 Okay. Thank you very much. I hope I think everybody can hear me. I got a microphone here. And what I'm going to do is try to provide a long-term perspective on decadal and even centennial timescale of climate variability from long, in this case, tree ring records. Now, just to preface, I'll go through this quickly. You know, we all know this from the AR4 figure. You know, the fact that there seems to be a clear divergence between the modeled greenhouse gas foreseeing effect on climate and what the background natural variability indicates. But as we all know, there's a lot of decadal to centennial variability in what I call, I'll be calling DCV in the paleo-climate data that may not be accounted for adequately by the climate models. And one can easily see that in the, this is again from AR4, but I don't think things have changed significantly since then. You know, the amount of uncertainty and amount of decadal variability there is quite evident. And it appears to be significantly different from what the climate models seem to be able to produce based on our understanding of 20th century climate variability and forcing. And over the past, millennium for a long time, solar and volcanic forcing was going to explain a lot of the natural variability at decadal timescales. And you can see here an example of model forcing by both volcanic effects and also a solar radiative forcing effects. And they seem to be able to reproduce to some degree the timescales of variability that we see in the actual proxy records of past climate variability. But there's an alternative explanation for all this too that has been proposed. And it's kind of an Occam's razor approach which says that basically a lot of the decadal timescale variability is really just an expression of kind of a red noise type or low order armor model type forcing of the kind that Hasselman argued for back in the 1970s. And it still casts the climate models papers. And Carl Wunch has argued that in fact this for practical purposes is all you need to understand the origin of observed climate variability in the climate record at the decadal timescale. And from a paper of his, there is a simple synthetic time series, effectively a random walk that he shows has this long timescale variability. No question about that. And the question is, is this sufficient to explain what we see in the observed data and also in the long paleo records? And I'm kind of calling this argument of Wunch's, the null hypothesis that basically everything is just all the decadal time variability, DCV variability is being driven by largely by red noise or lower order armor models of type persistent feedback models. And this is enough. But what about the alternative, alternate hypothesis and to get into this, I'm showing, first of all, an interesting figure from a paper published in 1976 by Murray Mitchell, a research climatologist with NOAA for many years. And back then, all that was thought about was that, you know, decadal and centennial timescale variability was largely driven by the solar, lunar and volcanic forcing and so on like that. And notice that there's no mention at all of climate modes. And so, NAO, AMO, that sort of thing. But since that time, that perception has changed considerably. And now there's another spectrum as found on a NOAA website. And we have the climate mode shown here. And now we don't see any solar volcanic forcing indicated. There is an important contributor to decadal timescale variability. I suspect my personal opinion is that there still is, we pretty much know that there's volcanic forcing going on at times of major volcanic eruptions. And the solar forcing side is still pretty controversial, I think, but I'm convinced that there's some element of that there too in the natural climate record. But there's been quite a change in what we think about the causes of natural climate variability away from solar forcing in particular. And it's been replaced now by these kind of natural internal modes that are indicated here. And the timescales you can see are of the kind that we have to consider here for this meeting. Some years ago, Mike Bann and Jeff Park published an interesting paper. This is actually in the reviews of geophysics, advances in geophysics paper, big monograph. And in which they use MTMSVD methods for looking for phase coherent, band limited signals in climate data, in temperature data series in particular. And what they found was that they could identify, in a geographically clustered sense, signals that were consistent with perhaps the AMO, the PDO, solar and ENSO. And this is all based on instrumental data back to the 1890s. But I thought that it was interesting because it shows that first of all, the instrumental record does in fact seem to have these statistically significant band limited signals. And that they're also very important. They tend to be clustered geographically to certain regions of importance. But the North Atlantic sector, as you can see, is very prominently shown here, as much because there's probably much more data in that area too. So just some comments now. I think there's little doubt that there's some form of band limited DCV in instrumental climate records, by lots of different analyses. And the sources of this oscillatory variability is not always clear, but ENSO, NAO, PDO, AMO, et cetera, is highly likely. And maybe solar too, but that's much more unclear as to how much variability could be due to solar. And there's pretty much no doubt that these records of these past instrumental temperature show some effective volcanic forcing at this time. The question now, is there an apparent oscillatory variability just a short term expression of stochastic as in red noise of lower order or armor variability? That's mostly a chance phenomenon showing up in the 20th century. And I call this the null hypothesis, something akin to what Carl Wunch was arguing for. Versus the question is, is there an apparent oscillatory variability indicative of long term band limited forcing in the system that we might be able to tell from longer records of climate from tree rings in particular? The instrumental records are very difficult to use to identify robustly the presence of even PDO timescale variability which is on the order of 20 or 40 years. And for AMO variability, we basically only have maybe one and a half realizations of that timescale variability in the instrumental records. So we don't really have much of a chance to test the existence of something like the AMO very robustly at all in the instrumental records. So this puts us in an uncomfortable situation described by Wright in 1971 as two rules of climate change and published in Weather. And the first rule is some feature of the atmosphere can always be found to oscillate in accordance with your hypotheses. And shortly thereafter its discovery, the oscillation will disappear. And I thought this was a brilliant description of the conundrum we're facing is that the instrumental record really isn't sufficient to tell us how important these kind of oscillations of oscillatory modes have been over time because the records are too short. So to investigate the longer term existence of oscillatory DCV climate modes, we need longer records and I'm going to be going into that now. And we'll be looking at tree ring reconstructed climate modes and so PDO, NAO, AMO. The null hypothesis as before is that the band limited variability exists in the observations by chance due to internal stochastic forcing and should not be expected to have persisted over time. If it disappears in this power spectrum, wavelet spectrum of these long reconstructions, we might suppose that this is correct, that the null hypothesis can be accepted. The alternate is the observed band limited variability persists back in time in the reconstruction ways that suggest some form of intrinsic or extrinsic quasi periodic forcing that is not easily explained by a simple stochastic forcing. And this would be, if we found this to be the case, this would tend to support this alternate hypothesis. I'm going to be going through some examples now. Six different reconstructions of climate modes produced by tree rings. And that's a reconstruction, two PDO reconstructions, an NAO reconstruction, an AMO reconstruction, and then one for the northern angular mode, the NAM. And there are many more of these out there and I don't claim that this is an exhaustive set of tests, but so the list is just illustrative. Methods of analysis, I'll use multi taper spectral analysis and wavelet methods to describe the band limited properties of the data. And the significance tests of the spectral peaks will be based on a simple red noise and all continuum model. The superior legion and consistency of MPM is the reason why I'm using this method. And the wavelet spectrum torrents and compo will reveal the degree to which the band limited variability indicated perhaps in the MTM spectrum carries back over time in a fairly consistent way. Here's a reconstruction of the DJF Nino 3.4 SSTs, reconstructing tree rings, goes back to 1300. And the spectrum of the instrumental data is right here and here's the MTM spectrum and here's the wavelet spectrum of the same series. And then the same MTM spectrum and wavelet spectrum for the full reconstruction. And you can see that, you know, this reconstruction is weakly interdicated with very little evidence for any sort of longer term molded to cadill variability. This is not necessarily meaning that it isn't there, but the reconstruction just doesn't show it in this particular case. Here with the, I mean not the reconstruction but the instrumental data aren't long enough to really show it. The wavelet power spectrum for the reconstruction however does show some fairly consistent presence of decadal time scale variability. Not in a very consistent sense in terms of always being significant back in time, but it comes and goes intermittently. Here's a PDO reconstruction back over a thousand years. And in this particular case, we do see in the instrumental data strong evidence for molded to cadill variability and that's to be expected. That was really an understood property of the PDO and you can see the MTM spectrum here showing that too, centered around 40 years, the main peak. The reconstruction shows very consistent long time scale variability. At the molded to cadill to centennial time scale and I freely admit that a lot of this power right here is in the cone of influence which means it's less reliably estimated in the way of the spectrum, but it's still there nonetheless and there's still evidence even in the 60 year time scale roughly speaking for this kind of variability to be present. Another PDO index reconstruction back to 1500 in this case using different data I should say shows more or less the same kinds of features as the previous example with clear evidence for molded to cadill variability in the way of the spectrum of the instrumental data and the same also for the reconstruction in the way of the spectrum of the reconstruction. So there seems to be some reasonably consistent presence of this time scale variability in both PDO reconstructions. December through March NAO reconstruction back to 1400 and this is a case of since this is an atmospheric index one might expect that there wouldn't be a lot of long time scale variability in either the instrumental or the reconstruction. In this particular case we do see some evidence from molded to cadill and in the power spectrum and the way of the spectrum of the instrumental data a lot less molded to cadill variability indicated in the reconstruction than we had seen in the other climate mode reconstructions and again this might be a reflection of the fact that this is an atmospheric index reconstruction not one that's directly tied to oceanic influences in a really strong sense. Okay. Yep. Okay. Finally I'll go through this pretty quickly. This is the AMO thing and once again we see strong evidence for molded to cadill to centennial time scale variability and I'm going to skip over a couple more of these and go very quickly because I'm running out of time. Just torn and trashed temperature reconstruction back 1500 years once again very strong evidence for molded to cadill to centennial time scale variability. Same thing for a Gulf of Alaska summer temperature reconstruction from tree rings going back over a thousand years and so I think what this is all pointing to is that the spectral analysis do show evidence for some clear presence of band limited to cadill to centennial time scale variability that would seem to eliminate rights to rules of climate change as being true. In this particular case the oscillations appear to come and go in time and what you might call an amplitude modulated sense but they never seem to completely go away. The preferred DCV band is difficult to define but it certainly occurs in the 30 to 100 year band which is suggestive of PDO and AMO type forcing that variability of centennial time scale is present but it's less robustly estimated because the series aren't long enough to really do a very good job of that. So the power specter indicates that there's probably a mixture of both band limited forcing and some sort of red noise, simple red noise forcing or even lower order armor forcing that can explain what we're seeing. Real quickly now let's go through one final thing showing the way that we might have a way forward and try to get a better idea about this kind of time scale variability through the evaluation of drought atlases that I've been working on over the years. If we look at the average reconstruction of drought and free rings in these boxes here and just smooth them with 30 year low pass filters, we note that there's a lot of interest in anti-phasing that goes on on 30 year time scales that is suggestive of some large scale organizing principle that's driving hydroclimatic variability in preferred ways over these northern hemisphere land areas. And an example of this can be shown right here in this reconstruction of drought in these three areas in the low and great drought period 1276, 1297 dry here, wet here which almost certainly means that the needy like condition was ongoing at the time. And then interestingly enough this pattern of dry, wet, which suggests kind of a negative NAO state. So I think the development of drought atlases will help us, will lead us to an improved understanding of the causes of a large time scale. I mean large regional patterns of decadal and centennial time scale variability in ways that will hopefully lead to improvements in our ability to both model and predict that kind of variability in the future. And with that I'm finished. Thank you.