Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ


Choose your language

InfoQ Homepage Presentations Risk of Climate Change and What Tech Can Do

Risk of Climate Change and What Tech Can Do



Jason Box and Paul Johnston briefly share several bold visions to slow down the pace of climate change to buy time and save lives.


Jason Box is Climatologist & Professor in Glaciology at The Geologic Survey of Denmark and Greenland. Paul Johnston is CEO at Roundabout Labs.

About the conference

Software is changing the world. QCon empowers software development by facilitating the spread of knowledge and innovation in the developer community. A practitioner-driven conference, QCon is designed for technical team leads, architects, engineering directors, and project managers who influence innovation in their teams.


Box: I'm here to talk about the story that ice tells about climate change. As it turns out, ice is a super sensitive indicator of climate change. There's plenty of other threads in the climate change story, but my past 25 years of working in Greenland allows me to convey this stark message about how quickly the ice responds to climate change. I'm going to follow that up with some solutions that fit within my repertoire or things that I'm doing in Greenland, ways that I think the tech can help with the climate crisis that we face. And so, let's get into it.

Greenland Climate Network

Greenland has this dome of ice on it. It has seven meters of sea level equivalent stored on land, and at the current level of a CO2 in the atmosphere, we've committed to losing most of this ice. So, there you already feel that at our current level of CO2 in the atmosphere, we have a big problem that I hope tech can help with.

My story began as a grad student privileged to install and maintain a network of automatic robotic climate stations on the inland ice of Greenland. That occupied me for 11 years and allowed me to tour the island almost every year. And, that qualified me for a job in Copenhagen, where I've lived now already six years, helping to maintain the complementary network of similar automated stations that are shown by the red circles. Those are concentrated down low where the surface melting is the strongest. And these stations, they're designed to capture the full surface energy budget. The amount of energy that drives melting. Also, the movements of mass snowfall and rainfall. They're also equipped with standard kind of meteorological measurements. So we get a lot of information from these with which we check climate models, we check weather forecasts, and we also do ground trothing for satellite measurements.

It's kind of like a ground truth system. And I think we're very fortunate that our society has invested in this 40-station network so we have authoritative numbers on Greenland's surface climate. Let me take you to the largest outlet where the ice is flowing under gravity into the sea. This largest outlet of the Greenland ice sheet produces 50 billion tons of ice per year. That would supply Los Angeles with its water for 100 years, every year that this glacier is driving ice out into the sea. As you can see from these lines, an area twice that of Manhattan Island has broken off just in the last 20 years. And that has actually reduced flow resistance. So now this world's fastest glacier has doubled in speed.

In the time that we were flying over this to one of our camps, I couldn't really appreciate it until I saw it with my own eyes, the scale of change. And the ice can break away basically instantaneously. The time it takes for a fracture to occur, it happens at the speed of sound. And that's instantaneous. The time it takes for the ice to build up is much longer. So one message is, the ice goes much faster than it builds up.

Follow the Water

I want to tell you the story that ice informs us. By visiting Greenland, we learned that, if you follow the water from its starting point at the surface, surface melting has increased on Greenland by a factor of two in the last 50 years, filling surface meltwater lakes like this one, amplifying sunlight absorption. That increasing meltwater supply is draining in to the inland ice of Greenland. There are unlimited options for the water to make it into the ice. And so, you end up with some really large flows of water that can actually hydraulically lift the ice while they're draining.

I want to talk, before we get to the bottom side, about what the effect is as the water drains through the ice, there's friction effects. Water, aerodynamic drag alongside the tunnels, it's draining through that, produces heat. That helps heat the ice internally. But more important is the fact that the water temperature is several degrees higher than ice temperature is internally. What's happening, that increased water supply is heating the ice internally and warmer ice is softer, it flows faster. And this is one of the processes I'll talk about that is leading to what we call the thermal collapse of the ice sheet.

As the water then drains to the bed, I mentioned that hydraulic effect because water is incompressible. As long as the water is supplied to the bed, it lubricates the flow and you get 100% speed ups during the summertime. And because the melt season has increased in time and in intensity, it's a geometric multiplying effect, it's flowing faster and longer. The water then reaches out into the sea in more than 200 outlets that are at least 1 kilometer wide. And in this photo, you can see, it's kind of a brown color, which is clear evidence that water was communicating with the bed of the inland ice.

And basically, all of that meltwater makes it out into the sea. You can see it pushes back the sea ice, here in the north. And that bay that you can see here is actually the results of a heat exchange that is forced by the water turbulently mixing with an Atlantic layer which is 4 to 5 Celsius at depth. You can see the effects of turbulence on the water surface like a Jacuzzi. That turbulent mixing is key for driving a heat exchange between the ocean and this water that is draining out at the bed.

In this illustration, the red arrows at the bottom symbolize the Atlantic layer, 4 to 5 degrees Celsius. The exchange of that heat being forced by the meltwater draining through and then entraining the warm water, and the freshwater being lighter than the salt water is forced to rise up this underwater ice cliff creating that enhanced convective heat exchange between a warming ocean and the ice where it is grounded. And, by eroding ice, undercutting it at the grounding line, that's the most efficient way really to reduce flow resistance and ensure that the ice moves faster into the sea.

Yet another process has to do with the fact that water is heavier than ice. It's more dense and existing fractures, if they can be flooded with meltwater, have the unlimited capacity to open up fractures as long as the depressions remain filled with water. And that process we call hydrofracture and it is responsible for some very large calvings of ice shelf. This 2 – 1 ½ Manhattan Island size area broke away in 2010 during a record warm summer. And, two years later, the next piece about a Manhattan-sized area breaks away. This is now no longer the largest ice shelf in the Arctic. That is now in northeast Greenland. This place is the northwest Greenland. This, again, shows how fast the ice can respond to melt, to heating.

Greenland Ice Sheet Mass Balance

I've published a reconstruction of the Greenland mass balance, the mass budget. This goes back to 1840. So you see the end of the Little Ice Age, on the left side, these positive values. This is when the Greenland ice sheet was growing a little bit. And then we have our warm period in the 1920s, an intervening cool period, which has actually a lot to do with a lot of coal sulfate aerosols being driven into the atmosphere shading the surface. That's been called Global dimming. That goes away with the Clean Air Act and the ice loss comes back with a vengeance. When you integrate that curve, you end up with this acceleration out of the Little Ice Age. It's not really surprising that it's accelerating, it's more of a concern - "Okay, well, how fast is it going?"

Why Is Climate Heating?

I want to just touch on why climate is heating. I think a lot of you know this, so very quickly. We have spiking greenhouse gas concentrations. This is 2,000 years of data. This is now confirmed, it's a faster increase in CO2 than during the Paleocene-Eocene thermal maximum. So in more than 50 million years, the CO2 is rising faster. What's largely driving this is fossil fuel combustion, coal, and then petroleum, natural gas, even bio fuels, our net increases of CO2 in the atmosphere. That's not all. Humans use fire to clear land, and then there's an increase in wildfire with climate change.

Cement production is about 10% of CO2 emissions. And I've learned that in the last 20 years, China has put out an equivalent amount of cement than North America used in the previous century. So, the huge construction in China is one of the effects that's driving up CO2.

When you add all the different human factors that influence climate, it's not just warming. The blue bars are the aerosol effects. Sulfate aerosol, as I mentioned, from coal combustion has a cooling effect. And various components including the sunlight, which has a very modest warming effect, climate change contrarian is often 0.2 solar output changes, but, in reality, it's more than nine-tenths of the warming is due to human activity. And climate change contrarians will say, "Oh, but the climate is full of cycles." That's true. What's happening now is humans are by far, the dominant factor in influencing our climate.

When you add up all these factors, it adds up to a net warming effect that's equivalent with little Christmas light bulb shining at 1.5 watts per square meter over every square meter of the planet, radiating downward 24/7. That is the net warming effect. If we look back 800,000 years, what I want to convey with this, you can see, is the spike in methane and carbon dioxide, which is far above any time in the last 800,000 years. And then the sea level effect, a sea level reconstruction from ocean sediments, basically this is all about observations, this is all real data here. What you have during the last 8 ice ages is a gradual buildup of ice on land, takes 80,000 years, and then a much faster loss of ice on land. And that reinforces what you saw earlier, the very quick breakaway of ice that's possible. So, we are perturbing the climate system in ways that the ice hasn't fully adjusted. And it will take a long time for that to occur but we're already observing the response of land ice, and I'll talk a little bit more about that.

This is another observation-based climate reconstruction in last 2,000 years, again, very interesting. The climate was cooling in the Arctic summer. This is Arctic summer temperature reconstruction based on wood and sediment archives. So, we were cooling, we were actually basically overdue for another ice age. And human activity, the Industrial Revolution produced that warming pattern you see on the right. And so this has been called the hockey stick because of that shape. By forestalling the next ice age, that's probably a good thing, but you can see we're rocketing off in into what's called a super-interglacial state. And what I'm going to show you in the next slide, brace yourselves as I tack on the next 100 years of temperature projections for 2 scenarios. One is Paris climate accord like, the other is business as usual.

The Next 100 Years of Temperature Projections

This is the envelope of what we can expect for the next 100, actually, 80 years. Here we go. Arctic summer temperatures are going to increase at least three more Celsius. And what this does for Greenland, unfortunately, is pushes the Greenland ice far beyond its threshold of viability. It then becomes just a matter of time until that ice is ablated.

The businesses usual scenario, at the top, is way off the charts. And so, our collective choice as a species is really where on that curve we are. The more we dial the curve down, the more time we save ourselves, we buy ourselves time by those investments, and arguably we will save lives by engaging in this, as a true crisis. The house is on fire, what do we do? We get out the buckets, we try to put out the fire. And so far, we're not doing a very good job because we're all kind of in a bit of a hallucination, and we're running off the cliff and we haven't yet noticed that we're in the air.

Here's Big Ben. We had a lot of sea level rise out of the last ice age. We have a stable sea level, amazingly stable sea level. This is part of the ingredients about how you can have a civilization developed where it has a stable, the sea levels just stays put. You can build your dock and you can build your port city in a place, you don't have to move your infrastructure around. And that was key in the development of civilization, as was a stable climate and predictable rains and so on. And a lot of that's about to go out the window. It's starting to go out the window, and we need to adjust to that. What we're looking at going forward is something like between a half of a Big Ben, or less, or more. Again, that's to some degree our choice. We have a limited time to dial this down.

Instrumental records show a sea level rise during the Medieval Warm Period, drop down into the Little Ice Age, and then we're going back up again. It's actually an abrupt increase in sea level rise. And it's not news. I think, to anyone who's paying attention that the era of sea level rise has begun. New York Times put that a couple of years ago.

How to Fix It?

Briefly, a couple of ways I propose that we get into solving this. These are really coming from my experience and as experts in your field, you can find your ways to engage this issue and help society adjust to this crisis. But the basic equation is we need to halt carbon emissions, the sooner the better. That's not enough because we have that problem of too much carbon already in the atmosphere, about 50% too much already in the atmosphere. So we need to remove an enormous quantity of carbon from the atmosphere. If we can do that, we can stabilize climate change. Very briefly, because this is an exciting, it seems like a kind of a win-win situation.

We can solve this 5x increase in data center electricity demand in the next 5 years that I've heard about. Mind-blowing? How are we going to satisfy data centers 5x increase in electricity demand? One way that I know, and I've spoken with officials in Greenland and this is as yet untapped, is Greenland hydropowered data centers. There's already data cables to West Greenland. They can run 6 terabits. They can be expanded - one goes to Iceland, one goes to North America.

Greenland existing meltwater that's in lakes, it can deliver 500-gigawatt hours 24/7 at a cost of about 4 cents a kilowatt hour. I understand that is less than the typical cost per electricity in most other places, and this is 100% clean energy. When you tap Danish energy, you don't know what mix is nuclear, what mix is biomass. This is 100% renewable, and it's more reliable than wind and sun because it depends on gravity, which is always there, and it's untapped.

I spoke with some Greenland officials, and I just want to flash up some of their material. They've published reports. They have a strategy for clean energy, they have policy in place promoting hydropower, and so that Greenland is kind of open for business. So I hope that you can share this message. Contact me, I can help you get in contact with Greenland's to put your data center Greenland and make it clean and green.

I got one more solution, then I'll wrap up. This one's smaller scale, but kind of dear to my heart because I've been busy planting trees in South Greenland. I was criticized for having a large carbon footprint. I thought that's legitimate, I need to plant some trees to offset my own carbon footprint, so in 2015, we planted 13,000 saplings. We've created a startup to do this and we're crowdfunding this now. Trees grow very happily in South Greenland. We've made, and we welcome you to join us to make it happen, offset your travel €40, San Francisco to London, but help us make it happen., thanks for listening to this plug. And, here's my contact info. And, I want to mention also that I'm here Wednesday morning, 6th floor, 7:15 breakfast, so we can continue the conversation there.

I think there's a bit of Q&A time and at this time, we'll transition to Paul Johnston. I was really setting up the problem, and you can see it's a big, big problem. And while he gets ready, I'll just say that I set up the problem, mentioned a couple of solutions, Paul's going to expand that much bigger. And we've had some really nice conversations about what is it that tech can do? What are the specifics? And he's been thinking a lot about it, he's been rewriting his presentation, so I think he has a really thoughtful and extended response to that challenge.

Climate Change Is Causing an Increase in Extreme Weather Events

Johnston: I think one of the things that we've all known is that climate change is having a huge effect on the weather. And this is Hurricane Maria, it hit Puerto Rico in September 2017. It was the second hurricane to hit the region within a few weeks. It was an unprecedented event. It hit the island, it had a devastating effect on the people. It destroyed homes, it destroyed livelihoods. It had a massive effect and cost an absolutely huge amount of money. It was the third largest bill in American history.

And climate change is just causing an increase in these extreme weather events. It isn't just the fact that the climate is warming. It isn't just the fact that the climate is having the effect on the ice and having the sea level changes, there are numerous other effects that are being caused by climate change. And this is the new normal. This isn't just something that's happening and we'll go back to being normal when we all fix this problem. This is the new normal. This is going to get worse. And this is what we've been being told by scientists for a very long time. And I'm not saying this to scare you. I'm just telling you now that this is actually what is the new normal.

There are companies out there, one called Mylan in Puerto Rico has actually realized this, and they've done something about it. Instead of going, "We're just going to stick our head in the sand and not do anything." They actually went to a company and said, "Right, what's our climate risk?" They went to this company, and they said, "Actually, your climate risk - you're in Puerto Rico, so therefore, what you need to do is you need to think about hurricanes." So they said, "Fine, we're going to reinforce our roof." So when the hurricane hit, all the buildings around them were destroyed and they're building stayed up. They looked at their business risk, they thought about it, and that would have been $150 million bill but the reinforcing of the roof was much less than that. And that was a huge saving for the insurer, and it was a massive climate risk avoided.

Climate Change Is a Business Risk

Climate change is a business risk, and it's obvious that it's a physical risk. There's obvious risks there that you can think of. Physical risks are the easy ones to think about, but there are other risks as well. And even newspapers like "The Economist" have started to talk about this, and they're talking about this in other ways - so, reputational risk as well. We'll get on to that, but we're at a tech conference, and we don't talk about these things because they're not really relevant to us, are they? But actually, they are. And I think we'll talk about those in the next few minutes. So, why am I talking about this? Well, I used to work at AWS. I'm a CTO, I talked about servers a lot, so ask me about the buzzwords later but I'm also an environmentalist. I've been involved with climate change projects for a number of years and this is something that's really important to me. I've done the research, and I've done the looking into it, and I've been involved in some projects, and I wanted to share this kind of stuff with you.

Personal Choices

And one of the things that we always talk about, and this is personal choices, “What are the personal choices that I can make to have my impact?” And there was some research done a couple of years ago. Some of you may have seen this that actually, there are some things that are really high impact and the top one is not having children. Well, I'd just like to point out I have five, so, basically, I'm a bit stuffed on this front. So I don't think it's particularly helpful actually, and I don't agree that you shouldn't have children. I think that we should be basically raising environmentalists and fixing the problem, not telling people they shouldn't be having children.

But the actual high-impact things that people talk about are actually quite important. But the thing that's really interesting on this is the low-impact stuff. Things like recycling, things like upgrading light bulbs, they're the things that most people think about when you talk about how to have an impact on climate change. Actually, the things that have the most impact are things like buying green energy, going vegetarian, taking fewer flights, that actually really simple things to do. And actually, as a tech industry, taking fewer flights, I think that's something we could probably all get behind. But we probably won't because we actually quite enjoy the jolly when we go out to the conference, don't we?

Community Energy

Another thing that I was involved with, as I realized that the personal effect of doing those things was a much lower than actually getting involved in larger things. So I actually got involved with the community energy organization, which if you're not sure what that is, it's just a small organization where a few of you get together and do something bigger. So, we, in my hometown, bought a large array of solar panels, it's about the same number as 100 homes, and we put together what's called a community benefit society. And we essentially became a small energy company. I learned a huge amount about how incredibly difficult it is to run a small energy company.

Basically, if you're not a big energy company, nobody wants to talk to you. But we're still going, and we've actually had a bit of an impact - something like 90 million cups of tea, I think was the best way of explaining it. But we have made some changes, and we have had an impact. And, I actually look at this as my way of offsetting my carbon impact.

Tech, Electricity and Emissions

Let's go back to tech. Tech essentially runs on electricity. Electricity creates emissions, very straightforward and very simple. Actually, most people in tech are essentially unaware and very disempowered when we start to talk about climate change. Most people don't even really think about it. And there have been reports over the years that actually we're in a real problem here. The internet of things says a whole load of things about how it's going to change home, so they're going to be better. We're going to have smart homes, and it's going to be the savior of everything. But actually, there are reports that say that by 2040, IoT and internet-connected devices could be 14% of global emissions. So let's think about that. We're actually going to save it by creating more? That doesn't make any sense.

And there are more of these kinds of projects and ideas that are producing statistics that don't seem to add up. The ICT industry could be 20% of global emissions. Data centers, which are the thing Ann Currie and I have been talking about for about a year now. They're roughly 2% of all carbon emissions, which is roughly the same as aviation. It's not a small thing we're talking about.

Whenever we have this conversation, someone comes up to me and says, "We're making our computers more efficiently. We're making data centers more efficient." The thing is, efficiency will not save us. Efficiency is not what matters. If you make something more efficient, people want to use it more, it just grows, and data centers are growing at a massive rate. You look at AWS as a proxy for this. That's AWS's growth curve up until a few months back. AWS is growing massively, the industry is growing massively.

And Jevons Paradox, if you have a look at that, "Increased efficiency leads to increase demand," and the data center industry and the cloud industry is basically showing us that that is true. Efficiency will not save us. We just have to use less or change the way we use it. And Bitcoin, something every time I do this talk this comes up. Basically, Bitcoin is using an enormous amount of energy. Have a look at the Digiconomist website. Look up Bitcoin energy consumption index. Bitcoin mining roughly uses the same amount of energy every day as Singapore in terawatt hours. Just to give you a bit of context, that means that the estimated carbon footprint per Bitcoin transaction is that, which is 192 kilograms CO2 equivalent, which is about the same as a VW Golf driving to Edinburgh and back from here. One transaction. That should horrify you.

Data centers are 24/7. They use electricity all of the time, and they're a growth market. Electricity providers recognize that they are an area that they can go and say, "Actually we want you to use our electricity. We'll give you a good deal." They're a growth market, they're a good area. And when you're talking about data and machine learning and AI, they know that you guys out there are people that they want to talk to.

Cloud and data centers do a brilliant job of abstracting something. They abstract your use of electricity and they abstract that electricity's creation of carbon emissions. You don't think about it. You don't do something like launch an instance in the cloud and then get a bit of information that tells you the grams of C02 that it's releasing every hour, do you? It doesn't happen, does it? Does anyone actually know how much carbon their instances release as they're being used? I bet you don't, unless you're using a sustainable region.

The Business Risk

We can talk a little bit about that, and that's my frustration. But we can now talk a little bit about the business risks. This is not just about getting frustrated because there are real business risks to climate change. And Forrester has done some research into this and it's worth having a look at their report. Just google "Business risks of climate change", this is what comes up and it's actually pretty good and worth the read.

But one of the things that's worth looking at is whether your cloud is climate neutral or carbon neutral. Do you know whether it's climate or carbon neutral? Have you actually even bothered to look? Because moving workloads to something that's carbon neutral may take years, and it's actually really difficult. And it may become more important reputationally over time. You could just imagine if we all of a sudden have to move things or if someone says, "If it's not, then we're going to charge you tax." Better to start thinking about that now.

What about financial risks? Again, if the world starts to decarbonize at some point and starts to say, "Actually, we need to move everything over", we're going to have to start paying an awful lot to make renewable energy or to create more nuclear power. So, do you really think that the cloud providers and the data centers aren't going to pass that cost on? They're not going to swallow it up. They're going to pass the cost on, they're going to make you pay. If you're already with someone who's using renewable source of electricity, then you might not have to pay as much.

What about operational risks? Do you actually know where your data centers are? It's quite fun because I used to work for AWS, and they don't tell you. But actually, do you know where they are? Do you know if they're climate-change proof? Is it a question you should be asking? Also, what about your suppliers? If you're in a supply chain and you're just-in-time supply is reliant on a specific supplier and they are not climate-change proof in some way, shape or form, that's quite important. What about their technology? What about their data centers? And there are regulatory things as well. What about financial things? All of those elements matter.

Technology Matters

Technology's underpinning all of this. Technology is the core of how we do business nowadays, and it's only going to become more important. We, in this room, are the kind of people who are going to be leading the charge in the climate change revolution. And we have to be doing this kind of thing. We have to be asking these questions, and we have to be thinking about it, we cannot ignore it.

Cloud Providers and Renewable Energy

If you ever think about the major cloud providers, there are six major cloud providers. And I'm talking about cloud because there are some easy wins that we have here, so I'm just going to quickly go over. The major cloud providers are AWS, Azure, Google, Alibaba, Oracle and IBM.

AWS is the largest cloud provider out there. They actually have four public regions, which are, in their words, sustainable. And they are Ireland, Frankfurt, Montreal, Oregon, and they have a fifth one which is the Govcloud in U.S. which are 100% sustainable, which is interesting. They are planning to go 100% sustainable at some point, but they have absolutely no timescale on that. There is a bit of a problem with AWS, which is the fact that actually, they do a very good job of doing deals with their energy suppliers, which means that local people of those energy suppliers have their prices go up sometimes. It's been an interesting one, that one. But they're doing pretty good. So if you have a workload on AWS, consider putting it into the sustainable region simply to get your renewable credentials, because I think it's really worth doing. I've had a number of conversations on Twitter with people going, "Where should I put my data?" And I've gone, "Stick it into one of the sustainable regions." They've gone, "Actually, that's a really good idea." It's not a difficult thing to do.

Azure is pretty good. Microsoft offsets their carbon with carbon certificates. They do really well here, and also their employees offset their flights as well. They are a pretty good organization for this, and so they're carbon neutral by default 100%. This is really good. If you're in Azure, well done, you're 100% already. This is really good. However, it's not perfect. Offsetting is fine, carbon credits are fine, but they don't necessarily offset very well, so, it's an imperfect way.

You'll see why in a minute because we're talking about Google, which is essentially currently the gold standard because Google are currently the biggest corporate purchaser of renewable energy in the world. And we love Google for this, they are brilliant. Basically, what Google do is they check on roughly an hourly basis for their data centers, and they try to offset by buying local energy in the locality and renewable power in the locality for that hour. So they try to offset as much as possible in their area. And it's really quite incredible what they're attempting to do, and they're really pushing to try and make energy more efficient and more available for other providers as well. So they're really working in this area to do some amazing things.

Alibaba, they are Chinese and they are very big in Asia as well. We found it very difficult to find any information about what they're doing in terms of sustainability, and it's very difficult. Because they're based in China, which has a very dirty mix in their electricity grid, so they have a lot of coal-based power stations. That means that their electricity that they're using is very emissions-heavy, which isn't very good. And that's very hard, because essentially on this basis, that's not very nice, not very good for the environment. So, please try and keep away from Alibaba from an emissions point of view.

Oracle, proudly, very green. They are in this and they claim a high energy efficiency which is good but as you said, efficiency doesn't save us here. And some regions like the UK are actually 100% renewable. So if you want to use Oracle, then put it put it into some of the regions that say they're 100% renewable. And their goal is 33% by 2020, which is admirable but still a little low, they could do better. But they do also provide decent reporting. So, well done, Oracle.

IBM have actually a very long-standing commitment to trying to be environmentally friendly, so we like that. And they procure 50% of their energy from renewable sources, which we like. However, still not that high and you can't put a workload into 100% renewable data center. So, again, something that could be improved.

We've got some data on some of these things. And overall, almost all have some way of doing it apart from Alibaba, have some way of using either renewables or some way of putting a workload into a sustainable way. And our estimation is somewhere around 25% to 40% of all clouds is powered by renewable power.

But we actually don't have an independent and auditable data. So you guys, I wouldn't be surprised if you didn't know or wouldn't know where to even find this information, because it's actually really difficult to find. It's really impossible to know where to look. And what is renewable is different between different providers.


We have some real problems as well because there's data that shows that we're going to have roughly five times growth in the next seven years in terms of data and data centers. We're going to have a huge amount of capacity problems, and we're going to need a huge amount of increase in energy use and consumption. Where's that electricity going to come from? It's not going to all come from efficiency gains, it can't. How are we going to deal with that problem as tech? And actually, now that you know, how are you going to deal with it from a carbon point of view? Because offsetting, the idea of offsetting is you use electricity somewhere, and then you buy renewables somewhere else, and then you offset the amount you use. It's imperfect because you still release greenhouse gases in some way. And it doesn't make it better. You still have a problem, it just kicks the can down the road. You're not really dealing with the direct issue. You need to fix the means of production in the end somewhere. Someone somewhere has to do it. Possibly, customer pressure's going to do that and make them actually do something about it.

It’s About Us

And actually, we in the tech industry, we want to be saying things like, "We want 100% renewables on our instances." And not only that, those big providers, these companies are not small companies, they are huge. Some of them are like the biggest companies in the world.

Google are already doing some amazing things in this space. DeepMind, for example, are improving the value of wind energy by 20%. This was last week, I think it was. They use machine learning to make their data centers energy usage better. So there are some amazing things going on out there. We can encourage this. We should be encouraging this, we should be talking about this stuff.

What Should You Do?

So in the end, what should you be doing? You should be assessing your risk. You should be looking at this. This is the easiest risk to look at. Where are your data centers? Where are your cloud instances? Which cloud are you using and why? You should be talking to your suppliers and your customers too. You should be encouraging your cloud to go fully sustainable as quickly as possible.

Anne [Currie] and I have written a white paper which is there, which is Please, go and read it. The Ethics White Paper here is the table that is the most useful one which tells you which of the providers are good and bad in our opinion, and bear in mind, this is an opinion. Google and Azure are at the top. AWS gets a C, and everyone else comes a little bit lower down.

And we also have a pledge, which we would like you all to sign if you would like to get involved in this. Google sustainable service by 2024, it's the only thing that comes up because there's nothing like that anywhere else. And that's what it looks like. Please do sign up.

A few things to leave you with. Tell your provider you want them to use renewable energy so you can have sustainable service by 2024. And, please offset your travel to this conference with Greenland Trees. Again, €40 will offset one San Francisco to London flight, that's €40 to get rid of that carbon. That's a really good thing.

And just to leave you with a couple of things, climate change is the biggest issue we face as humanity. Extreme weather is the new normal, and it's a current and future business risk. Something my friend told me, we are the first generation to see the effects of climate change firsthand. And probably the last generation who can make a difference. And climate change is our problem as a tech community. You are no longer unaware or disempowered. And we can make a difference. So let's change the world. Thank you very much.

Questions and Answers

Roxanne Beverstein: I'm Roxanne, I'm with QCon, a co-founder of the company. And we are committing a $5,000 match to anyone who contributes to the And we already have over $5,000 in that crowdfunded project. Jason wants to get $40,000 to plant, I think he said, 10,000 trees. And he said he could make a dent in climate change if we did that.

Johnston: Let's do it.

Participant 1: Thank you very much for that. That was interesting and suitably scary and refreshing to hear in this context. A lot of what you said had to do with mitigating our carbon footprint. But in what ways can we, as software engineers, directly use our tech skills in order to tackle climate change? And I guess you touched upon it with the DeepMind, making things more efficient. But are there any other applications that we can use?

Johnston: I will sort of give you something that Jason was telling me earlier, which was that, there isn't a huge amount of money in doing the research. So there is data that comes out of all of that information in the Greenland ice sheets, and he doesn't have all the money to pay the clever data scientists to do all of the work. So I know that he's going to say, if you donate a bit of time to look at some of the data for him, and to help him with that, then he would really appreciate that. I don't know how you organize that, some work that one out. But I know that you have a huge amount of data that you can't, you don't know how to code yourself in totally.

Box: Yes. I think the way it could work is to build a few processors. So, it's an algorithm that does something. And the data scientist could produce such a processor and then we could put it into work. The idea is to make it robust enough that you could walk away. And then that processor will be processing European Space Agency image data [inaudible]. So we have some pretty specific areas that we need some help from data scientists.

Johnston: That kind of thing, I think is we should be looking at using technical people who can do that kind of thing for.

Participant 2: My question actually is directly to what you just said. Is there a way you could set up a website where I can simply send people to? They can actually help but make it easy for them? Like a node that's able to do the Bitcoin transaction, as a person, as a developer, I could just go on that website, and say, "I have a free Sunday," so I'll go and do something for you so that make it super easy for developers to jump onto the train.

Johnston: Okay. Who would be interested in something like that? Please raise your hands.

Box: Sounds like fun, right?

Johnston: Okay, so that's something brilliant. So that's an action that we need to take away and do something with. That's fantastic. We will have a conversation and figure that one out and feedback somehow. We will do something about that.

Participant 3: Developers can get done more things when there is a website we can go to.

Johnston: We need a place to go.

Box: Yes. We need to make a crowdsourcing data scientists for the environment or for the climate kind of site, where we put on a bite-sized problem to solve that you guys could do for fun.

Participant 4: Where's the discussion of nuclear energy in here? For like data center has huge baseline load, nuclear reactors [inaudible 00:49:59] huge baseline load or baseline generation. Why is this not featuring in the discussion of renewable energy or at least the neutral energy?

Johnston: Some of the clouds use nuclear as a sustainable source of energy. And it is a debate as to whether or not nuclear should be or should not be used as part of that. The problem is, and this is a whole other conversation, the problem is the use of nuclear is actually quite difficult in terms of creation of new nuclear energy, and it's mainly political. And certainly, in the UK, if you talk to people within the UK, they'll tell you that four nuclear power stations have just been multipled, essentially, in terms of creation of new nuclear power station. So there's a whole big problem there as well. But yes, it's a separate debate, but it's one that I can have with you afterwards.


See more presentations with transcripts


Recorded at:

Apr 08, 2019