Ecosystem loss, extinctions and climate change are ongoing challenges to life on Earth, and coming up with a plan to tackle their effects requires an accurate picture of what's happening where, and who is involved.
Prof Tim Haas has taught and refined such models for years. In his latest paper, he lays out the case for a model unifying human behaviour, climate and ecosystem data, the computational power required to run it, and the credibility criteria any model should meet to prove its worth.
Read the original paper at: https://doi.org/10.1371/journal.pone.0226861
Ecosystem loss, extinctions and climate change are ongoing challenges to life on Earth, and coming up with a plan to tackle their effects requires an accurate picture of what's happening where, and who is involved.
Prof Tim Haas has taught and refined such models for years. In his latest paper, he lays out the case for a model unifying human behaviour, climate and ecosystem data, the computational power required to run it, and the credibility criteria any model should meet to prove its worth.
Read the original paper at: https://doi.org/10.1371/journal.pone.0226861
This is an automated transcript.
00:00:11 Will
Hello, I'm Will. welcome to research pod.
All models are wrong. But some are useful.
This quote is generally attributed to the statistician George Box and predates the crisis of science communication. We are currently facing where pressing issues such as pandemic response, climate management and ecosystem loss rely on the accuracy and acceptance of mathematical models.
Professor Tim Haas has taught and refined it. Such models for years. In his latest paper, he lays out the case for a model unifying human behaviour, climate and ecosystem. Data also lays out the computational power required to run it and the credibility criteria any model should meet to prove its worth.
And despite all of his research topics somewhat bleak nature, he's remained an optimist. Our interview began with a quick chat about how he wound up in his current position, bridging topics of ecology, behavioural psychology, business and statistics.
00:01:13 Prof Haas
I’m an associate professor at the Lubar School of Business at the University of Wisconsin, Milwaukee. I have worked in environmental statistics pretty much my whole career.
Before I went back to Graduate School, I was an engineer in the high speed disc Industry in Los Angeles, but I decided that I wanted to pursue research questions that I had about how.
Humans make decisions in groups and that led me eventually into my degree in statistics. I was hired because of my interest in modelling human behaviour. My dissertation was on the handling of the Cuban Missile Crisis.
00:02:03 Will
So yes, let's move on to your work about the threat of mass extinction...
00:02:10 Prof Haas
I had a nickname years ago at a research laboratory. They called me Grim Tim.
At that point you start wearing large black cloaks son walking round with a side surely.
00:02:26 Prof Haas
Since my tenure in 1995, I've been developing models of how humans interact with while.
Life the literature on biodiversity is very dark if you start looking at the some of the papers because that the people in ecology realised that there is cause for much concern about all the animals on the planet. Other than humans, cockroaches and crows.
Who's the main driver of this issue of biodiversity loss? Is is habitat destruction, but the second driver that's important is poaching an you see poaching in many forms. For example, a hardwood exotic hardwood trees in in the Equatorial regions.
These trees are being cut down illegally, logged at at a rate that is so fast that a lot of forestry people aren't sure that these exotic hardwood species will survive.
My work has focused on large mammals. Some people are critical of studying the survival of large mammals because they call them charismatic species and they say what you're doing is just trying to get attention by appealing to people's emotions, because when they.
Look at poor little koala bears in Australia an they listen to the World Wildlife Fund commercial. They say that that this is appealing to emotions. It's not looking at the biodiversity of allsorts of animals that maybe aren't so charismatic on TV.
But my approach has been more to find data sets that are in the public domain That I can use to experiment with.
00:04:27 Prof Haas
I am building my my models of how humans interact with wildlife.
My main concern is to learn how to build these models that are.
They're believable and and less about a particular animal, so I've spent a lot of time on Cheetah in East Africa an the rhinoceros in South Africa, but that's only because I've been able to acquire data without having to spend lots of money that I don't have to get proprietary data sets. I was very fortunate about. Nine years ago, to build a relationship with South African national Parks and I've worked with the ecologists there.
To try to understand how the rhinoceros population is being affected by the poaching of that population by folks who sell the rhino horn to East Asia.
It's a cultural value and it's driving this poaching crisis of the rhino in South Africa.
But before I entered into that relationship, I had been able to acquire some data on Cheetah population numbers in East Africa, and that has led me to.
My first book, an an now a couple of papers including my most recent one on the effects of Habitat Destruction and poaching on the Cheetah population in East Africa.
It's been a very hybrid career that I have pursued, but I've been very fortunate that the Business School has accepted my work in environmental statistics in environmental decision making, even though it isn't directly research in business.
00:06:32 Will
As much as you've mentioned with the epithet of Grim Tim looking at statistics, extinction numbers and political data for an entire day, how do you keep your spirits up?
00:06:45 Prof Haas
I look at the world as a work in progress and we have now many many tools and capabilities that we could use to craft a planet that is equitable and diverse.
And we don't have to wreak the destruction that is so far.
What we see, so I may be naive, but I'm hopeful that once we harness the capabilities that we have developed.
Ann, focus them in the direction where they should go, that we can reverse these trends.
An we can make the world better so.
I look at the the disturbing downward trends of survival of different species and it is troubling and that is true. But you know, I've I've had a background where I've had a lot of challenges in my life and a lot of times things didn't look like they were going to workout.
And I guess my attitude towards the world is bring it on, so to speak. Things are not Great, but there is a lot of room for hope, so I kind of stomach the grim news because I am working on solutions. I think that might be the difference. It's kind of like a doctor seeing a patient come in who is in bad shape. The doctor has some ability to make things better.
A bystander who was not involved in patient care, I would understand, would be prone to just major depression that this poor person is in trouble an feel powerless, but I guess the difference is I don't feel powerless.
That might be the one thing that keeps me from being really depressed, because I feel that My efforts in a very small way might be able to help things, but when I look at bigger pictures and I've noticed this recently, obviously that I feel a sense of of helplessness in in in watching the news, and it's kind of a new feeling for me because I've always felt like I could do something.
Things look better now, but I I was noticing in past weeks that for maybe the first time in my life I felt like there wasn't hope. And I can see that a lot of people looking at the biodiversity trends would have a sense of despair, and I think it's well founded. But, you know, humans need to realise that the future is definitely in their hands.
In fact, the entire ecosystem of the planet is in the hands of humans, so whatever happens in the future will be caused by humans. That despair is connected to a sense of deep responsibility that every single one of us on the planet shares.
I'm working on ideas currently with how to harness this sense of of need to do something across large numbers of people who are not formally connected to political institutions. Because there's an awful lot of anguish going on and anxiety among people in the world who feel that they have no ability to affect change.
That needs to be addressed in a way where people can feel like they're doing something to make things better, and I think that would relieve a lot of this sense of despair and hopelessness alone. Among people who are concerned about the direction of the planet is going.
00:11:04 Will
Well, I suppose if you've modelled peoples behaviours and you understand there's a chance that we might be able to steer it feeding into policy.
00:11:12 Prof Haas
Yes, I think that the work in climate change is a super example of trying to get the word out that the planet needs some help.
And it's a lesson for all of us in how to try to Mobilise a large number of people on the problem that is planetwide and that has led to this paper that I have just written because what has happened to the climate movement.
It's not been able to convince everyone.
Roughly 30 to 40% of the people on the planet believe that climate change is going to be a bad thing.
Why doesn't everyone believe that? Why doesn't everyone believe that climate change is bad?
That led me to looking at this paper and developing some ideas of, well, how do you get a credible message across about something which is as complicated as a process that is both ecological and social?
And that the two interact. I think that that is a process That humans haven't directly dealt with very much up until now.
00:12:38 Will
Now, when it comes to trying to put together a model of politics and ecology and sociology like each one of those things by itself is not static. There's a lot of give and take and flow, and changing all of those. So how do you put all of those together into a unified approach?
00:13:00 Prof Haas
So about oh, 20 years ago I started to ask that very question of myself.
When I was an engineer, I had had to learn about dynamic systems. Often they used the phrase dynamical systems and these are systems which change through time.
Engineer is know a lot about how to model physical dynamical systems and the very first models were of simple things called, you know, just simple mechanical devices, the first one being just a spring.
When you stretch a spring and let it go, it oscillates through time and you can actually write down some mathematics that describes the behaviour of that spring through time. So I started asking myself, well, what sort of things happen with humans and wildlife.
An I start and I was completely mystified that I was thinking where would you even start with something as complicated as that, and I started with the newspaper I said to myself I don't know what kind of data I should collect on humans and wildlife, but I know where data is happening.
And that's in newspapers in countries that have wildlife. So I started to just watch all these online newspapers in East Africa and write down the storeys that they wrote about wildlife and how humans were impacting wildlife.
And I just kind of basically went to school of how humans interact with wildlife and I started being this newspaper reader. I started to notice some reoccuring actions happening.
An being reported in these newspapers. Of course, one of them was the storeys of poaching.
So I started to develop a model Of how a human might decide to go poach an animal, shoot an animal anu's take their skin or their teeth, or their bones for some cultural or or pragmatic purpose.
Odd, so the newspapers led me to form my data sets and they also led me to the kinds of things that my models should be able to receive an produce. So it was kind of a pull yourself up by your bootstraps. Approach to research.
00:15:44 Prof Haas
The ecological side was much easier because he ecologists have been studying wildlife populations for 200 years. So the kinds of data on wildlife that was much easier to get ahold of an there were models of water called population dynamics. Models of wildlife. These models had been around for a long time and it was easy for someone trained in mathematics like myself.
It was easy for me to find those papers, read them, and write down their models. So now the only question was how do I glue the two models together? So that took years for me to figure out how to make the two models. Basically talk to each other.
And my work with the folks in South Africa really helped me learn how to place the poachers inside the rhino population dynamics models.
So now I had the poachers impacting the survivability of rhinos through time and the person I worked with down there, Doctor Sam Ferrara.
I remember one night he says to me he says, well, Tim, you've connected the poaching with the ecology. And I remember noting that he thought that that was kind of a new idea.
00:17:21 Will
That's interesting, seeing as my understanding of poaching is part of kind of the life cycle of humans in an ecology.
00:17:31 Prof Haas
Yeah, you have a whole field called human ecology. There are a lot of papers now being written that.
They are doing the same thing. I've been trying to do and that is to say humans are part of an ecological system. You can't avoid writing down some equations which represent their activities just like your writing down equations about how rhinos or wildebeest es are surviving.
But there's a lot more people saying we should do this. Then there are people who are actually doing it.
Then there's no wonder why that is, and it's largely because of the Western University education system. If you look at it, and that is a graduate student and there is a Department of Ecology or a Department of Political Science, and they are encouraged to write a dissertation which is exclusively in that domain.
They are not encouraged to write dissertations that bridged more than one Domain. Because the idea of a PhD dissertation is to prove to the outside world that you're capable of doing research. So you are discouraged from crossing boundaries. But the world is all interconnected across these disciplines.
So you don't get too many people that are well trained in more than one field.
And that discourages building models of, for example, how humans interact with non human ecology. You can say that humans are part of an ecological system. That's great. I totally agree, but show me some mathematical models of how you have implemented that idea or that concept.
Hard pressed to find actual mathematical structures that contain equations to tell you how humans will interact with non human ecological systems.
00:19:56 Will
Let's get into some of the specifics from the model in this paper. Then like to start off with something that you mentioned earlier with regard to climate change acceptance. I'm not going to deny ability, but climate change acceptance. The idea of credibility in a model.
00:20:12 Prof Haas
So in science. You see the scientific method? Well, the scientific method. Really works as follows.
One scientist builds an experiment in a laboratory. And they generate some results.
And they write a paper and they say I did this. That and the other thing in my laboratory and I got the following output.
Other scientists read this paper and say this person made a mistake. I'm sure he did, so I'm going to build my own experiment. That this person has described in their paper, and I'll prove this person wrong, so they go ahead and do that, and they discover that the first person was right.
So that's called replication of results in science builds on replication. Earth results from experiments in laboratories.
That's how quantum physics got discovered. If you look at the history of quantum physics, you discovered that there were a whole bunch of scientists, all of whom didn't trust each other and were convinced that the other person was wrong. But they were able to replicate their experiments in their own laboratories, and eventually they grudgingly admitted that the first group of scientists were right.
Once that happened, quantum physics became a fact.
The problem with climate change is there's no laboratory.
If you had a laboratory where you could duplicate the proclamations out of the Hadley Research Centre. About the temperature is going to go up 10 degrees in 30 years. Then eventually all the scientists in the world would say this is correct.
You can't do that with climate change because it's about something that hasn't happened yet. So what do you do? You build mathematical models that predict what the temperature will be in 30 years, and then you say to the world these models are correct because really famous scientists built them.
People outside the scientific community say We're not impressed, we're not. In awe of these scientists, so we don't think they have any credibility. It's it's the old game of the priests say such and such is true. Therefore it's true. The problem of climate change research right now I see it is you have a large number of people who don't respect these scientists. So what do you do?
I said to myself, you show that the models match historical temperature data.
You can then say look my model tracks the history of temperature really well.
There's no good reason to think that it won't track the temperature into the future.
You don't have to use authority to believe me.
Who cares if I got my degree at Cambridge or at Cornell? It's irrelevant. My model matches temperature data from 1850 to the present.
If you could say that people would say, wow, that's kind of like how physics got built.
So what I think the climate change crew should be looking at is publish your source code of your climate model. So that anyone can run your model. See for themselves what you did to build that computer programme.
00:24:16 Prof Haas
Also publish so that everyone can find it all the temperature data from about 1850 to now and all of the greenhouse gas emissions data.
Put all this in the place where everyone can get to it.
And then use the five techniques of my paper, which I didn't really invent. They're just good scientific practise and subject the model to these five tests.
If the model passes all five tests, proclaim to the world that there is no data based reason to not believe the models.
You don't have to be in awe of the scientists who built them. The models do a good job of matching historical temperature data. Therefore, when the model Sai in 30 years, we're going to be cooking the planet. You should believe it.
I think a lot of sceptics of climate change would have to drop their argument that it's simply an attempt to have poor countries get money from rich countries, which is currently one of the big conspiracy theories surrounding climate change. I think these people would have.
A lot less of a Position to stand on. If these models were shown to be Reliable in the face of actual temperature data.
00:25:49 Will
It would be nice to think that honesty and openness should and possibly could be the cornerstone for a deliberative process in science, and getting everyone involved scientists or not.
00:26:01 Prof Haas
Yes, see then you could get all these giant corporations to say we are both the problem and the solution. We are going to use our vast resources to shutdown the generation of greenhouse gases and replace them with technologies which we already have.
00:26:23 Will
To challenge you on that, that does require honesty and credibility from there half as well.
00:26:27 Prof Haas
Yeah, and see, that's where I'm kind of naive. I understand it. I'm saying that normatively if everything is open, people will be honest.
That may be a little too naive. I understand the concern there but but what other options do we have for going forward? Well, yes.
00:26:53 Will
When it comes to the specifics that you mentioned of having your data, your tools actually getting that work done leads to the topic of many tasks. Computing that came up through this paper as well.
00:27:07 Prof Haas
Yes, so a tiny bit of history about me. I spent the last six months of 1999 at the National Centre for Atmospheric Research in Boulder, Co.
As a visiting scientist and I worked in a statistics division there and so I got to know some of the principle investigators.
Who were responsible for building one of the five big climate change models in the world?
So I came away from that experience, having learned one very important thing among others. But one most important one, and that is these climate models require pro digus amounts of computing unbelievable amounts of computing. When I was at end car, they had this.
Oh, I guess 10,000 square foot basement in their building. That basement was chock full of 1 giant IBM computer. It was awesome that computer.
Could barely run three months of climate model simulation in a week. My paper get this. My paper says not only do you have to run that climate model over 30 years, you need to run it 100 times over the same interval of time. We have nowhere near the computing power that.
Can do that at this point in time, the year 2021. So my paper if you really look at it sceptically and you know something about compute.
You would just have this grim smile and say what this person is proposing can't be done because the computing requirements for big simulation models such as climate change.
Are way too expensive to run to even think of running them repetitively 100 times or more. But the statistical methods in my paper require that you can run the model multiple times over the same interval of years.
So my paper is really extremely edgy.
And challenging to the scientific community because what I'm saying is, if you want to do credable model building, you're going to have to fit the models to data using statistical methods.
But statistical methods require Lots of replication an for a big model you have to run the model many times, but climate models an my cheetah model in my paper they require. Days or weeks of computing. So my only hope Was we need to harness many many computers and have them work like horses in a team of horses.
So a cluster computer is kind of like a big chariot where you have lots of horses hooked up to it. Cluster computer takes advantage of many small computers. But they are all. Orchestrated to work on little chunks of the same problem.
And by doing that they can take an enormous computing problem divided up into little pieces and solve it by all of them working at the same time, but cooperatively. So my paper envisions. Computers that are 100 times bigger than the computers we have now.
All working on making models credible to people who don't want to believe them. That's really where the human race, in my opinion, needs to go because we now have control over entire ecosystems.
But we have lots of people who want to ignore the detrimental effects we're having. My way forward is create models that through subjection to these statistical methods an massive computing win over the sceptics, and now we can Mobilse their energy and money To fix the problems.
00:31:44 Will
I think you might need a little bit of marketing to put the Polish on to really sell it to people, but fingers crossed.
00:31:51 Prof Haas
You're absolutely right. Well, I, I think that the the actual implementation is very challenging.
00:32:01 Will
Imagine that human beings are in control of entire ecosystems pretty much globally, and there was another part of the paper about ecosystem management tool procedures.
And it was particular sentence that stood out to me about task based parallelism. Technology requires the following nine characteristics. And was that what you meant by having all of these distributed computers working as cooperatively as people should be working? Or is that leading onto something else?
00:32:34 Prof Haas
That's a good way that's I hadn't thought of that analogy, but you have a good point in the refereeing process, which, by the way you can read all of the referee concerns and all of my responses by clicking one of the buttons in this Open Access paper. One of the concerns of the reviewers were coming back to over and over again was too for me to justify why I chose a technology called Java.
And because of this concern of the reviewers, I went into a very thorough study of the present computing technologies that are available.
And the the challenge of getting lots of little computers to work together in cooperation that challenge has only recently been addressed in a way that is effective.
And that was the technical part of my paper, and that was actually why I gave the paper the title idea.
So when I teach students, I'm always trying to give them tools that work, and sometimes my students don't like this because it means they can't just sit back and think conceptually. They have to get their hands dirty and work on a computer, but to me, unless you can demonstrate that you have the necessary tools that work.
You're just talking, you're just coming up with issues that maybe other people will solve. I'm never happy with that. I guess it's my engineering background. Unless you can come up with a suite of tools that you can show will do the job. I don't think you have much to contribute, so I spent in my paper allot of time reviewing free computer tools that anyone can download. And use them to harness the power of many little computers who can talk to each other.
And so that the idea of using Java spaces, or maybe if you like to use the new Berkeley invention of Ray, go for it. Pick one I in my list of of I think 5 technologies is not exhaustive, but I did make sure that all the ones I listed were free.
Anyone who has an Internet connexion an knows something about computer programming could, with a little effort, be up and running in maybe two months.
00:35:20 Will
Is the case study at the end of the paper about cheetahs EMT simulator kind of the proof of concept for this then?
00:35:27 Prof Haas
Yes, yes, that is exactly it it was.
It was a proof of concept. It wasn't to try to fix the cheetah challenges, it was simply to say look here is a real.
World human wildlife problem. I have data an I have a model so I'm going to put my money where my mouth is an I'm going to take my methods and an exercise them on this real world human wildlife interface problem and that was simply to demonstrate that my ideas can actually be used and come up with a practical incredible model.
And one of the methods which I've considered very key is this idea of having sceptics propose.
Alternate model output.
And the model are being forced to answer that sceptics concerns.
What I've noticed with climate change modelling is the lack of effort on the part of the climate change models to respectfully address the concerns that sceptics have. Instead, what you see is the climate change community.
She placing sceptics on the Greenpeace lists of climate deniers.
Well, that's going to get us nowhere. If you ostracised a person because they are critical of your model, that's not going to shut that person up. That person is just going to be resentful more angry an more vocal about how they don't believe in your model. So what I tried to do in my paper was to bring the sceptics to the table and say, look, why don't you come up with a scenario that you think my model. Can replicate. That you don't want to believe.
So here's one for climate change guys. Let's say you could find the climate model. That predicted no temperature increase. If you could do that, you could say that the climate models have nothing to bring to the policy discussion that would be damning. No ones done that.
00:37:56 Will
I feel like we should weigh in with that. Well, there are possible credible reasons for someone to challenge the findings of climate change models if they are coming from a place mentally or morally of that climate change was a lie invented by lizard people from the centre of the earth to inject you with 5G, then those kind of concerns could be quite, I'd say, comfortably disregarded.
00:38:20 Prof Haas
You know this this, uh, epistemological approach of discounting the message because you can discount the source. I'm sorry that that doesn't fly with me. I don't care if a serial killer comes up with a critique of a model that holds water.
It's unfortunately going to be true.
If a sceptic and say look all you have to do is adjust these 10 parameters in your climate model an you'll get no temperature increase.
Let's say you could do that and you got no temperature increase. Then the sceptic would say look it's easy to tweak your model to give no temperature increase, so who cares about your model?
I think that result would be damning. I think that would that would require the climate change community to admit that their models have nothing to say about climate change.
00:39:17 Will
If I win, that challenge can be like you say, statistically, an rigorously levied, then that would be a turn up for the books.
00:39:25 Prof Haas
An I think that that if that was successfully met to be more positive.
Then those sceptics would be silenced. It's an opportunity. My paper represents an opportunity to the climate change community, however. Well, I might mention early in the refereeing process, my paper had more explicit commentary on climate change models. I was encouraged by a sympathetic referee to remove that material.
Which I did and he said. Or she if you leave this in, although the person didn't say this in so many words.
You will get so much push back that this paper will never be published.
And so I very diplomatically removed most of the commentary on climate change models an I restricted myself to wildlife interaction models. But if you look at the title of the paper an you know something about.
These models the phrase political ecological. What is a climate change model? But a political ecological model? I don't know how many people are going to notice that, but it's as plain as day.
00:40:57 Will
As it stands, the paper is out there. Your tools, your models, all of the data that you've put into it is out there and like you say, the paper is free to access. Who can join in with this?
00:41:09 Prof Haas
In my positive view of the world, I think that if lots more people got involved in the issues surrounding building these models, and in particular building models that have policy relevance.
That we could start to. Broadcast the idea that we are living in a very interconnected world with the non human. Gang, all the critters and plants of so. I think anyone who is doing analysis or policy making using science and that is everybody from Environmental Protection agencies through Fish and Wildlife institutions an and of course across all the universities you have 10s of thousands of scientists and programmers, all of whom could take the ideas of this paper to heart.
And not just build models, but make their models credible.
In my paper I have an explicit way to identify the most practical ecosystem management plan.
Which is another way of saying an effective policy for managing an ecosystem so that it doesn't just become a wasteland, and so that my idea is the last.
Take home of my paper. It's in there.
So the key message of my paper is simply. It's not enough to make a model, it you've got to make it credible an my 5 methods that I given the paper I think is the way to make a model credible and hence the policy recommendations might actually be implemented in the real world.
00:43:09 Will
If we don't do that, if we don't get our house in order, what is the alternative?
00:43:15 Prof Haas
I teach a course in business impacts on biodiversity to graduate students an if we don't do something.
We're going to kill off our reefs. There will be no large mammals left an it'll be pretty much us. Cockroaches and crows do we want to leave the planet to future generations?
Including our own children that is devoid of anything other than 12 billion people, zillions of cockroaches and lots of rats and crows.
Do we really want that to be our future?
And lose all the wonderful diversity and beautiful things that we have on this planet that make it so valuable.
Humans can make the planet either a paradise or a big junkheap. We have the power to decide.
If we don't start doing something intelligent about managing the planet, we're going to get a junkie, and that's entropy. In physics, the idea that entropy down deep philosophically says everything will go in it towards the direction of less complexity.
So the loss of biodiversity is going in the direction of less complexity. The only thing that can stop that is adding intelligence.
00:44:44 Will
I hope as many skywatchers do, that there is someone intelligent out there listening to this.
00:44:52 Prof Haas
Me too, I really appreciate this opportunity for you helping build this podcast.