FULL INTERVIEW - Dr David Evans on CSIRO's climate models

[00:00:00] Malcolm Roberts: hi i'm senator malcolm roberts and i'm in brisbane but i'm interviewing dr david evans and i'll tell you more about david in a minute he's in perth western australia and many of you will understand or know of the late professor bob carter who was looked upon as one of the world's premier climate scientists he's a palette was a paleontologist with a fabulous record day bob carter told me personally one day that dr david evans is one of the world's top five climate modelers and we'll see why in a minute david is one of the very few models the only model on the spectacle side of the climate debate maybe there are others there are others who work on the models as a side apart from their normal science activities i'm going to read his background to make sure i get it right david was previously and this is significant the lead modeler in the australian government's carbon accounting unit that produces australia's carbon accounts he developed the full cam model which is the world leader in estimating carbon flows in a country's biosphere principally in forestry and agriculture full cam was also promoted up the clinton foundation for use in the third world so david is his work is known in the in the uh in the whole of the global climate area his formal educational qualifications and this will get you he has earned six university degrees in 10 years at sydney university and stanford university i mean getting one from standard stanford is fabulous getting six degrees from sydney and stanford in 10 years is just unbelievable all these degrees are essentially in applied mathematics and statistics and include a phd from stanford in electrical engineering david is currently investigating climate models while running a small business in the financial sector and writing a book on some novel aspects of mathematics and engineering welcome dr evans thank you senator robertson could you please we're going to focus on models as you know dr evans could you describe in broad terms what computerized [00:02:06] David Evans: numerical models are well let's start with models i mean a model is just uh it's a mental construction it's some aspect of reality it's an understanding we have may or may not correspond to reality but the crucial point is that a real a model exists in your head whereas reality is just out there and as children we learn you know hundreds of models as we grow up and as adults we're so used to our mental models that we barely ever notice them as models anymore uh for instance here's a quick example of a model a food gathering model you know you'll find berries in those trees over there but if you hear anything hiss then run away uh when we first encounter something we usually try to model it perhaps it works like this perhaps it works like that and as we learn more we discard models that don't work and you know our current model usually has the most detailed up-to-date information our best understanding of the situation now numerical models are just models that involve numbers and you know if they involve a lot of numbers well then we want to use a computer to do the number crunching uh days of sitting there with a calculator are long gone uh so numerical models senator roberts are just bundles of calculations that predict something about reality that we're interested in for example um our model of the of the earth and the sun developed a few hundred years ago was that the earth revolves around the sun once a year isaac newton about 400 years ago discovered gravity and from that we built up a numerical model of the solar system whereby we can predict very accurately all the planetary orbits and we can use that model to send spacecraft out through the solar system um a useful model predicts reality now some models are more useful than others because some are better at predicting reality than others all models are simplified versions they simplify reality and just focus on the aspects we're interested in uh a model that was accurate all the time would just have so many details we usually can't be bothered with it it's just as complicated as reality [00:04:19] Malcolm Roberts: so what's the main aim and purpose of modeling then is it to understand the world around us or the topic that's being modeled it's more than understanding a model reflects [00:04:29] David Evans: understanding the purpose of a model is to predict reality so we don't have to experience it to figure out what's going to happen so if we can model the the the results of various courses of action we can choose the course of action that's best for us without having to experiment with each path [00:04:45] Malcolm Roberts: okay okay so what we've seen is the models have been discredited uh what we know is the models have been discredited discredited in the climate sphere um comes down to something you talked to me about once before and that was about theory versus empirical data can you elaborate on that it seems to me that's where the conundrum is now well how do you know if a model works a model works if it predicts [00:05:09] David Evans: reality and it doesn't work if it doesn't predict reality i mean it's obviously it's on a spectrum the better the better a model works the better it predicts reality so a model is a theoretical construct it's a product of our minds and evidence is data about the world it's reality do the two combine or are they at odds with each other a good model there's good agreement between them for instance our model of the solar system is incredibly good it's quite a simple model and it's very accurate down to the finest detail we've ever observed so we strongly believe in that model uh most models aren't a hundred percent they don't work all the time in complete detail you know close enough is good [00:05:50] Malcolm Roberts: enough in most cases so you you've mentioned that a good model makes accurate predictions so the ability to make predictions accurately is a part of science and it's a part of assessing science but the ultimate arbiter of science is empirical data so hard measurements physical observations and and particularly when presented in a framework that shows cause and effect so when we can prove cause and effect empirically then we know how that process works is that still accurate then very much so [00:06:21] David Evans: um there's always a tension between models and realities you know our minds and our computers work with models but reality serves up evidence and they've got to match closely to have a useful model right and so [00:06:34] Malcolm Roberts: if there's a disagreement between a model uh prediction and the empirical data providing the empirical data accuracy has been checked and validated then uh the model is just is dismissed as wrong because the empirical data prevails yes what sort of people then because there's been criticism of some of the climate scientists uh who are modelers as just being remote mathematicians or statisticians rather than hands-on climate people now we've seen many air many areas of climate um on both sides of the argument uh we have practical people and on both sides we have to do some remote work as well so what sort of people that tend to become [00:07:15] David Evans: modelers ah senator roberts we're all right modelers just only some of us don't make numerical models more complicated than a household budget um yeah look sure some of us make numerical models on computers uh yes there'd be mathematicians engineers statisticians physicists chemists financial analysts things like that now most people in those areas use their knowledge to uh to solve problems maybe conduct research but a few of us more focused on encoding that knowledge into computer programs or models that can calculate useful answers um there is something of an art to making models that can be applied in almost any field um it's kind of inevitable the phenomenon you say whereby hands-on people haven't got enough time or expertise in life life's too short to be a modeler as well uh modelers can't be across everything i mean there's only so much time in a human life to learn so uh it's it's hard to master you know it and modeling and computing at the same time as completely modeling are mastering subject areas but a good model would be made by people who have those [00:08:23] Malcolm Roberts: skills and if they don't have the uh the connections with the real world or the physical world then they will have people who do understand the physical world to give them the the basic theory and they'll [00:08:34] David Evans: encode that remember a model is basically comes out of your mind it's a mental construction and so the people making the models the programmers the modelers have to have a good understanding in their minds of what realities what's going on they've got to understand the latest science and the principles [00:08:51] Malcolm Roberts: and all the rest of it so just because a modeler happens to be a mathematician or an engineer we should not dismiss it what we need to assess is not his his or her background but the quality of the [00:09:03] David Evans: model's predictions the only thing that matters matters is whether a model predicts reality all the [00:09:09] Malcolm Roberts: rest is just fair so i've learned from you in the past that there are two types of models in the climate area so the ones that i knew of before talking to you about that were the general circulation models what what do they model and what is it what does general circulation mean can you tell people i understand what it means if you could just tell people what general circulation models are [00:09:33] David Evans: well in the climate it's very important to uh to know about the circulation of air in the atmosphere and the circulation of water in the oceans uh so hot air rises clouds form cool air sinks mountains get in the way the complications are endless with that circulatory flow now simple models of the climate can kind of ignore circulation you can sort of get by and get a rough answer without even worrying about it too much but if you're going to get beyond a certain level of sophistication in your model in your climate model then you really need to simulate those circulations so the term circulation model is really a badge that tells you how sophisticated the model is it turns out that have been able to to simulate circulation reasonably well tells you that it's a very sophisticated model and so a general circulation model a gcm gcm that's what i'm going to call it is a big numerical model of the climate it simulates the air and the water and it needs a large computer and a fair bit of time to simulate the climate for a few decades because it does bazillions of computer numerical calculations and they're brought down sorry keep going look when people talk about the climate models what they actually mean is the gcm's these are the big models that whose output is produced to to show people and what have you now the first gcm's by the way were made in about the mid-1960s and because computers were so lame back then and the deep the models were very simple those early gcm's could be run on any modern pc now in a few seconds what they do is they petition the atmosphere into cells which are maybe five kilometers high maybe a thousand kilometers on each side latitude and longitude and they they have a number like temperature the average humidity the average cloud content in that entire cell and they might those early gcm's might calculate how the properties of each cell varies say from week to week and they might do it for a few decades the modern gcm's are a lot more ambitious a modern gcm has uh about a million lines of computer code it's not written by a single person it needs a small team of people to write to to write a gcm it uses grid cells that are only about a kilometer high now so there's about 30 or 60 of them stacked up in the atmosphere uh the grid cells have shrunk to about 30 kilometers on the side um they compute the temperature the humidity the wind speed cloud content all that sort of stuff for every grid still on the planet uh every 20 minutes for the simulated time like say between 1900 and 2100 200 year gap they'll compute that every two for every 20 minutes that's a sort of level of detail [00:12:19] Malcolm Roberts: they're getting down to okay so but the models even though their grid sizes are coming down which is wonderful news they're still they don't climate scientists um even non-modelers don't understand fully the climate processes even the flows of fluids both in the air and the and the ocean so still there's still a lot of assumptions there to feed into that grid because data in data out [00:12:43] David Evans: it's not just assumptions it's approximations Malcolm if a grid cell is 30 kilometers by 30 kilometers there's an awful lot of things that can't simulate accurately these are things called sub grid processes like updrafts an updraft might only be a few tens of meters or maybe a kilometer across well you're average that over a 30 meter by sorry 30 kilometer or 30 kilometer cell or clouds clouds are much smaller than a grid cell you can't simulate individual clouds yet in a in a climate model you're nowhere near it so there's approximations that are made and yes you touched on a point there in principle we there are some things about the climate yet that we do not fully understand we can mimic them in certain ways but we don't fully understand the physical processes so those do form limitations on our climate [00:13:29] Malcolm Roberts: modeling yes and and the the inability to model clouds is seen as a particularly bad bad error in the gcm's because clouds just slight increases or decreases in cloud cover around the earth can significantly affect temperature very much so clouds are a known a known problem a big source of uncertainty [00:13:57] David Evans: all the global warming we've seen in the last 100 years could be explained alternatively by about a half about a half of one percent change in cloud cover wow and that's that's easily that's easily occurring [00:14:10] Malcolm Roberts: well maybe it's hard to measure yeah um what about uh the radiative um factors i remember writing writing this some years ago in its 2007 science report the ipcc the un intergovernmental panel on climate change itself published table 2 11 chapter table 11 and chapter 2 showing the purported levels of understanding of 16 factors claimed to affect radiative forcing assumed in computerized numerical climber models of the 16 factors that they said had an effect they claimed that two had medium level of understanding one has a claimed high level of understanding despite empirical evidence to the contrary the remaining 13 have low or very low levels of understanding so that seems to me that the ipcc is itself admitting that over 80 of the factors that are the basis of the ipcc's unvalidated computerized numerical models have low or very low levels of understanding about radiative forcing [00:15:09] David Evans: uh well that's sort of true but it's sort of misleading as well in the sense that overall the climate scientists are pretty confident they have the general picture of what happens correct but an awful lot of what happens in the details we're still not sure of it we can we can sort of simulate it but it's it's not it's not 100 yet yeah there are a lot of deficiencies in climate models and the climate modelers and the climate scientists will readily admit that but but the but they're pretty sure that they've got the general theme [00:15:38] Malcolm Roberts: of it correct okay uh before we get on to the second type of models what is validation of a model and why is it [00:15:45] David Evans: essential and how is it done ah well validation is really simple it's like comparing reality with a model so you get your evidence and you see whether your model predicts it that's validation um now the climate models are a bit of problem here because climate is is a uniquely difficult area uh climate changes over so slowly and we've only been collecting decent climate data since the satellites went up in 1979 all the data before that is as you go back further it becomes increasingly uh uncertain and problematic and even the stuff since the satellites not all of it's good and not it's never sufficient you always want more data and there's only one historical record you can't do experiments with the climate like restart the earth and try something else and so on so you only got this one historical record to work with um the the models have been and now no models really validated in the sense that if you start off with the conditions that existed in say 1980 and you ran a computer gcm and you ran it forward say to 2020 to today so 40 years none of them can predict the climate today from the 1980 climate even given all the inputs that occurred between 1980 and today none of the climate modeling is not that good yet what a gcm can do though is it can produce outputs that look like they well they might be realistic that's the sort of thing that happens and the global the global temperature average that they produce more or less follows the historical the historical um record so they're sort of you know they're getting in the ballpark um they make a lot of regional predictions like what happens in say africa or australia or something like that uh those are pretty pretty random they're um they're not very good they're a little better than random guesses a little better but not much um different gcm's produce different results sometimes very different especially when they're asked to simulate the future so yes it's tough [00:17:47] Malcolm Roberts: so because they're not validated they can't be used for predictions and forecasts but they can be used for projections now the general public doesn't understand the difference so projection is just a wild guess [00:17:58] David Evans: well sometimes a projection projection it's sort of a compromise uh senator roberts look policy makers and politicians want firm predictions and they pay the bills the climate models and the climate scientists know they've got a very complicated system on their hand that they're trying to model and they don't fully understand it even in principle and the compromise is that we're not going to call them predictions but all right we'll call them projections and it's it's a best attempt with what we currently know and they're pretty confident as i said about the general thrust of things even if a lot of the details particularly regional predictions are at this stage only slightly better than guesswork [00:18:42] Malcolm Roberts: and and there are some regions in the world where the temperatures are decreasing at the same time as other regions the temperatures are increasing so uh it's very very difficult to to pull out something [00:18:52] David Evans: regionally from a global model global averages seem to be easier to predict in some ways than the details of the regional projections yet but but think about the way they're trying to model the world they're doing it in little grid cells all across the planet so if they get the regional projections wrong how can they get the the global average right is that just chance or is there some deeper principle at work we're not sure we're not entirely confident [00:19:16] Malcolm Roberts: and that's that's the significant thing and you're being very honest and open about it to have a good model you need to understand the processes that you're modeling and we don't understand climate processes so we hope that we're taking a step at it i'm not saying we shouldn't model because we should model because we will eventually maybe one day with computing power or complete change in how we compute models or layout models we'll actually have a very accurate tool maybe so um you know so let's continue but at least at the moment climate models are not at the point [00:19:47] David Evans: where they can make predictions accurately yeah they're in their infancy and i think it's a bit unfair to say we don't understand climate processes we understand most of them fairly well but we don't understand them completely and there are some we only have a passing understanding of where you know passing we can produce programs that mimic them but we don't really understand them and and also the interaction [00:20:09] Malcolm Roberts: of some of those um processes we might understand the processes but we don't understand their interactions and there are so many processes that too so this this thing i know i know what you're going to say here but i'm going to ask you um in the in the un ipcc's last three reports that's 2001 2007 2013 there was only one chapter in each of those reports that claimed warming and attributed to carbon dioxide from human activity in 2007 that 2001 that was chapter 12 2007 that was chapter 9 2013 it was chapter 10. now i've read each of those chapters because they're the sole chapters claiming warming and attributing it to our carbon dioxide and what struck me was the complete absence of empirical data but what also struck me was that they would call the output from a computer model data and then they would use that to discuss the word data their output from computer model is not data it's just hypothesis of course you're quite correct [00:21:16] David Evans: you can't the output of a climate model sorry it's an exaggeration only numbers derived from instruments in the field are data right increasingly however the situation is complicated by the fact that uh those numbers from the field gathered from instruments in the field are often adjusted adjusted to fall more in line with the theory which is what they're expecting and they're very confident about and there's also another process whereby they're supplemented sometimes you don't have data at all for some area at some time and so instead they substitute data from or sorry they substitute numbers that ultimately come from a computer model so increasingly there's a sort of blending of theory or numbers that come from theory and numbers that come from instruments in the field and they tend to label them all as data [00:22:10] Malcolm Roberts: which only further confuses lay people yeah and and i know that you're very pure in that sense that that data is measured observed that's it full stop in the story you've got to make a clear distinction [00:22:22] David Evans: between them the outputs of a model are not evidence they're just the results of a calculation of a mental construct all right that's quite different to evidence which is a measurement of reality calling model outputs data is like confusing your daydreams with reality now i had a conversation with another australian senator uh about 10 years ago and she's still in the senate and i remember her after the after one of the climate scientists had uh given a small presentation of modeling outputs and she was picking up the shape of papers and she was waving it around and saying so this is evidence this is evidence how do i how do i account for this evidence and she had it entirely wrong she's a lawyer by training so you know we can maybe make an excuse for her but model outputs are not evidence [00:23:08] Malcolm Roberts: and that's i don't think we can make excuses for lawyers because they should understand what the word evidence means but anyway um so how the climate models doing the gcm's before we start talking about the ff fffs what how are the gcm's doing what is the level of agreement between models and reality so now [00:23:26] David Evans: we're getting to the place where we we think we talk about reality versus the models how are they doing what are they doing i suppose they're there let's divide them to three main areas um the big important area that everyone focuses on is global average temperature is it is it going up how fast uh how much higher is it going to go um the climate models are tuned to match the historical record temperature record that is the uh the model settings or parameters are adjusted so that the model's output best resembles what we're recorded actually actively happening so while it's reassuring that the gcm's reproduce approximately the historical temperature record it's not really all that significant um the record of the gcm's at outright prediction of global temperatures is weak uh it's not without some skill but it is weak they've persistently overestimated global temperatures every time except from the period from 1975 through to 1998 when the planet actually was warming quite quickly that was followed by a pause in the warming a period of about 15 years where it didn't really warm much uh we don't really understand why that occurred even in retrospect the model certainly didn't predict it um the estimates of warming haven't been that dramatic uh it has generally warmed since the 1940s so i suppose any warming prediction is kind of look going to look kind of right um um you might overlook the overestimation uh especially if the historical records adjusted a bit um but i think privately the model builders must be disappointed at the performance of the gcm's at predicting even global average temperature Now, the second area I'm going to talk about is more important for the purposes of validation, and hardly ever gets talked about. It's a little bit esoteric, but let me explain. The second area is about the upper troposphere. Now, the climate models get it entirely backwards, which is the source of a lot of angst and controversy. The upper troposphere, we've got the ground, the upper troposphere is from five to 12 kilometers, generally speaking, roughly. So it's up there, but a fair way up there. And it's a crucial area to global warming. The reason is that about 50% of all the energy that the Earth radiates to space is radiated or originated in the upper troposphere. That is, molecules in the upper troposphere emit that radiation and off it goes into space. About half our energy comes from there. So the upper troposphere is the major cooling element of the planet. So if you model that wrongly, how can you model global warming correctly? It's kind of important. The climate models say that as the surface warms, the surface of the planet, the upper troposphere will become moister and warmer. But the evidence is that since the 1970s, as the surface has indeed warmed, the upper troposphere has instead slightly cooled and slightly dried. Now, this evidence comes from weather balloons, which we also call radiosons. These are the only instruments we have with enough vertical resolution to measure the sort of changes we're looking for. They've been operating since 1948. Nowadays, there are over 900 launches of radiosons per day globally. Over 30 million radiosons have been released. 30 million? 30 million. We're not talking about some flash in the pan, one-off sort of thing here. There's a lot of redundancy in this system. Now, no instrument is perfect, but the picture that the radiosons paint is clear enough. There's been a mild drying and cooling of the upper troposphere since the 1970s. Not the warming and moistening predicted by the climate models. Now, you might not have heard a lot about this. Okay, it's a little bit esoteric, but it might also be because it plainly demonstrates that there's something fundamentally wrong about today's climate models. Today's GCMs are flawed in some way because they're predicting something that's really quite at odds with reality. And it's an important part of the whole puzzle. The climate scientists are obviously a bit embarrassed about this. They will explain away, they have labored mightily over the last two decades at least, three decades, to explain away the discrepancy as bad data. They emphasize how unreliable radiosons are. They make some legitimate points. Furthermore, if you squint hard enough at certain satellite data, well, it kind of looks like it's not actually getting that much drier and cooler. All right, so, you know, maybe. But the satellites don't have the vertical resolution to look to measure the changes we're looking for, and so they aren't really relevant here. If you want to do further research, this controversy sometimes goes by the name of the missing hot spot or the tropical hot spot. Now, in the climate models, these trends in the upper troposphere are responsible for over half of the global warming attributed to carbon dioxide. Carbon dioxide warms the surface and then the trends in the upper troposphere amplify it by a factor of two or three. So, if they have the trends in the upper troposphere backwards, the amount of warming we're going to get from carbon dioxide is only a fraction, maybe a fifth, of what is currently predicted by the GCMs. So, it's a very important point that gets right to the heart of global warming and global warming and climate policy. You hardly hear anything about it. Third, third area of reality versus models is in regional predictions. And as I've mentioned, the regional predictions of GCMs is only a little bit better than random guessing. So, that's definitely a sign that things aren't correct there either. So, all up, I'd say the climate models are doing a pretty ordinary job. They've definitely got some skill and surely they get a lot of stuff right, but there's still some things that are probably fundamental or definitely fundamentally wrong. It's more than just a little bit of refinement and bigger computers that's going to fix up. So, if there is something fundamentally wrong, [00:30:06] Malcolm Roberts: how can we model global warming correctly? We can't, can we? [00:30:12] David Evans: All right. Now, I'm going to tell you something that most people don't know and surprises people a little. Look, the world has embarked on our current climate policies of reducing carbon emissions, despite the disagreement between the GCMs and the upper troposphere. On the face of it, that seems odd. Why would you follow the advice of models that plainly don't work? That's the position of climate skeptics. Why are you doing that? But what it suggests is that the GCMs are not really the reason that the world is doing this. And that is, in fact, the case. It's not the GCMs. The actual reason that people are afraid of dangerous warming due to increasing carbon dioxide is a much smaller model that can be calculated with a pen and paper on the back of an A4 envelope. This basic physics model, it's just based on basic physics, the sort of physics that you learn, perhaps, in undergraduate physics. It's known as a forcing feedback model. It's part of a family of models known as the forcing feedback framework. And we'll sort of talk about a generic model from that family as an FFM or forcing feedback model. Now, it's actually the FFM on the back of an envelope that convinces climate scientists about the dangers of carbon dioxide. It's that that the world is following. No one is convinced by a million lines of computer code. Computer code is notoriously buggy, often does things you don't understand, even the people who programmed it. And a million lines of it is an absurdly high amount to trust, especially for models that haven't been validated and don't seem to work 100%. Okay, the GCMs, as you as you said, will one day be brilliant with sufficient refinement and understanding. We're not there yet. The GCMs are interesting, they're impressive, they're high tech, but they're not ultimately convincing, because they don't work properly, and because no one trusts a million lines of computer code. It is the FFM that motivated this whole debate. In 1979, the political establishment in the United States was approached by the climate scientists, pointing out that going, "I think we have a problem here with carbon dioxide." President Carter of the United States did the right thing. He said, "Oh, okay, well, let's set up a blue ribbon committee to look into this." And he appointed all the best people in the United States to look into the state of climate sciences and see whether carbon dioxide was indeed a problem. He set up a committee called the Charney Committee, and they produced the Charney report. Charney was the lead meteorologist in the United States at the time. And they looked at it, and they produced a report, and the first thing in the report was a quite, it was a blessedly small report, about 30 pages, but the first thing was they looked at the basic physics. They looked at the FFM, and they used the FFM, and they said, "Oh, we're going to get this much warming. It's dangerous." They also looked at a couple of the two leading GCMs at the time, and the two leading GCMs at the time disagreed slightly with the FFM. One was too high, one was too low, so they said, "Oh, split the difference. It's in the middle." And so they pretty much went with the FFM. And the FFM had a certain amount of uncertainty about it. And by the way, the latest assessment report in 2013, AR5, has the same level of uncertainty that was expressed back in 1979 in the Charney Report. In some ways, we haven't made very much progress. We're still following the FFM, and we still have the same amount of uncertainty. The GCMs, in a sense, have contributed nothing in 40 years. I mean, we've learned, but we're still basically playing with the FFM, and it's still this basic physics model that is guiding world policy, because the FFM unambiguously indicates that at the rate we're increasing carbon dioxide, we will get dangerous global warming. Now, the FFM uses basic physics, nearly all of which has been validated in laboratories. Laboratory-tested physics is the best knowledge that humans have, because it's replicatable. Anyone can do the experiments. You can do them in your lab under nice controlled conditions. You can see what happens when this happens, and you get knowledge about how radiation works and all this sort of stuff. And that is what's plugged into the basic physics model, essentially. So, as a result, our climate scientists are so sure that they're basically correct because of the FFM. [00:34:46] Malcolm Roberts: So, how do we explain then that as China and India rapidly increase their amount of carbon dioxide produced, and human production of carbon dioxide has increased dramatically in the last 25 years, but we've had basically a flattening of the temperature. So, how does that conundrum resolve itself? [00:35:11] David Evans: Well, I mean, it gets to the real state of the climate debate. We're saying that the climate debate is essentially, in an intellectual sense, maybe it's out of view of the public, but in an intellectual sense, is in an intellectual standoff. On the one hand, the climate scientists have their FFM. They believe cause dangerous warming. It's theoretical. It's based on basic physics. Climate skeptics look at the evidence, such as the stuff that you just quoted, and say, well, it doesn't really seem to be working. Are you sure about that? Seems a bit wrong. So, some people go with the evidence, some people go with the theory. It is a classic case, and we've had this before in science, of theory versus evidence. Now, the climate scientists acknowledge that some of the data doesn't fit their theory. Not idiots, but they would point out, and rightly, that all climate data is somewhat dubious. You see, to collect climate data, you have to put a whole lot of instruments out in the field. And they get rained on. They get washed away with water. They get dried out. Things degrade. They lose calibration. Instruments need replacing. The instruments that replace them are slightly different from the original instruments. Now, for studying climate change, we're looking for small changes in the climate. So, back comes your data, and it's got some small changes. But are those small changes due to some degradation of the instrumentation? Or are they, in fact, due to climate change? Or maybe a bit of both? And you can always argue the toss. Always. So, here's what the climate scientists do. They are so implacably confident in the basic physics model that they feel justified in just ignoring evidence that disagrees with them. That's just why they do it. They go, well, look, your evidence, you know, you collected it like this. There's some doubts about that. We know for sure that basic physics is right. So, that evidence, we're going to discount it a bit. Furthermore, they feel justified in adjusting some of the data sets a little bit. And 99% of the, well, there are problems with the data. There are problems with the data. So, we need to fix it. And 99% of the adjustments move the data towards the theory. But consider, if the fixes to technical problems were unbiased, then by chance alone, half of the changes would move the data towards the theory and half would move it away. But that's not what's happening. So, we have a situation where the climate scientists are informed by a basic physics model and they're pretty certain they're right. And so, when you point out the carbon, the temperatures didn't rise as much, they mumble a bit and talk about the temperatures maybe not being measured properly. And, you know, there are little unknown things about the climate that maybe make it very off, you know, natural variation for a few years and things like that. But they are quite sure that they're basically correct and that it has in fact been on a warming trend now for about 100 years. [00:38:08] Malcolm Roberts: Okay. There are a number of things that I would raise in that. First of all, if they're claiming the data is wrong, how do they resolve the fact that their models are also based on data? Was that data wrong too? Because that wasn't even adjusted back then. So, the foundation of the model must be [00:38:27] David Evans: wrong if they're saying that it's wrong now. The basic physics model comes out of the physics laboratories by and large. Yes, what I'm saying is, oh, I see. That's replicable data. Okay, we can experiment with that all day long and really figure out and get to the bottom of it. Climate data is much harder because you only get one shot at it because history passes slowly and the changes occur very slowly. You know, we don't have hundreds and hundreds of years to do it over and over again and really collect data. There's a real paucity of climate data, Senator Roberts. There just isn't [00:39:00] Malcolm Roberts: enough of it. We always want more. Well, then the other question it raises also is that while we understand the behaviour of carbon dioxide in a closed lab container, that's quite a bit different from the dynamic open atmosphere. And we know that forcing, what did you call it? Forcing feedback models, that rely upon water vapour because it's quite rightly pointed out, carbon dioxide is just 0.04% of the of the Earth's atmosphere. So people say, how can it possibly have much effect at all? And the answer from the climate modelers and the alarmists is, well, hell, it's amplified because of water vapour. But we know that water vapour in a practical sense, it is a temperature modulator. So it decreases the maximum temperature and increases the minimum temperature, generally speaking. And sometimes water vapour can have a warming effect. Other times it can have a cooling effect, but the net effect of water vapour in the Earth's climate is as a coolant. So I don't, I can't see how they can take a lab and say, well, what applies in the lab, we will just move into a dynamic atmosphere, open atmosphere. [00:40:10] David Evans: Well, I'll show you briefly how they kind of did it. Oh, by the way, a very small amount of, if you just take a drop of dye and drop it into a glass of water, it immediately turns the glass into something opaque. So a very tiny amount can have an effect on radiation and its transmissibility of radiation. What you're talking about with water vapour is what is one of the feedbacks. In fact, water vapour feedback is the main feedback. And water vapour feedback can kind of be worked out. We know from our understanding of radiation, how adding extra carbon dioxide to the atmosphere makes it more opaque at certain frequencies, and that causes the surface to get warmer. We can then work out how much extra evaporation that causes. And we understand how much water there is in the atmosphere and the saturation levels of the atmosphere in humidity and so on. And we assume that relative humidity stays about the same, which, you know, it borne out in practice as well. So it's a reasonable sort of thing to do. And that is behind the idea that the upper troposphere moistens and gets warmer as the surface warms. And that water vapour feedback, Malcolm, dominates all the other feedbacks. The water vapour feedback is much greater than the sum of all the other feedbacks together in the climate models. Now, the terrible worry for the climate models is that maybe there's some feedback out there they haven't thought of, right? Maybe there's some feedback out there that's changing things in ways that they hadn't anticipated, and it's been emitted from the models. And that's what caused the temperature fluctuations of the last few years. It caused the pause, it causes this and that and the other. But they don't know for sure. So in answer to your question, they think they have it about right from just from applying basic physics, but they can't be entirely sure. [00:41:59] Malcolm Roberts: Right. And I would say that this must be a big question mark in an open, dynamic atmosphere, especially because there should be changes then as carbon dioxide increases, there should be changes in humidity around the planet. Or you're saying, but that doesn't seem to be happening at all. And certainly, you've just pointed out the key problem they have is the upper troposphere. [00:42:21] David Evans: That is a key problem. And it's also connected with the cloud problem. Yes. Well, I mean, water vapour changes and so does clouds. How are the clouds changing? Well, we're not entirely sure. So there are problems. [00:42:35] Malcolm Roberts: So you've used the term standoff, Dr. Evans. So we're an intellectual standoff right now. Yes. What's going to resolve that? [00:42:45] David Evans: Well, I suppose that eventually, in the fullness of time, we'll find out who was right. So by the year 2100, it will have warmed several degrees Celsius, if the carbon dioxide theory is correct. So if it has warmed that much in the official temperature records, or if our lived experience of temperature indicates that it's a lot warmer in 2100 than it was back in the year 2000, then the theory will be judged correct. And that's one way of resolving the standoff. Another way is if temperatures don't go up nearly as much as they say they do, or the official temperature record continues to rise, but people's lived experience of temperature says, not really, it's kind of the same, then we'll sort of disbelieve the theory. There is another alternative. There might be a mistake in the basic physics model. And at any one time, at any time, someone might find it and come forward and say what it is. And that too could resolve the argument rather quickly. Are you capable of doing something like that? Oh, definitely. Definitely. That's what I guessed. You know, I understand climate models very well. I understand the FFM very well. I know where to look for the mistake. If I had funding, I reckon I could find it. If I had enough funding to be able to hire a few people and really get things done, I reckon it might take about a year to find the mistake and to make the case publicly. [00:44:14] Malcolm Roberts: Well, then, if there's an entrepreneur out there or a philanthropist, take note of this because we've got one of the best minds on the planet right in front of us right now. And he needs funding. And think about the investment that returns that would make. If Dr. Evans could find the error out of a small investment for a wealthy philanthropist, it would save the planet trillions of dollars. And I mean that literally. Trillions of dollars. I'm not exaggerating. One bit. Trillions of dollars. That's a very healthy return, plus all the destruction that we're going through and will be going through as we destroy our energy sector and many other sectors, including agriculture. So if there's anyone who's willing to come forward, let's have them because this is really significant work that needs to be done. So Dr. Evans could actually end that intellectual standoff. He's got the ability to do it and he's got the understanding. David, is there any other topic you want to cover? [00:45:15] David Evans: Oh, not really, Malcolm. Senator Roberts, that was very kind of you to do this interview. [00:45:21] Malcolm Roberts: Thanks for letting me speak. The other thing that I would like to talk about, a couple of things if you've still got time. I know you're busy. Underpinning whether it gets warm or not, regardless of that, underpinning the whole assumption there, the underpinning assumption is that warming is bad. Yet the scientists have labelled past Earth's past far warmer periods as climate optimals because they've been so good for civilisation and the environment, the natural environment. So whether the models are right or wrong doesn't matter because if carbon dioxide raises the planet's temperature, wouldn't that be wonderful? [00:46:04] David Evans: It might, yes. The period of the ancient Greeks and Romans was known as the Roman optimum and it is mostly believed that the temperatures were about one or two degrees Celsius higher than they are today. And of course, all of civilisation only grew up in the last 10,000 years since the end of the last ice age, which ended about 10,000 years ago and then 100,000 years prior to that were an ice age. So it would appear that we humans really need warmer climates or thrive in warmer climates. [00:46:40] Malcolm Roberts: So, second last question I've got is you're married to Joia Nova, who's looked upon around the world, a number one website, I think, for climate bloggers in Australia or climate generally in Australia, won awards internationally around the planet for her work, a phenomenal talent, a very effective communicator and dogged and just persistent and very high standards, always insists in doing properly reviewed work. Now, she's one of the world's best skeptic and she draws attention to inconsistencies in the alarmist work. You're one of the world's, well, you're the only well-known climate modeler in the skeptic side of things. You're both working voluntarily. [00:47:30] David Evans: Pretty much, yeah. We funded ourselves for several years off the earnings I made from the Australian Greenhouse Office working on the other side, doing their carbon models. That's kind of run out over the years. We sort of got stuck in this longer than we ever expected to. We kind of got cancelled a few years ago, so we don't really have a lot of options. But yeah, we sort of live on the smell of an oily rag, so any donations are being, you know, gladly accepted. Joanne's website is at joannenova.com.au. It's one of the handful of leading climate skeptic websites in the world and it's the leading climate skeptic website, I'd say, in Australia. Definitely in Australia. Yeah. Definitely. Yeah. [00:48:18] Malcolm Roberts: And as I said, both are meticulous in their thinking, in their standards. So, I'm going to go and patronise joannenova.com.au. Last question for you, Dr. Evans. What, what gives you satisfaction? What, what are you passionate about? Because you've, you've worked in many areas. You've got a very inquisitive mind, a very bright mind, very capable mind. What, what powers your heart? What, what energises you? I think what motivates me most is discovering stuff. [00:48:58] David Evans: The thrill of discovery, I'd say. Writing it up afterwards is a bit of a, is a bit of a bore sometimes. But finding out, finding out new stuff and, and understanding stuff. Yeah, it's the thrill of discovery is what drives us on. So you love the truth. Yeah. Yeah. And, and it annoys me too when people vaguely disobey. I guess that's why I come down on the empirical side of the theory versus [00:49:25] Malcolm Roberts: evidence question. Well, thank you very much, Dr. Evans, for your time and especially for sharing your knowledge so freely because people who talk about models, whether they're for them or against them, wouldn't have, wouldn't have understood some of the things I previously didn't know some of the things you've raised today. You've raised a number of significant things. And I think a lot of people get a lot of benefit from that. So thank you very much. And finally, thank you so much for, I know you left the, the greenhouse gas, uh, carbon farming modeling for the government and left a career there because you wanted to, you wanted to be on the side of truth. And so I thank you for your courage and your integrity and for your hard work. And please say hi to Joanne when you see her, please. Thank you, Senator Roberts. Thank you, Dr. Evans.