A Day in the Half-Life

Climate Modeling

April 22, 2022 Berkeley Lab (Produced & hosted by Aliyah Kovner) Season 1 Episode 7
A Day in the Half-Life
Climate Modeling
Show Notes Transcript

For decades, scientists have been able to predict future Earth conditions, like rainfall and temperature, with impressive accuracy using computer programs called climate models. These models are helpful at telling us what might happen to our weather depending on how much we curb greenhouse gases emissions now, and they can be used to study how much human-driven climate change plays a role in big events such as Hurricane Harvey or last year’s Pacific Northwest heatwave, compared with our planet’s natural processes.

We hear about climate models all the time, but how many of us know how they actually work?  In this episode, we peel back the curtain, discussing where these models came from, what they can do amazingly well, and their current limitations. And our guests talk about what it's like for them, personally, when their work is doubted, minimized, or politicized. After all, climate scientists find themselves in the hot seat a lot more often than other scientists. Today's guests are experts not only in the science itself, but also expert at staying cool under pressure, communicating their science with the public, and laughing off the negativity.

Featuring:
Jennifer Holm, a research scientist in Berkeley Lab’s Climate and Ecosystem Science Division. Her work focuses on modeling terrestrial ecosystems, with an emphasis on tropical forests.

Michael Wehner, a senior scientist in the Applied Math and Computational Research Division. A veteran in the field, Michael used to write climate models, and now uses them to study how human-caused climate change impacts extreme weather events like hurricanes. 

Aliyah: 

Welcome to a Day in the Half-Life. I’m Aliyah Kovner. In this episode, we’re going to peel back the curtain on climate modeling. 

For decades, scientists have been able to quite accurately predict future Earth conditions, like rainfall and temperature, using computer programs called climate models. These models are helpful at telling us what might happen to our weather depending on how much we curb greenhouse gases emissions now, and they can be used to study how much human-driven climate change plays a role in big events such as Hurricane Harvey or last year’s Pacific Northwest heatwave, compared with our planet’s natural processes. 

But I for one, rarely stop to think, wait, what is this model? How does it work, what does it LOOK like? And what is like to BE a climate scientist, to work in a field where your research is shaded by politics? Where your hard-earned study findings are hotly sought after by news outlet A, then dismissed out of hand by news outlet B? 

We get into these questions, and more, with my guests Jennifer Holm and Michael Wehner. Jennifer is a research scientist in Berkeley Lab’s climate and ecosystem science division, who uses models to study how Earth’s terrestrial ecosystems will respond to climate change. And Michael is a senior scientist in the applied math and computational research division, focused on extreme weather in a changing climate.

 Hi Michael and Jennifer, thanks so much for being here.

Michael:

Thanks for having us

 

Jennifer: 

Yeah, happy to be here

 

Aliyah:

All right, I'm gonna jump right in. What, on a basic level, is a climate model?

Michael:

Well, I'll take that one, Aliyah. A climate model is basically our best representation of the climate system. It's based on many, many years of research, in fact, but in a nutshell, it's a computer model that a computer program that solves a set of equations that represent the various aspects of the climate. For instance, in the atmosphere of the ocean, the equations are called the Navier-Stokes equations. They're they're 19th century fluid mechanics equations, and they're, they're configured such to represent the the flow of the atmosphere. And some of the processes are very explicitly represented, like, you know, sort of the law, large scale circulation and, and, and heat transfers between the various parts of the climate system and other parts have to be approximated and what we call parameterized.

Aliyah:

And so what are these models really good at doing? And what are their current limitations?

Jennifer: 

I think these models are, are really, really good at simulating, you know, so many different and detailed processes across the whole planet like Michael was mentioning, you know, they, they simulate a lot of different components of our very large complex earth systems. So everywhere from ocean and atmospheric circulations to sea ice, to the land and biosphere. They're really good at telling us individual processes, but also all the way up to these large global patterns. You know, once we put all these individual components into the code and into the model, you know, that's what gives us these large global patterns.

Jennifer:

And because all these individual processes are pretty robust. I think that these models are good at showing us, you know, future projections of what the world would be like under different scenarios of climate change. And we know that these projections, you know, are pretty good because we are able to do hind casts. We are able to show how well we can predict things in the past that have already happened. So, so this paper's a little older now, I think it was out in 2019. But there was a paper that came out that showed that these models, you know, a whole group of these models actually have done a really, really good job at predicting exactly our rise in temperature. So, they looked back at these models that had been published a decade or two ago that predicted the future. And now this future is, is here, you know, that, that maybe these models were started in the seventies or the eighties or the nineties, and they didn't know what warming would be like in the 2000s, but now that we have the observational data, we went back and looked at it and it's tracked it pretty well. Almost exactly. So yeah, so, so now these models are also used as show future projections.

Michael:

Indeed. That's what I was going to add as well, that, that, you know, what do we want, what do we really want to do with these models? And, and what we want to do is, is understand the climate system, but at the end of the day, we really wanna know what's gonna happen. You know, how can we prepare society for the inevitable climate change. And you know, climate models have been the subject of much criticism because there are a lot of approximations, they are big complex things that a lot of people don't necessarily understand. But perhaps their most important validation has been these studies go back and look at the simulations that were made 20 years ago or even longer. And seeing that, yeah, they they're predicting that they predicted now 20 years ago, quite accurately. And so that gives us confidence when we try to predict project, what, what future climate change might happen um, in the next 50 years or the end of the century.

Jennifer:

And with that looking out into the future, they are, they have been really good at showing us this very direct tight link and tight connection with increasing greenhouse gas emissions and rising temperatures, that these two are basically lockstep increasing with each other. And since we know this, these models can tell us, you know, how much left of this total carbon budget do we have to stay with within, you know, a reasonable 1.5 degree warming or two degree warming. So we could even look at how much we could emit before we hit some of these, these critical, critical marks.

Michael:

I think it is also important to talk about uncertainties from the climate models and in our projections. And there are various sources of, of uncertainty and the models are, are one of those, but as we try to project into the future, the farther we go into the future, the biggest uncertainty is not coming from climate models or understanding of the climate system, but understanding of what humans are going to do. And so are we going to have a future where we have, again, reductions in greenhouse gas, emissions, carbon dioxide, and others, or we're gonna do a business-as-usual kind of scenario. And those are two very different futures. 

Aliyah:

So, let's talk about the history of climate modeling. This is something I've been thinking about a lot because the 2021 Nobel prize in physics was shared by three scientists, Suyukuro Manabe, Klaus Hasselman, and Georgio Parisi, who helped lay the groundwork for climate modeling. And the Nobel committee cited work that started in the sixties when Manabe led the development of physical models of the earth climate up into the eighties, when they note that Parisi discoveries on the physics of complex systems, allowed scientists to create these hugely intricate models that are composed of all these physical processes and phenomena kind of mixing together and interacting. But how has the field evolved since then, since what they cited sort of ends in the eighties?

Michael:

Suki Manabe’s efforts were of three really at the time in the, in the early sixties. One of the others being Chuck Leith at, at the Lawrence Livermore National Laboratory, and Akio Arakawa at UCLA. And and what, what these three did was really kind of bold is they, they tried to put the, these equations of motion that describe the atmosphere onto a computer. And indeed, the climate models we have today pretty much directly are ancestral to that. 

 

Aliyah:

That's cool. 

 

Michael:

And there's also, there's been a lot of improvements of course, but that basic idea is essentially the same. And indeed, you know, these ideas didn't originate with Manabe and Arakawa, and, and Lee. So one of the, one of the first to actually try to put these equations on computer was with the, one of the very first computers, the UNAC was John van Neman, who thought we could predict weather. And of course we can, and, and that has become very successful. That was done just after the second World War in the 1940s. And prior to that was a remarkably prescient book by L.F. Richardson who envisioned predicting the weather with computers, but computers, of course, this was in the 1920s and computers then were not machines or transistors or even tubes, but rather were people. And he envisioned, he, he envisioned an auditorium full of computers. Each of whom would be assigned a small portion of the, of the area where you wanted to predict. And there would be a conductor in the middle coordinating these people to to, to synchronize their computations. 

Aliyah:

So that's interesting because it sort of reveals that the desire to do climate modeling and the, how should I say this, the, the curiosity to model the climate and to try and put that into action has sort of existed since alongside the concept of a computer since the very beginning. And I'm assuming that a lot of the evolution that has happened during all those decades is because of evolution in computer technology. Is that right?

Michael:

Well, actually it even goes back farther than that. I mean, the idea of, of climate change was, was really first more, most clearly stated by a guy named [Svante] Arrhenius back in 1904.

Aliyah:

Oh wow.

Michael:

And of course he didn't do this with computers he did, or with computers, either people or, or machines, but rather by himself with, with pencil and paper. And he worked out a key quantity that we still talk about today called the climate sensitivity. And, and the way we define it today is: if you double the amount of carbon dioxide in the atmosphere, how much does it warm? And his estimate of that stands today. It's well within the range of what the most recent Intergovernmental Panel on Climate Change stated. But, you know, he also was relying on work from others prior to him. And so the, the curiosity about the about, about the world we live in – climate being part of it – goes back centuries.

Jennifer:

Yeah. I guess I kind of, yeah, I was thinking a little bit more kind of the development of, of more recent climate models. We basically just started really simulating these general circulation patterns in the atmosphere only, and then slowly added on different realistic parts of the planet: you know, started adding in ocean circulation patterns, and then ice, and then the land, and then eventually becoming more and more detailed to really represent our entire earth and globe. And you know, now we've gotten to this really cool kind of spot where, where it really is doing very, very detailed processes, like all the way down to microbes in the soil to photosynthesis well inside each individual leaf of a, of a tree. So, so yeah, so they've become really detailed over time.

Jennifer:

I kind of have a quick different segue if you don't mind me kind of talking about this, this is slightly different but I thought might be interesting for the, for the people to hear.

 

Aliyah: 

Yes.

 

Jennifer:

Where alongside the history of climate modeling itself is also the history of just what we as society, and as humans, have decided to do with this scientific information, and how we have tried to stop climate change. And I kind of like to think that there are three thematic era eras really. Other people have talked about this more eloquently that I have, but these eras are really, you know, mitigation, then adaption, and now kind of where we are is in losses and damages. So, we know that greenhouse gases will trap radiation and it'll lead to warming and have catastrophic consequences. And you know, the greenhouse effect is physics, period.

Jennifer:

So, all the countries of the world got together at these United nations conferences and understood that they needed to try to mitigate, you know, or stop, stop emitting these greenhouse gases. So this kind of happened in the 1980s and 90s. So, that's the first era. But then all the countries, you know, went back home and didn't do what they agreed on. You know, we kept emitting, emitting, emitting, greenhouse gases over time. So then I think we kind of entered the next era where it was starting to get into adaptation where we know we still need to mitigate, but it wasn't really happening. So it's like, well, how can we adapt to climate change? You know, how can we try to help adapt to sea level rise, to oncoming droughts, to oncoming floods. And we see that part of the world is gonna be impacted and especially poor countries of this planet are gonna be impacted. So, what can we do to try to help adapt to that? 

 

Jennifer:

And I think kind of now in the last couple years, we've entered this last era where climate change is here, it's happening, you know, people are getting displaced. So, you know, we can't really adapt to something that's already happening. So now we're in the losses and damage era where I think we do need to try to help for losses and damages of livelihood, agriculture, homes.

Aliyah:

Yeah. And you both, I mean, as part of your roles as climate scientists have been, you know, you've been working while these social and political have been happening. So, speaking about that, what is it like working on the, and sort of seeing the front lines of climate science interacting with policy?

Michael:

Yeah, it's been a great privilege of mine to have been an author on the fifth and sixth assessment reports of the Intergovernmental Panel on Climate Change [IPCC], as well as on the second, third, fourth, and now fifth reports of the United States National Climate Assessment. And as a scientist, these are, these are fantastic opportunities because you're in a, you're in a place with some of the best scientists in the world. And it, it was a tremendous learning experience for me each time. But, you know, as an individual, as a, as a citizen, it's also, it's clear that there's a tremendous responsibility that comes with this as well. Right. This was dawned on me when, you know, not long after my, my participation in the fifth assessment report of the, IPCC, when I was in a in a big meeting and somebody shows a picture and it's one I made and, and they were talking about things that I said, and it's like, oh my Lord, you know, people are, you know, tens of thousands, if not millions of people see these pictures and, and read these words.

Michael:

And you know, one is, one is struck by that responsibility. And so you, the, the care one has to take to make sure that what one says in these reports is, is as correct, as we know, and trying to, to really synthesize what the community at large is trying to say.

Aliyah:

And Jennifer, I heard that you were invited to COP26, which was the most recent, large global climate conference. What was that, what was that experience like? 

Jennifer:

Yeah, exactly. I, yeah, I had the, the pleasure and just the great opportunity to be able to attend this United Nations climate change conference. There was, you know, the largest commitments ever made to mitigating climate change and they, they were really trying to say, we need to make strong emission reduction commitments by 2030. And for the first time fossil fuels were called out – the wording went back and forth a little bit of what the exact wording is gonna be in the Glasgow pact, where it called for a “phase down” of unabated coal use and a “phase out” of inefficient fossil fuels. There's a lot of talk between this phase down and phase out where, you know, we would hope there would be a total cutting out off, of these fossil fuels’ use. But you know, they're just slow, slowly making progress, I guess. 

Michael:

One of the things that I observed, you know, from distance not having been at the COP, but, but sort of as a result, was something that I've been saying for a very long time and others like me, that dangerous climate change is here now. It is not our grandchildren's problem. It's not our children's problem. It's our problem. And as, as Jennifer mentioned, the loss and damages part of this is, is large. You know, we we've had tremendous disasters in our own country, you know, hurricane Harvey cost somewhere around $90 billion. And there's plenty of work on this that demonstrates that the rainfall in that storm was increased dramatically by climate change heat waves in the Pacific Northwest. 

Aliyah:

Yeah, I was stuck in that one.

Michael 

You were stuck in the heat?

Aliyah 

I was, yeah, it was awful.

Michael:

I'm glad you survived. Cause many people didn't, I mean,

Aliyah:

Yeah

Michael:

Yeah. It, it, people are, people are suffering, people are dying, people are losing their, their livelihoods. And this is happening now. Of course, one of the things that Jennifer mentioned is that, you know, the people who suffer the most at, from climate change are, are the poor, right? The poor in our country and the poor in other countries. And, you know, arguably these are the folks that, that contributed least to the problem. You know, it's the rich countries, you know, that, that have the highest per capita consumption of fossil fuels, but the rich countries are more can more easily adapt to these threats than the poor countries or in, in our country, in our own country. The poor people who are, who, who are in lower income groups are less able to adapt.

Aliyah:

And so, for you both to be at these big climate summits whether it's COP26 or, or one of the other recent ones, I mean, I imagine it's can be extremely frustrating to be at the 26th climate summit and still have there be basically couched language about decreasing use of fossil fuels. And so I'm wondering just how you both sort of deal with that pressure of having to just calmly press on with the science that you're presenting even when the recommendations and the kind of necessary steps aren't being taken. How do you, how do you sort of stay calm, keep your cool scientist hat on and keep going with the data in these frustrating situations?

Jennifer:

Yeah, yeah, you're exactly right. That it can be, it can be frustrating at times, you know, that we, we, we know that science, we know what's happening with climate change with warming temperatures, with increasing floods and droughts and storms. We, we see this and you're right. It, it can be very frustrating to maybe see not as much action being taken and a big kind of takeaway that I did see at COP26, which was wonderful, was the huge rise of youth activism. It was incredible to see all the youth that was there, the protests that were happening outside the different marches and how the youth is really galvanized. I mean, they know, I mean, this is the world that they're inheriting that we're leaving to them. And they, they get it. So, the, the youth activism was very, very wonderful to see, and we, and we need to listen to the, to these people more, I think, and really listen to people in different parts of the globe that are experiencing climate change so much firsthand and listen to their stories and try to have more, more empathy and more compassion.

Michael:

Yeah. I agree with that. I think we also have to recognize that these are really complicated problems,

Aliyah:

Right.

Michael:

And we have to acknowledge right off the bat that fossil fuels have been good for us, you know, in that, you know, our standard living that we have now is so much better than it was a hundred years ago or 200 years ago. Life expectancies are longer. You know, we live a, in a very privileged time in many ways because, you know, we can, we have, we have leisure time. You know, we, we enjoy long lives compared to our great, great grandparents. And fossil fuels had a lot to do with that. Right. But at the same time, I think we have to realize that, you know, it's unsustainable.

Jennifer:

Yeah.

Michael:

That we can't continue a business as usual and continue to pollute the atmosphere in the way that we do. And, you know, we have changed our behaviors in the past. You know, we, we recognized back when I was in high school that that chlorofluorocarbons – the propellants in hair spray and the, the gases used in refrigerant – were destroying the ozone hole. And the Montreal protocol was, was adopted and signed internationally. Led by the United States, in fact. And we reduced the, the production of these to, to very near to zero. And the ozone hole is recovering. Now this is a, that was a simpler problem than replacing our energy production system. But I'm a firm believer that we can do that. But it costs money. I mean, it costs money to change from chlorofluorocarbons to their substitutes. You know, we still have refrigerators, we still have hairspray. We just have it differently. We're still gonna have electric lights and we're still gonna have cars, but we have to do it differently.

Aliyah:

As, as a scientist involved in these complex negotiations involving governments and long-standing financial investments, you're both facing pressures that few other scientists in other fields really deal with as part of their day to day work. What is it like for you personally, to have your work be something where members of the public, politicians, and media are continually seeking your opinion and often disagreeing with you from, you know, how do you sort of handle those pressures and what is it like for you?

Michael:

Never read the comments? 

 

[Laughing]

 

Michael:

Actually, my brother loves to because he, you know, all these funny things they say about me, he thinks they're amusing. It, it is sort of like doing science in a fishbowl. Everybody knows something about the weather. 

Aliyah:

Or thinks they know something?

Michael:

Well, it's a typical opening line at cocktail party. And, and indeed everybody does have some experience about the weather, but climate is different in that, you know, it's not today's weather, it's the sum total of behavior of weather over a long period of time. And so, climate change is really something that unless you live to be, you know, old enough – and I have some uncles who are in their eighties, who, who believe they see the, the effects of climate change where they live – it's, it's hard to really, you know, feel climate change until you really look at the numbers and look at the data. Jennifer, I'd like to hear what your, your thoughts on that.

Jennifer:

Yeah, yeah, exactly. We, I think we do have an interesting position as scientists that yeah, [do] interact more with the public and with media and with politicians. We just do the best that we can as scientists by literally just sticking to the science, sticking to the truth. And just remembering that all these climate models are based on physics. They're based on mathematical equations and chemistry and biology. They're not political models, they're mathematical equations. And then something else that I, I like to do is that I've been trying to talk just about climate change just to family members, to friends, just more and more, and just try to really normalize talking about it and just about very common, different things that we all experience. You know, that we all like to be out outside maybe. Right. You know, and maybe if people like to go fishing or skiing or hiking, just kind of talk about things that are happening with climate change in our everyday lives of, ‘Hey, have you noticed that there's less snow pack this year? Like that's weird, I wonder why.’ This is actually advice from Katherine Hayhoe, who I love, and I love listening to, she's a great atmospheric climate scientist, and she's great at climate communication. And this is also one of her big advice is just try to normalize talking about it. It's, it is a large, global thing. But if we talk about it more often with people, it won't seem maybe so scary. And the more we talk about it, the more we'll start acting and maybe, maybe changing the choices that we make and maybe hopefully how we vote, and things like that.

Michael:

Yeah. I think, you know, as scientists, you know, when we're wearing our ‘scientist hat,’ we need to be very objective about what we do. And, and some of my colleagues feel, you know, that's as far as we need to go. But you know, we're not just scientists, we're also citizens. And you know, this is the kind of thing that, you know, we've spent, we spend our careers thinking about and learning about. And what we've learned is that there's a clear and present danger. And in my view, it would be irresponsible if we didn't sound the alarm. People, people sometimes say, oh, well, ‘You are an alarmist.’ Yeah, it's like, ‘Well, what I see is alarming, what do you want?’ So it is hard for some scientists to be able to speak publicly about the dangers we see, and, and that's okay. You know, not everybody has to, not every scientist needs to do that if they're not comfortable with it, but, you know, if you are, you know, sound the alarm. Whether it be about, you know, hurricanes or ecosystems or, or heat waves, you know, if you see something burning, you know, you should do something about it!

 

Jennifer:

Yep

Michael:

But it is kind of funny when you talk to somebody, you know, who's in some other field, I was, I was talking to some, I was at some computational meeting with scientists from all different disciplines. And there was this guy who had written some interesting papers about gravity waves, and the evidence for gravity waves is pretty strong, but it isn't anywhere near as strong as is for climate change! But you know, he doesn't have people standing up at a room and calling them a liar about gravity waves because nobody in the right mind knows what a gravity wave is. You know, I only barely understand it!

Jennifer 

Yeah, that's so true. But people are like, oh, gravity waves! That's science, that's hardcore science. That is right. We won't dispute it. But you're right. That there is. 

Michael:

So how cool is that? 
 
 

Jennifer: 
 Right, which is great. But there's so much more evidence for climate change.

Michael: 

There aren't that many gravity wave deniers that I've met.

 

[Laughs]

Aliyah:

Yeah. Yeah.

Michael:

Well, it is cool stuff though. I have to say.

Aliyah: 

So, I'd like to transition back a little bit more toward getting into the nitty gritty of these models a little bit. When we were talking earlier you know, you were saying how they're based on a lot of equations that are reflecting physical laws of nature. So not anything that's like up in the air, these are physical laws of nature, but you know, then it's a lot of them being put together. And how do scientists make sure that that sort of overall system is accurate when putting new things into the model? Like, Jennifer, you're talking about how a lot of new Earth systems have been added recently and maybe new sophisticated equations get developed. So I'm just curious on how like a model get, gets updated with all of that? And then how you basically do quality checks to make sure that everything's working correctly?

Jennifer:

Yeah, something that I think is really, really cool is that all of these different codes, all the different equations, all the different chemical reactions you know, are the, the laws of thermodynamics, the Stefan-Boltzmann law that shows that there is a greenhouse effect, you know, all these different equations, once they come together, they make these emergent features in the model. So the emergent features or the models actually creating the storms, the droughts, the amount of snow, the rainfall, you know, we're not telling the model, ‘Oh hey, put X amount of rainfall over India’ or ‘X amount of rainfall over China.’ That's the emergent feature, which is really, really cool that I think it's all the equations then saying, oh, this storm will form, this drought will form. These forests will grow here. So we also want to compare those against observations, which observations have really, really helped us a lot to come back and check, you know our climate model results. And we have national weather stations, at least in the United States all over the United States that monitors for us. 

 

Jennifer:
 So we're able to check, you know, each year, what is the temperature change? What is the what's the relative humidity of the air? And we have a lot of remote sensing data. So data from satellites that show us, you know, like it could show us very detailed stuff all the way down to soil moisture from space. It could show us how forests are changing and growing over time. So we have a lot of these, these observation observational data sets that we'll use to really kind of benchmark or, or test the model against.

Michael:

Yeah. I think the key here in, in how having any confidence in climate models is when you compare to the observations. But that being said, Europe and the United States or Western Europe, United States are, are well observed. Many parts of the world are not. Most of Africa and South America, and large portions of Asia are very sparsely observed. And so that's where satellites actually can play a role in, in enhancing our knowledge of the real world that we can use to help validate the climate models. And of course, you know, the climate models aren't perfect by any means. You know, there are a lot of things that we don't get right. But many things we do. And, when assessing the confidence in our projections, when has to take that into account.

Jennifer:

Exactly. Yeah. And kind of touching on what Michael was saying, how observations aren't in all parts of the globe, you know, is very true. Like I, I happen to work more in tropical forests and in tropical locations. And Michael is very right that in South America and Asia and Africa, we just don't happen to have very long observational records or there just haven't been as many tower data sets or, or these weather stations. So, it can be a little hard. And especially when these big tropical forest ecosystems are so important for the climate system. And I quickly wanted to touch on, there is a little bit of a mismatch between what the tropical forest carbon sink is doing and observations and what the models are predicting, you know, tropical forest can, I could quickly go into this if you want?

 

Aliyah:

Yes

Jennifer:

Tropical forests are so important because they're pulling in all this CO2 pollution from the air, and they're very important for cycling water all across the world. So they perform this rate, you know, ecological service for us of removing CO2 and they're this natural biomass sink. But this biomass sink has actually been decreasing over time. So this sink, you know, maybe even of the Amazon forest, we're predicting that it might even reach zero by maybe 2035 and the forest isn't becoming this carbon sink as much because there's increased forest mortality, more burning, you know, deforestation, the plants are shifting to these shorter lifespans. And when you looked at the predictions from climate models they're actually showing the opposite response. They're showing that our tropical forests, even after counting for the negative radiative impacts of climate change, the models are still predicting a strong, positive land carbon sink, you know, due to a plant productivity, increasing, due to CO2 fertilization in the models. And there's this mismatch. So we never, would've known that if we didn't have observations. We were like, we need to correct that. That's what, that's what we're working on, we're correcting that.

Michael:

Jennifer, this is, this is a great example of what makes science fun. 

 

Jennifer: Yeah. Yeah. 

 

Michael:
 If we knew all the answers, this would really be boring. Yeah. But we don't know all the answers! And, and when you have these situations where, you know, you've got conflicting information from the observations and from the model, for instance, or from theory for that matter… You know, that's why we do what we do. That's why this is a rewarding career. I mean, it, it, there is this aspect that we discussed about, you know, how this is an important problem, but it's also an interesting problem. Yeah. And, and I think that's what motivates people like Jennifer and people like me is, is finding a problem that's hard and trying to make progress on it.

Aliyah:

And so, because you're always noticing either discrepancies or agreement between the models, it sounds like so-called quality checks are really happening, like all, all the time. It's not like someone develops a model and is like, ‘All right, we're good!’ Like , ‘Here it is! Ready for anyone's use.’ It sounds like there, it's always an iterative changing thing. Is that right?

Jennifer:

Completely. Yep. It's always iterative. We're always collecting more observations and more of these hard-to-reach areas. And the world's just constantly changing. It's changing faster cause of climate change and, and humans’ activity. And also with that, as Michael talked on, you know, statistics, we're improving our statistics that we're using with these models and improving our own computational speeds. And something cool that we're doing right now is we're starting to put machine learning into the models. One quick example with machine learning is we're actually helping to predict wildfire modeling.

Jennifer:

So kind of all these emergent processes I was talking about before that come out of the model is the model will predict, you know, the total area of land that's burned and the model uses all these different equations of how much wind is there, how much fuel moistures in the ground, how much vegetation is there that will burn. And those are all really hard things to predict on very fast time scales. So we're using machine learning to help us figure out and predict what that total area burned is by pulling in observational data, in time, with the machine learning with the climate model. And it's in, and it's interactively working together. The machine learning will work, work with observations to help us improve our predictability of wildfires.

Aliyah:

Michael, I understand you, you do a lot of work on extreme weather events and Jennifer, obviously yours is, is Earth systems. So can you both give me some neat examples of things that are, you know, today's super-computer-driven climate models are able to do that maybe you could only imagine doing 10 years ago?

Michael:

Oh, that's, that's really an easy question. For me, it’s hurricanes. 

 

Aliyah:

Okay! 

 

Michael: 

I can remember the first time I saw a hurricane in a simulation. It was probably 12 years ago. I was doing, I was trying to push the model’s fidelity from about 200 kilometers down to 50. And the code kept crashing. Hmm. And I didn't understand why. And I talked to a friend of mine who had run it at this resolution and he says, well, go ahead and look at the simulation. So I made a visualization, an animation of this. And I was looking at temperature at 500 milli bars, which is, oh, fairly high up in the atmosphere. And I saw this blue dot moving across the across my screen, across my map.

Michael:

And I realized immediately what was happening, that this was a hurricane. And of course, at that time, the common wisdom was climate models don't make hurricanes because they're not high enough resolution. And of course it turned out that that was, that was sort of a stepping stone that, you know, we then started running at 50 kilometers. And soon after that it went down to 25 kilometers and it's become a whole new generation of climate models that I call ‘tropical cyclone permitting models.’ You know, the, the hurricanes are there. They've got a lot of realistic properties. They don't have everything, of course, ‘cuz 25 kilometers is still pretty big for, for a hurricane. Nonetheless, it looks and smells like a hurricane. And so that's enabled us to do lots and lots of things. The attribution was one, but also looking at the long-term trends.

Michael:

So attribution is usually said is with another word called ‘detection,’ we call it ‘detection and attribution’ or DNA for short. Where we, we first go to observations and detect some sort of change. And then we go to the models and run the models with, and without climate change to see if there's a signal. If, if we run it without climate change and we see that detected trend, then we say, well, it wasn't climate change. It was something else, natural perhaps. But if we don't see it in the worlds that don't have climate change and we do see it with the world that do have climate change, and after a lot of fancy statistics, we'll come up with a statement and say, ‘We can attribute this phenomena to human changes to the composition of the atmosphere,’ usually. So what happened when, you know, we got computers powerful enough at the National Energy Research Super Computing center at Berkeley Lab – NERSC – that really changed things, you know, so now we could do these kinds of studies and, and make assessments of how climate change affects hurricanes.

Michael:

And so it is interesting. It didn't answer all our questions. The question I think that, that these models have really answered pretty definitively is that the most extreme hurricanes, the category four and five hurricanes will become, become more frequent and more intense with global warming. The question that's still open and there's been a lot written about this by me and a lot by others is how many hurricanes will there be, you know, overall. So category one through five and the tropical storms, which, or named storms, so how many named storms will there be in a warmer world? And the models, almost, almost all the models suggest that they'll be fewer, really. And, there are reasons for that that I think are true. But not everybody believes that. And so again, this is one of those things where science becomes fun because I say this and my friend Carrie says that, and they’re not the same. And at the moment, we don't know. And so that that's something where computer models play a big role in, in increasing understanding and, and the fact that computers continually get more powerful and, and faster and bigger really drive that.

Jennifer:

It's so funny, Michael, I think I remember that time period as well, [when] you and your group simulated an actual hurricane. I was like, ‘Oh my goodness. Look at that?’ Like, you could see it. I mean, it is really remarkable that these models can do that. Cause like I was mentioning earlier, earlier, we're not saying it put a hurricane here over the Caribbean, like it just does that. It's like, wow, it did it.

Michael 

Yeah. Outta thin numerical air! I mean, it was, it was definitely an aha moment. 

 

Aliyah:

Wow. 

 

Michael:
 And it was, it, it, I remember it very clearly.

Aliyah:

I think that shows how complex these systems you're working with, that, you know, 12 years ago, the technology couldn’t even do that. And then today still you are digging in and analyzing what it means and, you know, kind of adjusting the models to have that capability. And that the answer, you know, it hasn't been solved yet.

Michael:

What what's been rewarding about it too, has been involving some young scientists: training young scientists who have taken this technology and gone much farther than, than I ever possibly could have imagined. And so, you know, it's not, it's not just all about me. I mean, it's about, you know, us as a community, and to see that has been an extremely rewarding experience at this point in my career.

Aliyah:

Jennifer, have you had any of those like aha moments with, with models in your career, where you're just like, ‘Wow, I can't believe we can do this to now! I can't believe that it's showing this?’

Jennifer:

Yeah, yeah. I have. And I think kind of, as Michael was speaking to this, it kind of goes hand in hand with our computing systems also allowing us to have this greater resolution in the model, this greater fidelity. And earlier on, when we were talking about the history of, of climate models, you know, our computers were also more basic and they have really progressed over time where we have these very sophisticated supercomputers. And the specific climate model that I work on, it's the Department of Energy funds it and runs it and develops it. And we're able to use Department of Energy Super Computers, and we're creating our model able to run on a brand new type of supercomputer called an exascale, and exascale computers are, are almost ready.

Jennifer:

They're gonna… what it refers to, exascale computing, is computing capability of at least one exaFLOP. So this is 50 times faster than the fastest super computer around today. And it'll just allow for so many scientific discoveries. I touched on this earlier when I was talking about [how] I model more terrestrial ecosystems, that at first they were just treated kind of as a green blob, just like one, one big leaf, basically that respires. Like the whole, the whole ground really was just a leaf that respires oxygen and CO2 and water vapor and all that. But a big kind of aha moment for me is now with supercomputers, we can really simulate individual trees across this entire planet. So individual tree processes, even with multiple canopies, with lots of leaf layers. And what's really cool, now the model that I'm specifically working on is that these trees will compete with each other, for resources, for water, for space, and they grow and they die independently on their own in the model and they'll grow and die and compete at different rates with climate change or with disturbances, you know, or with land use change.

Jennifer:

And a cool thing that we're bringing in in the model is that these plants are going to be able to grow in new parts of the world. So, so with climate change, you know, the biomes are shifting. These biome boundaries are shifting. The Arctic’s becoming warmer. So shrubs are going further north into northern latitudes, [and] you know, [the] snow pack is less. So tree lines are going up higher into the mountains. You know, tropical forests are being pushed into new hotter climate regimes that they've never experienced before. And we don't know what this new world will look like. And so, our group that I work with, all the terrestrial ecosystem modeling side is we're starting to, we need to account for that

Michael:

Jennifer, that was the one that's like, ‘I didn't know they could do that!’ Yeah, yeah. That was the one, that was the moment for me.

Aliyah:

Oh, that's great. So one thing that I've been curious about from listening to you to both talk is, you know, and I think a lot of people listening, don't… It's hard to conceptualize what these climate models really look like, and how you interact with them. And, in my mind you're logging onto a computer and it's like ‘EarthSIM Version One,’ and it's the entire Earth. It's like a full model and you're both using the same model, but I'm guessing that's probably not right. So, what are the models actually? You know, how are they sort of divided up? Is there one big one that you can work on little parts of? And how many of the sort of big, really well developed models are out there in the world? Is it just a handful or are there actually a lot of them?

Michael:

That's a really good question. There, there are a lot of them there. The question about how many of them are independent is, is much more complicated because a lot, lot of models are derivative of some other modeling center. 

 

Aliyah:

Okay. 

 

Michael:

So, for instance, the model at the National Center for Atmospheric Research has many descendants, including the Department of Energy's E3SM model. So they're closely related. Similarly for the UK Met office model from the Hadley Center, very successful model, and it's been propagated to other countries. So, the Australian model is closely related to that one. I may be the last person who actually wrote a climate model from scratch. 

 

Aliyah:

Really? 

 

Michael:
 I did it. Yeah, I did it when I was much younger and, and quite a bit crazier. And it was about 300,000 lines.

 

Aliyah:

Wow!

Michael:

Of Fortran. And we abandoned that model when I froze was the entire United States.

 

[Laughing]

 

Michael:

And it I think I could have fixed it, but it seemed like, it seemed like the days of one person writing a climate model are basically over. And so, so climate models are generally written by large teams of people. And so you know, so there, so there's two groups of people. Actually, there are three groups of climate scientists. There are the people who write them, who are usually the unsung heroes, because they're writing Fortran rather than papers or C or Python, or what have you, rather than papers.

Michael:

And so they're not highly published. They're sort of unsung heroes and who, who are not always in the forefront. Right. They don't get on TV for instance. And then they're the people who make, who make the runs, and then they're people who analyze the runs. Okay. You know, there may be a little more overlap on that. And so, one of the things about climate science, that's maybe different than other science efforts is that there's a large database of simulations that are publicly available, that anybody can get. And so there's a large community of people that never run climate models, but analyze them.

Aliyah:

I see.

Michael:

And so, it's different than other fields, you know, where, you know, the models are much more proprietary.

Aliyah:

That's cool. I really appreciate that thought when something's so important. It's nice that it's, there's more access and more openness.

Michael:

The Department of Energy can take some credit for this. This goes back to my days in Livermore, there was a visionary scientist named Larry Gates. And Larry Gates started something called the Program for Climate Model Diagnostics and Inter-comparison. But he went out to the various modeling groups, climate modeling groups, and he said, ‘Look, you got these great big climate models. They make all this data. You don't have enough people to analyze all this. Why don't you let us distribute the data to the scientific community? And you know, you'll get some credit and we'll learn something.’ And that became a project called the Atmospheric Model Inter-Comparison project, we, which is morphed into something called the coupled Model Inter-Comparison Project. And now we're on version six of that. 

 

Aliyah:

Wow. 

 

Michael:
 And it's become successful, I think, way beyond what Larry even dreamed.

Jennifer:

Yeah. Yeah. I'm so happy. You, you brought that up. I was thinking about bringing up the Coupled Model Inter Comparison Project and it's, great that it does bring together all these different modeling centers from all around the world. So anybody can look at it and compare them to each other. And I think in total, are there something around 40 ish or so?

Michael:

Yeah. About 40 yeah.

Michael:

You know, and again, okay. Yeah. It's question about how many are independent, but, but the beauty of this system is that any graduate student at any university can get on their computer and download this data. 

 

Aliyah:

That's great. 

 

Michael:

And, and that means they don't have to have the great big monster computers that we're blessed with at Berkeley Lab. You know, not all these analyses can be done on a laptop, but a lot can, and a lot are. And you know, thousands and thousands of PhD theses have been based on this that would never have been possible if Larry hadn't had this idea to make this important data public.

Aliyah:

Yeah. That's, that's great. And I’m so happy that someone was there at the right time, in the history of this field to yeah, make that decision to say that. To give, to put us where we are now. And so, are you both working on the same model or different parts of the same model? Could you even call it the same? When you do, like Jennifer, when you’re working on investigating the carbon sink in the tropics, and Michael, you're looking at hurricanes, is that the same model?

Jennifer:

Yeah, well, something that we do is that we can work in different parts of the model independently from each other, which I think is really cool. And we've set up the code in that way. So again, this is all just computer code. It's all written in Fortran, like Michael mentioned. And it's, it's funny that we're still in the old school language of Fortran. You know, I hear these, these new fangled languages are out there somewhere, but we're still stuck in the, for Fortran language of the 70s. But yeah, so we can, 

 

Aliyah:

If it works!

 

Jennifer:

Yeah. It's, it still works. And that's what we usually started all these codes in is Fortran, but yeah it still works. 

Michael:

It's still the fastest.

Jennifer:

Exactly. Thank you! I try to say that as well. Michael, it's still the fastest. 

 

Aliyah:

Hey, that's nice. 

 

Jennifer: 

Yeah, but yeah, but I could work in just kind of land component system. So I could work on things with the soils and the forest and soil moisture while other atmospheric scientists can be working on atmospheric chemistry. And there is this kind of ‘coupler’ part of the code that can bring together all these different components, that can bring together the land or the atmosphere or the ocean, and make it all work together as one coupled system. So independently we can be working in different parts simultaneously.

 

Aliyah:

To close it out one last quick question: what skill or, or talent, or even just, curiosity, what kind of questions do you think the next generation of climate scientists could bring in that would make this even better? Grow further?

Jennifer:

Ooh, that's a good question. I'm gonna pose one that might be a little challenging. And I don't know if any new young scientists will want to take it on because it's dealing with humans. And I don't know if people want to deal with human decision making and human actions, but that's where I think the next direction, really, of climate modeling needs to go is – we're starting to think about this just a little bit – is to try, starting to add in the human element.

Jennifer:

So yeah. How, where, where humans are deciding to live, where their cities are growing, you know, our economic choices, how GDP is changing and you know, which cars we're driving, where we're emitting things and that feeds into the climate system. And, and that's a whole nother can of worms. That is very, very challenging. Because humans are, you know, finicky people. We don't know where we're going to live or what choices we're going to make in the future. But if I think that that could be something really interesting, because I mean, it's bringing in kind of social sciences and with the physical sciences, and this is kind of where we're needing to go also with, with trying to address the climate crisis is trying to account for what humans are doing. You know, like how we're going to try to mitigate climate change. If we're not mitigating climate change, you know, what are the adaptations that we need to do? So yeah, that might be my, my advice.

Michael:

I actually have a very similar thing to say. I think that the interaction of the physical climate sciences with social climate science is really an exciting new area. If for no other reason than the language of social scientists, science, is different than the language of physical science – so there's a communications problem that is actually interesting to try to bridge. 

Aliyah:

Yeah. Yeah. I think that's a great point. A great thing to end on, is just, you know, with so much science and science communication, it's about letting people see themselves in the science, how it affects them. So in the future, people could sort of see where they are in these models and maybe that will make a big impact. Yeah.

Michael:

Yeah. I hope so. I hope so.

Aliyah:

Great. Well, thank you both so much for being here. It's been a pleasure to speak to you both.

Michael:

Thank you.

Jennifer:

Yeah. Thank you. This has been really fun. Thanks.