 Okay, so I'm going to move to much bigger spares, the climate system, to solve the drought. The code I'm going to show you is also publicly available, but it's millions and millions of lines, so it's not for the human part. And I want to talk about capturing paleoclimate changes, and particularly to fill in tropical Africa. Because in tropical Africa, we have times like this year, in the 1960s and 1980 when we had severe drought in the Sahara region, it was very devastating for the people of the Sahara. And then we have times like this in the past, 7,000 years ago, where we can reconstruct what the lakes look like, the lakeshore, which is relatively small now. It's huge, right? They look at shorelines, the rainfall finishes. So we have these contrasting periods here in terms of water, and water, rather than temperature, is what's going to be of concern in the future, too much or too little. So most of you are not familiar with the Sahara region. On the top there is a time series. It's the Sahara Precipitation since 1900. It's kind of the anomalies in precipitation. So here is a very marginal region. They get most of the precipitation from about June to September or October. And so they don't get precipitation in one year. It's hugely detrimental to the agriculture. And many of you get these time periods, like we had in the 70s and 80s, where they had boundaries of drought. It's really an extreme condition. There are indications. You can see a little bit on the precipitation record here that extends to point 13. But also in terms of photographs and satellite observations, that suggest that there's an increase in seasonal raininess since the 1980s. The cause is still debated, controversial. Some would still say, you know, they've improved their agricultural practices and that has led to some gradient. There is some evidence of that. But we've also had increase in greenhouse gases since then. And there's both the direct effects right on the main surface, but also the indirect warming and the sea surface temperatures nearby. So quite a lot of models led us to look at some of these effects here. In particular, there's a lot of emphasis on how precipitation will change in the future. And this will and really all of Africa. But of course we don't have data for the future. We do have sort of drainage to involve in the future climate change. So we can use the past to decide what's the perspective of how it might change in terms of looking at the mechanisms, evaluating our models. And we can use the past in particular to talk about kind of these racial, interracial changes from weather ingredient to dry weather please and what processes and forms that change. And also how abruptly could it change from a humid period to a dry period or vice versa. And so just to, so those of you who are not familiar with the geologic climate record, this shows the care in Africa. We're looking from 120,000 years ago to present. And we have records here off of main cores that measure transport both by the preview system but also the only one of particles, in this case single measuring grain size, of Western Africa. And so what they see as they look through this core is they have these periods around here where they have wetter conditions and then drier in between. And the wetter conditions correlate quite well with what we know about processional forcing of inflation changes over the northern hemisphere and in particular over North Africa. So that's when we have, oops, I shouldn't use this, their inflation in the summer months there, you had wetter conditions in Africa. But you can see there's a lot of noise there. There's a lot of else happening besides that. And if you look at a record here, this is a record of sea surface temperatures in the north of Antarctica, in this case off Portugal, that showed periods of warmer conditions or cooler conditions and in these kind of merrier conditions, these millennial type variations correspond to these colder events in the north of Nanter. So we get prior again those colder events. I just want to try to reproduce this. This is some results from a relatively simple model primer, it's an atmosphere ocean vegetation model, where they give it the orbital changes. And what they find is if you look at the central region here from about 10 to 20 north, they see these variations in the dash line here in precipitation that correspond to those increased summer installation over north of Africa. And they also see that they have in particular more trees growing on the other side of precipitation. So they have a green in the Sahara. Sahara doesn't quite look green, but it does get more western. So they're able to reproduce this pattern of a greener, whether Sahara, Sahara, just by giving the model these insulation changes. But we also, as I mentioned, have these periods where we see these drier periods on millennial type spheres. And that is also the model here, is another set of records of western Africa that's off the Senegal River mouth. They measure kind of the ratios of aluminum and iron to silica. And from that, it can come up with whether it's wetter or drier, or more humidity, going out before drier, going down. And what they find is when it's drier, it corresponds to what we call airic events, where we have ice-crafted green in the North Atlantic that suggests a lot of ice dirt, going out into the North Atlantic from the ice sheets of the North American mountain and cooling the North Atlantic and then making the region North Africa drier. I should comment, a lot of these figures geologists don't have time going the wrong way. So this is 60,000 years. On the right here, going to present on the left. I'm a primatologist by training, so when I show you some of the figures from our model, I'll point out that they don't activate. But in the case, I just look at the arrows here. So what the modelers have found is that when they add wet water to the North Atlantic, they cool the North Atlantic with shifts in the inter-topical convergence. So in the right topical range, both sides, you get wind here in these regions of North Africa with this chorus measuring. And just for those of you that don't speak the geography of geologists, particularly geologists, each one is most patient maximum about 21,000 years ago. And only seen is this tanker that started just after about 12,000 years ago. We're still in the Holistin. And most of what I'm going to show you now from our results are showing you results starting at the last place from last month. And then that was, so just for the doing of the Holistin. And in particular, I'm going to show you results that have to do with the first part of what's called the Africa-Humor Period. So it's a period, from about 15,000 years ago to about 5,000 years ago. And there's lots of evidence that Sahara, the Sahara, had more vegetation, the lakes were higher. And so that's what I'm showing you here in this next slide. Now we're zeroing in here from 20,000 to zero. And which of these figures are suggesting less rainfall and then higher rainfall as you go vertically? So if we look first at a record here, it's kind of an iconic record produced by Peter Domenico. It's a measure of dust in a marine core, just also Western Africa, where the scale here is reversed. So it's less dusty here during the African-Humor Period, more dusty before and after. And you can see the strong correlation again with the insulation. In this case, part of here was the June 21st insulation at 30 north. So this is the less summer insulation peak in here, so greater solar warming here of rural Africa at about 11 o'clock and then coming back to the present day. And what we have found that a number of our geologists have found are that also in the northern hemisphere suggest much better conditions. This is from the Kermak, they looked at lake levels at 6,000 years ago from during the African-Humor Period and we see always plus signs that say lakes were much higher than present. What is a puzzle that can be intriguing when I started working on this project with my colleagues who are geologists is we also see here higher lake levels here in the southern hemisphere. And in particular, this record here from Lake Tenderuta, which is this blue diamond here, also shows this African-Humor Period. So dry, and then about 15,000 years ago there was much weather and then dry out again. And this case, there's sediment cores in this lake and they measure deuterium of leaflets. So the vegetation that's growing around the lake. And this is a puzzle because what we know about the orbital cycles is it should be out of phase where this should be drier during this African-Humor Period because the northern and southern hemisphere are out of phase with this cycle. We can see this somewhat if you look at it. And now on time it's going left to right. Okay, so we can see this if you look at the insulation here going from the first place to maximum to 11,000 years ago. There's again a 20 north there's a 10 degrees south during this summer so like a less summer insulation. So we should have a weaker most of the south of the equator, stronger north of the equator. So what else is happening? This was part of a project we did called Trace where we simulated from 21,000 to present. And in addition to the orbital forcing we also have greenhouse stresses changing here. So the blue line here is carbon dioxide which increased from about 185,000,000 parts per million states like that to about 17,000 years ago and then rose up to almost present a 270 parts per million industrial. That then also with a few more levels increased from its local communities to its around 11,000 years ago. And then the final piece of forcing that one needs to think about is we had large ice sheets over North America and we had large ice sheets over in this area. So these ice sheets were no longer in the climate warned with that water going into the North Atlantic out this anode of air had more ice as some going into the southern. There's a lot of unknowns about that so this is just the scenario we used in our simulations we started adding that water at 19,000 years ago and quite a bit goes into the North Atlantic to about 15,000 and then again at this time period about 12,000 years ago so just a number there if we're going to go on. So the model we moved, I would say there's millions of lines of code but the vision we moved is I would say one and a half versions before what you'll hear about tomorrow from Jean-Claude from about the community climate system model three it has an atmosphere model with about a four degree wetland resolution we include a land model we include the vegetation as a full ocean model and a full sea ice model and with climate change to have all these interactive components to understand the system and I should say to one model from 22,000 to present took about 14,000,000 four hours on a super computer that the Department of Energy has done in Oak Ridge so it's a huge effort to do this to put the one on my record task and what I'm going to show you then are three simulations that we did one where we moved through with all these four things to see how well the model does and compare some good data and then to drill down a little bit on the individual four things two ones, one with orbital only and one with greenhouse gas only starting at 17,000 years ago so as I said we first have to see what we're comparing here to is from the second to the bottom panel is a wonder of proxy data that we have for Africa black and blue and our simulation in red and so first looking at some records of what was happening in North Africa the top one is the second panel is again that humidity index that I showed you earlier now just looking for the shorter time period and the bottom one here is the dust measurements from Peter Domenico and so we compare to precipitation simulated by the model in red or the total nuclear index in this third panel and you see the model does so well kind of dry conditions a little bit dry here about 18 to 15,000 years ago and then getting a roughly much more wet or more moisture available not about just after 15,000 years ago than starting wet we can also look at a like 10 to a nuclear record and also a like a shallow record in terms of what was happening in south-east equatorial Africa or the region called the African Great Lakes region and again you see that both the model and data suggest this kind of dry, getting a bit drier but then getting much wetter here right about 15,000 years ago and the third is just another record from the Congo recently and later on just kind of this is a measure of what the Atlantic ocean overturning circulation that the data is doing and you can see here that the response to that notewater that we put in the North Atlantic where basically that circulation becomes very sluggish and then we cover it right at the beginning of the North Atlantic period so to kind of look a little bit more broadly than just that data that we did is for a number of records compiled by geologists who have collaborated with in effect lake levels they have pollings that they retrieve some cores they have geochemical integers and they compiled those records where they had a record that was from 22,000 to 10,000 and those are powdered here and what we did is look at the empirical or empirical analysis pattern to try to look at the modes of variability and what are dominant modes of variability we used for the model for Canada we used for that we're not looking at the more annual trends we're looking longer temporal so what you see in both the data on the top and then the model result on the bottom is kind of this point African weather conditions in both over all of Africa except we'll be right here in kind of the southeast part of southern Africa but pretty much weather everywhere in Africa and if we look at the time series of those EOS patterns what you see is kind of this in both model and data this higher conditions here from 21 to 17,000 years and an increase in precipitation and moisture until we ran with kind of this small kind of retreat here in this time period we called when we got this so both are showing this kind of same pattern of the moisture increase and we wanted to kind of understand that better by zeroing in on two regions one kind of the northern Africa and one here with a great life stream and comparing kind of the drought period at 17,000 years ago to the weather at 11,000 years ago so what we find is a model result I should say and what's spotted here is the left panel with northern hemisphere summer north Africa most of the precipitation with the W increase in June to September and then on the right is southern hemisphere summer because again the monsoil nature of the rainfall means that most of the rainfall there occurs from about October to March and so what you see when we include all the portions is that we have this broad region of weather conditions for blue and green for 11,000 years ago and it's compared to 17,000 years ago but it stands across north Africa into India and during this southern hemisphere summer similarly it's much further by extending over all of Africa so we have increases in rainfall in both if we look at the bottom two panels where we only include the orbital variations now you can see what we kind of expected from what we know about the processional cycle in terms of north Africa you get this increase in precipitation that's moved there over the southern so there you're actually not only an increase but kind of a gradient of the rain belt while if we look in the southern hemisphere here we see actually decreases so the orbit in the southern hemisphere as we expected not enhance precipitation which disagrees with the data on the other hand if we look at our simulation where we have only the green house gases changing now we find that in both regions and now look at these middle panels here we get an increase in precipitation so it's actually the green house gases that are moving in this coherent response both north and south of the equator just looking at the time series rather than the two short shot differences what's shown here is the change in precipitation simulated by the model from twenty to eleven thousand years ago the green house was the orbital only one so you'll see that we're sitting here in North Africa orbital is starting to increase precipitation at about seventeen thousand years ago while again over the Great Lakes and the southeast equatorial we actually will change due to orbital if you look at the blue curve you can see that over the single region it also continues to be about forty percent of the increase but it's the only thing that's causing the increase in southeast equatorial Africa but what's interesting one is to look at the red curve which is all the forces and in that case the orbital green house in the Sierra green house can explain kind of why we haven't got the conhuman period the answer to that really is explained by something else that is a very important increase in both regions and so we can explain what if we look at the second mode of variability in the buildup and in the models today there's a little missing but you can kind of see here kind of how there is between what's happening here in North Africa and what's happening here over the Great Lakes region it's much more easily visible in the model data as I like to tell my data things if you give me four dimensional data at space points may not be correct at the time though of course the data also has some interpretation but in any case what you see here in the model if we look at the second mode of variability we have these coherent changes between northern Africa and the Great Lakes region here but we have the out of these relationships as we look across the shoulder along the west coast of tropical Africa okay at the time series then and we can see that this pattern where we have these positive and negative we have go to the opposite sign here at about 17,000 years ago and then this record recovery and then again at about 12,000 years ago and I'll just tell you in the interest of time that this is correlated to the notewater in the north Atlantic reddit and the slowdown of that thermal haemian circulation so that when it slows down like in previous modeling results we find the shift of the tropical rainbow south so we get drying conditions north and reddit south and then when it restarts it's just that and we find that the coherence that we get actually with the atmosphere and that motion is too slow to transport that similar one with the shape of the form there so the atmosphere that transmits that signal across Europe and leads to this coherence in this response which I won't show you here in the interest of time so you just have to take my record so I just want to finish then with kind of we're excited about the results but we're not perfect like most models what we find is that although we get the meaning of the Sahara up here we don't get as far north as the data people would like so this shows some reconstructions of how much precipitation will feel increased at 6,000 years ago so about 300, 400, 500 millimeters per year and not only our model but none of the models that have participated in the IPCC assessment reports I can correct it that what meaning or that what meaning it was so are there lines at this for feedback that may be missing or not representing well um vegetation most of the models have a fairly simple kind of climate envelope type vegetation scheme but the newest model in CESM is going to try to look at the ecosystem demography so it may be that vegetation if you cover it also for albedo for transpiration so for models that suggest that if you can get enough vegetation that'll be a positive feedback on precipitation so can I say also the more the amount of precipitation, evaporation transpiration almost all models still keep soil present which might be okay well to 2100, 2200 but when we're talking about past changes where you've had the vegetation change you'd expect the soil composition to change not so much a a feedback that may be a reason why we don't much the lake records is an extensive study where there were 1500 paleo hydrology records suggest that a lot of the increase in the lake started about 9000 years ago and there's a time lag until they fill up so including a good groundwater hydrology scheme could be important and dust but just starting to improve that but with a more vegetative surface you'd expect greater soil moisture to gradually decrease dust production there's both positive and negative radiative feedbacks but what they have found for modeling in the recent 50 years is that if they don't include dust they cannot reproduce the rainfall over northern Africa so just to end up my last slide is what I'm showing here is a plot from the last IPCC it's a precipitation stored by global temperature so it's percent per degree change in global mean temperature by all models out at the end of this century and what you can see is a panel that looks remarkably familiar to what we see during the deglacial period where you have weather conditions here predicted by the persistent models in northern Africa into a red rough and then down here into the the great lakes changes so we found that these greenhouse gases are important to do our pattern during the deglaciation it also seems to be the pattern that the models want to produce for the future and then just that we found that the onset of the effort from whom it was synchronized to what was happening in the north of Africa that can bring sociations so there is a possibility that climate change is in the ocean full of structure but that overturning circulation could change so probably not as dramatically just a few caveats in terms of future nexus in terms of the future projections they're learning this land cover changes they can change what the models are projecting into the future and there are changes and there are cells in that that are still unknown and the results I showed were from one model but we're excited that it worked that so well so thank you and we're going to move into further back to that so I think we should be aware that this is here for us to be able to monitor the data so the data has very different resolutions but we try to come up with a tragedy about the same resolution on both I'm sorry but we have one more one more do you think if you started to tell me that there was a few caveats there were a few caveats that you were able to take a hold of and push this back well actually a part of doing this South America the problem isn't as much data but I think we're doing more and more we have a really fantastic team data the other the other part is South America is moving more difficult it's got the huge andeans there there's a fairly well cut one to the bottom but even in South Africa we didn't trust on himself but there has been some work looking at that and with this simulation if you're interested I can give you the reference on that it is important to have your faces important Mr. Bradley and