two different protocols. While if you are more into a mud flow kind of system where all the grain may not have seen the light, at the same time you expect the quartz age to look much younger than the feldspar age this is at two different temperature. 

So here are some results. This is an ongoing work, so we have many more samples to date, but here is just to give you a brief overview of what the age of these deposits are and we're pretty confident that the age of all these sediments are all ranging around 80,000 years. Fortunately, the quartz was also saturated, so only a minimum age of around 70,000 was obtained. So far on these samples here, if you look at the these two ages obtained by feldspar using these two different protocol, you see that here they're pretty consistent within the range of error at two Sigma while here in this area, they are less consistent, which makes sense because the type of sediment are here, it's much more muddy and as I said before, we expect a less bleach sample here when it comes to mud. 

Okay, so just to summarize, what we've discovered so far is like all these luminescence ages indicate, very fast deposition probably very quickly in around 80,000 years ago and intense regression and transgression during the late Pleistocene. And we obviously have a fault somewhere that we probably be able to find using GPR. Evidence also of a tilted fault block by the shore and if I want to push the reasoning a bit further, I think that if these faults are still happening, perhaps it's still moving sediments toward the lake, and I think it's something that we need to consider when we think about the water level that is raising from the lake. We need to also bear in mind that we may have had some sediment coming into the lake due to tectonic. So, we have to do more work with understanding the tectonic around the lake Turkana over time. Too early to tell if we'll be able to trace where the skull is coming from. But lucky if we would, maybe one day.

Okay, for this second case study, I just want to take a few minutes to talk about this other dating method that we are developing. And I worked a lot on this technique during my PhD and further after that that is called infrared-radiofluorescence dating, also applying on potassium-based feldspar. In this technique we are measuring the signal using a radiation, not a light, different from OSL. And just for you to understand the concept, I made this little graph and remember before I said the grains can act like batteries and with optically stimulatedluminescence we look at how much charge are in the battery. In the case of radiofluorescence we're looking at how empty the battery is and I'm not going to explain all the physics behind it but behind it. But somehow it allows us to date further back in time than traditional dating methods. 

We have used this technique on lot of different sites have used it on this site that you guys may be familiar with at Shanidar and Iraq. I use it also on site in Saudi Arabia where other luminescence dating technique couldn't work and we managed to get good results using infrared-radiofluorescence. But really what I'm interested in is to try to unlock the true potential of this dating technique. And to do that I'm using samples from the Koobi Fora formation. Again, I probably don't have to convince you how important the Koobi Fora formation is; Exceptional fossil records, chronological range up to 4 million years, there have been more than 20 volcanic tuffs that have been dated by Argon/Argon in the past. So great, great sequence to work on and test our methods. Here, I really like this illustration that I borrow from National Geographic that the thing illustrate very well how fossils are usually dated in Turkana. 

So, by comparing where the tuffs are coming from and you clearly see that in some cases it might be a bit difficult, a bit tricky to obtain a good date if you have only one tuff and also, you're lucky if you at least have one tuff. So, wouldn't it be nice to have a technique that could allow you to directly date the sediment in which the fossils were found right? That's where I'm interested. In 2018, I went to Koobi Fora, I collected a lot of samples from the sediment in between tuff deposits and here I'm showing you just a few results that I got from this dating methods because I'm running out of time. The technique yeah, we were very happy with this result because it’s very current with the dates that were previously obtained on the Silbo tuff and we also were able to collect to get dates from the sediments underneath the Kale tuff, which we can discuss that later. 

Okay, here at Stony Brook, we are only 20 minutes away from this great facility, the National Synchrotron Light Source at Brookhaven National Laboratory where have been using the oldest sample of Koobi Fora formation to try to understand how we can extend further back in time the dating technique. And we found that the shape of the radiofluorescence signal is related to the mineralogy of the grains and has a direct impact on the age limit. And this has been published by one of my postdoc, it's currently under review. But what we are extremely excited about is in next month we are going to use a new detector or install a new detector, a beamline to be able to do dating and elemental mapping at the same time at the resolution that was never studied before that is less than a micron. Here I don’t know if you can see for you, view who are back in the room. But the samples here are this tiny grains of sand that are stick on this little tape. And every single time I go to the synchrotron with my samples, only a few grains of sand and I feel that it's incredible. That's what an incredible journey these grains have been through. Some of them are like 4 million year old and now we're using them to establish next generation of luminescence dating technique. And I really like this idea and yeah, makes me kind of proud. Thank you very much for your attention.

View all Geology and Geochronology videos