 Thank you, Matt. We have Matt Palmer online continuing in some way the previous presentation by Karina and talking about, are you ready to talk now, Matt? Yeah, yes, fine. Just let me get the presentation up. Okay, yeah. So it's a presentation on ocean heat content and the earth energy balance inside from climate models. Thank you, Matt. Okay. Right. Can you see my screen? Okay. Yeah. Okay, great. So yeah, thanks. Thanks for the invitation to speak at the meeting. I'm sorry I can't be with you. It looks like a great meeting. So I'm gonna talk a bit about ocean heat content and earth energy and balance from some work we've been doing with climate models. So this is some work that I've done with colleagues, Chris Roberts, Doug McNeill at the mental office, and also Freya Gary, Roberto Fernandez, Bill Bowd, the University of Reading, and Jonathan Gregory. So the outline of the talk, I'm gonna give a bit of context and talk a little bit about global warming hiatus, which I'm sure is something that we're all pretty familiar with now. There's been a lot of papers written about this subject. And then gonna talk, there's all two main parts of the talk after that. So the first is looking at pre-industrial control simulations from CMP5 models and what they can tell us about variations of Earth's energy budget. And the second part is really about forced simulations. So this is mostly Freya Gary's work, but also some of the other literature that's out there. So I'll be looking at historical RCP 8.5 simulations. And then at the end, I'll summarize the findings and I'll sort of outline what I think are some of the open questions in this area. Okay, so the global warming hiatus is this stalling of the surface temperature rise. So if we look, hopefully you can see my pointer here. In this latter part of the record, so I'm sure you'll all be familiar with, the surface temperature hasn't been rising as quickly as we would expect based on climate model simulations. And if we look at a background warming rate to give us some idea of what the rate is, since about the 1970s, we've seen a warming rate that's about 0.2 Kelvin per decade. So that's what we might call a background warming rate. And that's a useful number to have in mind when we look at some of these model simulations later. And so in terms of the hiatus itself, there's basically two ways that we can bring about a sort of slow down in surface temperature rise. So the first is a change in the net radiation at the top of atmosphere. So this is a change in the rate of energy convergence within that system. And some of the recent published estimates put that about 0.6 watts per meter squared. And the other potential sort of causal factor for a hiatus on these sorts of time scales is a rearrangement of ocean heat content. So just rearranging the energy of the system. Okay, so moving on to the first part of the talk, the first result section. So this is all about pre industrial control simulations based on cement five models. So the approach that we're going to use is to assume a linear combination of forcings and internal variability within the real world. So we can analyze multi century cement five pre industrial controls simulations to address these questions. Okay, so the first thing we're going to ask is on what time scale does net to a come into balance with changes in total ocean heat content? The second question is what is the potential role of internal variability in net top of atmosphere radiation for the recent pools in surface warming? And then the third question is what's the potential of ocean heat redistribution for the recent pools in surface warming? Okay, so I just know that all the analyses that we present here are highly idealized studies. So we're not taking account of realistic observational sampling. In this work, we're just trying to get a first order look at the potential for rearrangement of heat and variation in net top of atmosphere radiation. Okay, so this is one of the key plots that was published in a paper that came out last year in environmental research letters. So this is looking at 10 year trends in surface temperature against the trend in total system energy. So the way this is plotted is the 10 year trend in earth system energy is on the y axis in what's per meter squared. And then the 10 year trend in global average surface temperature in kelton per decade is on the x axis. And what we can see is that while there's a positive correlation, in, in general, the trend in global surface temperature is a fairly weak, weak indicator of what's happening at the top of atmosphere. So you can see that in these quadrants, I'm pointing out here that they're actually of opposite sign. And I think on average that the number of decades that are of opposite sign is about one third of the distribution based on all the same five models. Okay, so conversely, if we look at full depth ocean heat content, we find this very robust relationship between the trends in 10 year earth system energy and the 10 year trend in a global ocean heat content. And this of course is because the global ocean is the earth systems primary heat store and it's dominant on these timescale. So if we were or if we are able to monitor global ocean heat content well, then we have a very robust indicator in terms of keeping track of synergy imbalance. Okay, so you could having got this result, you can ask the question, well, on what timescale does the ocean become the dominant term in the energy budget? And we attempt to quantify that here. So we're looking at the correlation between the trend in the earth system energy content and the trend in total ocean heat content as a function of trend length in months this time. So we've got a whole bunch of cement five models are plotted up here. And what you can see is for the vast majority of these models, the correlation starts to saturate out at about 12 months or so. So it looks like based on the cement five models, the ocean becomes the dominant term in this energy budget on a timescale of about 12 months. And this is relevant to some of the other papers that have been published in this area, such as the Lovatel study was published a few years ago in nature Geoscience, where they're trying to confront these series top of atmosphere satellite measurements, which track the changes in net to away and comparing that to changes in ocean heat content for consistency. Okay. So the other thing that you can look at. And this time is it's on a decade or timescale again, is the depth of the ocean that you need to sample in order to get the best estimate of the net to away. So we have, we have four panels here, but they're all showing the same thing. It's just different models in each panel with the total range of the model behaviors is indicated by the shaded regions. So without getting into too much detail, basically, what we're trying to show here is how the uncertainty in your estimate of the radiative imbalance decreases as a function of depth. So as you might expect, as you integrate over more of the vertical column, you're able to retrieve a better estimate of what the net top of atmosphere radiation is. So I guess the key thing to say about this is that the models vary quite a lot in, in their behavior and particularly the information content with depth changes quite a lot, you know, it's quite variable among models. So you might have to integrate deeper in some models than others to get a reliable estimate of net to away. Okay, so thinking back again to the height of specifically one of the things we wanted to do was to look at the the magnitude of trends in a bunch of different variables and as a function of trend length. So what we have on the y axis is the magnitude of the trends. So we're, we're trying to plot both sides of the distribution. So we have both positive and negative values. And then what we do is we extend the trend length in years along the x axis. And you can see this reduction in the magnitude of the trend as a function of trend length, which is sort of what we'd expect from, from internal variability. So when we look at surface temperature, which is a, you know, a very familiar variable, what we find is that time scales of about 10 years, it's very easy to accommodate a trend of a few tents of a Kelvin per decade. So it's, you know, as previous authors have found, it's fairly straightforward to eradicate the 0.2 Kelvin per decade trend sort of background warming trend just through just just by invoking internal variability alone. I should say that that these dotted lines are the maximum trend that we find for any model. So it's just to give the absolute upper bound on what we would find from any Z with five model within the archive. So this is the surface temperature in some ways as a result we've seen before. But what was very interesting and definitely I found very surprising when we did this work was the magnitude of the total energy trends that can arise just just through internal variability alone. So when we think about the height and some of the literature and some of the external forcing mechanisms which have been talked about generally in the order of about 0.1 maybe to 0.2 watts per meter squared. And we can find that on a 10 year time scale that actually you have a similar magnitude of changes in net to your wake and arise from internal variability alone. So it'd be very interesting to to understand further how this comes about. OK, so I'll move on to the next slide now. So this is all about the ability of ocean heat content to or the ability of the ocean to rearrange heat basically. And this again is expressed in watts per meter squared relative to a surface area. So if we look at the fluxes across the 100 meter isobath, then we have at decade old timescales we have a sort of an order of magnitude changes that can be supported about 0.1 to 0.2 watts per meter squared. So these are fairly big chunk of the 0.6 watts per meter squared that we that we understand as being the approximate magnitude of us energy imbalance sort of right now. So as we move to deeper levels again as we might expect our capacity to rearrange heat is somewhat reduced. So we have you know have a lesser we have we find lesser fluxes across this across 1800 meters. And this is chosen of course to be roughly commensurate with the depth of the Argo profiling floats. Okay so I'm just the next few slides really to just give you a qualitative feel for how the the character and magnitude of ocean heat rearrangement varies among the CMIT 5 models. So this is these are just hovmuller plots over 200 years. We have depth on the on the x-axis sorry the y-axis and we have time of 200 years on the on the x-axis. So you can see I mean if I just flip through some of these so these are the different models you'll see different patterns and you'll see different depths over where the variability seems to be active and you'll see different timescales of the vertical propagation of features. So I think really all I'm trying to say here is that the choice of CMIT 5 model really matters for the for the ocean variability and there's a lot of diversity out there. Okay so to summarize quickly the pre-industrial control simulations the ocean becomes a dominant term in the energy budget on the time scale about one year. The changes in net TOA associated with internal variability or border 0.1 to 0.3 watts per meter square squared on decadal timescales. This is large compared to what we think the current imbalances and the ocean heat rearrangement across the various isobas I discussed is substantial but it doesn't seem to be as important in these model simulations as changes in the net TOA itself and there's a slightly smaller magnitude. So compared to previous studies you know 10 to 15 year height is could could be explained by internal variability alone and there's we published a paper in Nature Climate Change this year which I'm going to briefly show some plots from but I think Matt Collins is going to talk about this in a lot more depth tomorrow and then surface temperature trends are not reliable indicator changes in a system energy and hence global warming on decadal timescales I've said something about the representation of internal variability varies considerably among the cement finals which is obviously something we should seek to understand better. Okay I'm really going to skim through this in the interest of time because I think Matt's going to talk in a lot more detail about this but basically cement finals show different flavors of height events but they all agree that the Pacific is a key region we we know that following a height is decade is a greater change of accelerated warming to do with the reversal of the trends that we see in the Pacific and that here's a couple of fingerprints showing a composite of it which is dominated by you know a PDO like pattern and to first order the the accelerated warming decade really looks just like a reversal of the cooling PDO type like pattern that we see for the height to seize and then I guess something that we've probably been I guess many of us have probably been thinking about a little bit is wondering about whether there's a role for for external forceings during the height is so and this is a slide from Roberto Bilbao and he's showing here the observed rate of sea level change based on the satellite altimeter relative to the global mean and the hatched reason regions are regions where the models so the trends are inconsistent with with unforced variability in at least two thirds of the sum of five models so clearly this might be indicative of a role for external forceings but equally it may be indicative of it might it may be indicative that the internal variability that's represented in the sum of five models is too small okay so I think I've been speaking about for about 15 minutes so I'll try and go through the next few slides fairly quickly so this is about looking to the future really and and also the sort of historical past so this is a slide from Freya Gary so this works been submitted to environmental research letters a few weeks ago I think and really we're we're looking here a bit of spatial patterns so this is based on the pre-industrial control runs and really we're looking at the from from left to right we're looking at different depth layers so the upper 700 meters then 700 2000 meters then 2 to 2000 to 4000 meters and it's just an idea of the magnitude of the variability and where that's happening and so you can see that as you start to get deeper particularly in this layer we see a lot of variability in the Atlantic and I guess for this deep deepest layer there's a lot of that while all the models show that the Atlantic is a key sector the gist the gist model gis gis e2r there's a typo there but you know shows that both the Pacific and the Atlantic could be important so this is looking at the variability I should say that the premise of this work is about monitoring the earth's energy imbalance through ocean heat content changes so the other thing that's important as well as the variability is the emergent signal that we're trying to trace so again as from a left to right we're going increasingly down through the water column and again the the Atlantic is emerging as a place where there's a lot of action as well as the southern ocean and this is some work from Celine Housset who's published nice papers looking at the Seaman 5 bottom water properties under our cp8.5 so for our purposes particularly the bottom temperature that we're interested in we see this and what was apparently a strong southern signal which seems to be spreading northward so this is reminiscent of some of the observational work by Perky and Johnson for example so there's some interesting parallels there okay I'm so I'm going to skip through the next slide this is just showing how the temperature varies as a function of model and and there are some differences but there's also some fairly common features particularly the northward spread from the southern from the southern ocean region and so this is going back to Prey Gary's work again so what this is doing these different colors are trying to show the bias in what's per meter squared that we have in trying to estimate what the true earth's energy imbalances based on the changes in ocean heat content observed to successively increasing depths and what we basically find so if we concentrate on this upper left hand panel for now it's still while measuring only the upper 700 meters may have been adequate for monitoring the change in earth's energy imbalance up to 2,000 or so as we move into the 21st century then we start to get it basically starts to become inadequate in that we get a large bias between between our estimate the other set energy imbalance and what we're able to monitor with ocean heat content change because the climate change signal is starting to spread into the deeper ocean and therefore we need to monitor more the deeper ocean in order to keep track of it and it's particularly the southern ocean and the Atlantic and southern oceans where we where where that signal emerges so we can we can gain quite a lot of the benefit of a deep Argo array by concentrating deep float deployments within those basins you know to start with to keep track of this energy imbalance and so the Atlantic in the southern ocean a key regions for ocean heat contemporary ability in the response to anthropogenic warming and the climate change signal does start to extend over the full depth of the water column and Argo like observations are adequate to monitor the earth's energy imbalance historically but we but we really do need deeper observations in order to do this over the 21st century and as I think we've seen previously that the patterns of magnitude of ocean heat content variability and change vary by seeing the climate model so very quickly these are some open questions how might realistic sampling of the ocean effect results presented here it's an important question from the monitoring point of view what is the relative importance of the deep ocean the ice-cut regions and the marginal seas that might be might touch on some of the work that Corino has presented and you know are there important missing processes such as our an Antarctic bottom water formation I think we're on the cusp where we're starting to look at a couple models where we have an eddy permitting models and it's interesting to think about how that might affect these results and whether there are other modes of variability which come into play there's some literature that you know is starting to come out about that and ultimately how do we determine which models are most like reality and I think you know that ultimately we need to aim for process-based constraints on the climate model projections because we want you know that's something we should be working towards so I'll leave it there and take any questions and thanks for your thanks for your attention