Caitriona Devery talks to Conor Sweeney about climate change, renewable energy and the science of meteorology.
UCD School of Maths & Statistics
Q: My first question is rather simple, but I think sometimes there can be misunderstanding. What’s the difference between weather and climate?
A: So the weather is exactly what’s happening at any particular place at any particular time. If you want to have a barbecue tomorrow evening, you want your forecast to tell you if it’s going to be raining. That’s very precise, in terms of the location and the time. So weather forecasts start off by looking at the world and how the weather is right now and use the equations of the atmosphere, the physics, and put all that into a computer model and they try and figure out how all the different parts of the weather are going to change, how evaporation and precipitation and all these things are going to change over the next few days.
Climate is very different. Climate is like what are all the different things the weather could possibly do. In Ireland we’re very lucky, because the climate is very mild, so it’s never really hot or really cold, and we don’t have hurricanes or typhoons or any of these extreme winds. For our climate we’ve got a range of temperatures that would be normal. To gauge the climate of an area we normally look at around thirty years.
The interesting thing that started since nominally the sixties but more recently since the nineties is to look at how the climate itself is changing. We use climate models to study this. If temperature was increasing, then your climate model still couldn’t tell you what it’s going to be like on a particular day in ten years time, but it could tell you that all the ranges of temperature that we know now will all get a little bit warmer in ten years time, or the range of temperatures could increase. A climate model tries to predict the range of possible things that could happen.
Q: How has meteorological science changed over the last few decades?
Weather forecasting and climate modelling have both improved and it’s to do with both technology and understanding. The technology one is simple, that’s just computers. We have computers that would have been beyond our comprehension, the power they have now would have been unthinkable decades ago. I started coding when I was around ten or eleven, on a computer called a Commodore 64.. We’ve moved to a massively more powerful stage in computing, where we have computers that have thousands and thousands of individual computer units in them that can all work together with lots of memory. Climate modelling is perfect for that because our models can be split. You can split the whole world into lots of little cubes, and each computing unit can calculate what’s happening in its cube and then swap with its neighbours. We can run codes on computers with thousands and thousands of nodes. Super computing has impacted both weather and climate modelling dramatically.
The other part of it is science and research. Our understanding of what exactly happens in the atmosphere has increased in leaps and bounds at the same time as computing. It’s really hand in hand. These are small things, like… what’s a cloud? Clouds are intriguing and until recently and even to an extent now they would have been a source of a lot of the uncertainty as to what’s going to happen with the weather and with climate. They’re composed of tiny water droplets, but these are microscopic scales, and our models can only estimate what happens maybe every one kilometre or even every ten kilometres, depending on how high the resolution is. But of course clouds are very important. If you have clouds, they reflect the sunlight back out again, so below the cloud won’t be as warm. Our understanding of clouds and all the different types of clouds has increased a lot over the past decades.
Q: What kind of technology do you use for this?
A: In meteorology we use pyranometers, they record the wavelength of radiation. The sun emits light at shorter wavelengths, like visible light, which we can see, and ultraviolet light, which is bad for us. The earth itself, which is cooler than the sun, is emitting longwave radiation, and that balance of incoming shortwave radiation from the sun and outgoing longwave radiation from the earth, from its temperature, that’s the core question of climate modelling. If we have these greenhouse gases that stop the longwave radiation going out, then of course we’re heating up the whole system.
Satellites allow us to see what’s happening all around the earth. Until the most recent generation of satellites, mostly they worked by recording different frequencies of radiation, different bands of shortwave and longwave. Before satellites, we had much less data about what was happening in the atmosphere. You had land stations and ships and aircraft. Then satellites came out, the earliest meteorological satellites would have been in the mid-seventies. We got the data from satellites before we knew what to do with them or how to put them into computers. It wasn’t until around 1980 that we could use all of the satellite observations in weather forecasting, so it’s about forty years of this era of satellite forecasting.
Q: How would you compare the pre-satellite era to the current one?
A: That’s a question that’s interesting to quite a lot of people at the Earth Institute, people who want data about the earth, people who are not meteorologists. People who have questions about land use or sea surface temperature or snow coverage. For a lot of them satellite data is really important. Suddenly you know the temperature of the skin of the earth everywhere. Before that you only knew it where you had a thermometer. If you take something like rainfall in Africa it’s hard to get at that historically because they don’t have many weather stations. We don’t have historical data. Satellites came along and you can guess from type and thickness clouds about precipitation. There’s a lot of uncertainty but it’s better than nothing. Land use, urban climatology, growth of cities, all of these things have benefitted from satellites. Before this you physically needed somebody there recording data.
The most recent satellites have active sensors on them, so as well as looking at what is emitted from the earth, they can also bounce signals off the earth, like radar, to investigate the atmosphere more closely. They can look inside storms. They can send a signal that goes right through a hurricane, it goes right through the storm. They get slices of data all the way up. Before that, they had to fly aircraft into storms, that was the only way they could get the data.
Q: Post-satellites, what happened to all the data collection points and stations?
A: That’s very important. Satellites are great for data but they are not what we call ground truth. We need to have observations on the ground to ensure that the satellite data is correct. A lot happens between here and space. There are uncertainties. If you have a thermometer and it’s there, the uncertainty is tiny. All meteorological stations have at least two of every instrument. If they’re both the same, you know that’s the temperature. It’s really important to keep those up. You can get by with less, but it’s really stupid because you’ve more uncertainty. The government have realised that now. Happily they are putting out more ground stations, but it’s cheaper now because they are all automatic.
Q: You worked as a researcher for Met Éireann, what was that like?
A: After I finished my PhD, in computational fluid dynamics, my first job as a postdoc was in Met Eireann when they set up their climate modelling team. That was great and I learned lots about the atmosphere. Met Eireann is in Glasnevin. That’s where all of their researchers are based. It was a great place to work. Everyone had physics or maths or theoretical physics backgrounds. I was surrounded by people who were very smart and very interested and wanted to do the best they could, wanted to figure out what was wrong and how to make it better. We were the research group, there was also the operational forecasters that were on radio and tv, there were also the people that look after all the instruments. It’s quite a small group of people, maybe 50, and we’d all meet for tea. The big thing I learned from Met Eireann was the sanctity of having tea at 11 o’clock. Everyone just got together, very quickly you get to know everyone.
Q: What kind of applications are there for your meteorological work?
A: I’m working in two areas. Most of it is in forecasting, for wind energy and more recently forecasting for solar, for electricity, solar panels. Wind energy has really taken off in Ireland. If you want to build a wind farm you want to know where it’s going to be windy, how windy it’s going to be, how much that wind is going to change. And if you are trying to get that electricity to everyone’s houses, like EirGrid, and you’re responsible to balance electricity you need to know when this electricity is going to come on the grid so you can tell some other producers to stop producing electricity.
I work on the prediction side. Generally speaking the time unit is every ten minutes, that’s the trading blocks, you want to know every ten minutes how much is on the grid. They want to know 48 hours ahead, what’s going to happen so they can plan for tomorrow’s electricity. More recently I’ve been working on prediction of solar energy. There are lots of solar panels in the UK and lots in Northern Ireland so within the grid of Ireland there is interest in predicting electricity from solar as well as wind.
'Wind energy has really taken off in Ireland. If you want to build a wind farm you want to know where it’s going to be windy, how windy it’s going to be, how much that wind is going to change'
Q: How is this electricity managed?
A: The fascinating thing about electricity is the electricity that powers our computers right now is being generated right now. It’s not being stored. All this energy flows all around the country right now, in real time. If anything goes down, they need something else to kick in immediately. Managing the grid is really quite impressive. Historically we had control, because we had coal and gas and a certain amount of hydroelectricity. We could always turn those on or off. But now, EirGrid has to take all the energy whenever the wind farms are generating electricity, and than makes managing the grid more difficult.
EirGrid are amongst the leaders in the world in integrating wind energy, so this non-schedulable energy, that you can’t turn on or off. Other countries, like for instance Demark have more of it, but they can just send their energy across their borders if they have too much. They also have a lot of big communal buildings that share heating so they can shove electricity into them if they need to. In Ireland we don’t have widespread communal heating, we don’t have neighbours that we can send electricity to. We are linked to the UK but that’s smaller capacity. We are like an isolated grid. That’s make us very interesting to people studying electricity all around the world.
Q: Can it be stored?
A: You can but it’s extremely expensive. The one we have that we use all the time is Turlough Hill. That’s very efficient, but it was built on a large grant and it doesn’t make any financial sense to build another one. If you’ve got too much electricity, you send it to Turlough Hill, they have these massive turbines and they pump all this water from a lake down here to a lake up the mountain. If they’ve got too little electricity they do it the other way around. It’s not a battery, you’re just using potential energy to store. It works really well, but it’s really expensive to build. Batteries in general are far too expensive to build. You can just turn off the wind farm, it’s called curtailment, but nobody wants to do that because it’s a waste of energy. If you do that you still need to pay the wind farm.
So the question is, what can we do with this energy? Lots of people within the Earth Institute and the Energy Institute are looking at how we can handle this. You can shove electricity into immersion heaters that have electric elements in them, into electric vehicles, if you’ve storage heaters, which may now become more popular, that’s another option. All of these things are ways of using renewable energy as soon as it happens, without wasting it. To change our energy systems will take years and billions of euros investment to implement. All of Europe is trying to do that.
Q: Is it frustrating when working on forecasting to have to reduce hugely complex probability statistics to very simple warning systems, like for instance the traffic light system used for the weather?
A: It’s hilarious because there’s so much numbers and maths and physics behind it, and then you just get green, orange and red. The colours are based on impact, not on actual weather. You may have a really warm temperature event but if that happens in the winter it’s not going to be a huge impact. But if it’s in the summer, that could have quite an impact. What I find interesting is that some people who are really good with numbers, don’t want the numbers. We can give them lots of stats, probability, uncertainty, but they just want a traffic light. Because they’re too busy. They’ve already got more information than they can look at.
I don’t find this frustrating, although it did surprise me when I found out. I think that it's interesting, it keeps me in touch with the real world. I could get lost down a warren of the best way to quantify this 99.9999% thing and could spend a year trying to do it and someone could come and say, we don’t care, just give me green, orange and red. It’s good to be connected. I am based in the School of Maths, and I think pure research is really important. To focus only on applied research doesn't make sense to me. You need to also fund blue sky research, so you are prepared for whatever the future holds. That’s not to say science shouldn’t be applied, we need to think in those ways to frame our research. But I don’t think the funding should require it.
'I think that it's interesting, it keeps me in touch with the real world. I could get lost down a warren of the best way to quantify this 99.9999% thing and could spend a year trying to do it and someone could come and say, we don’t care, just give me green, orange and red'
Q: What kind of implications has COVID-19 had on the atmosphere, has it been significant?
A: Absolutely. There’s a great NASA website called the Earth Observatory and they post a picture of the earth every day, there was one last week that was a satellite tracking pollution – these aerosols over India, and it showed, it was really striking, the reduction in pollution. There are lots of other things, there are less greenhouse gases because of fossil fuel emissions, so there will be impact, there’s less transport, less pollution on transport lanes. This is a health disaster, but that doesn’t mean it’s not an opportunity for science. Now we have a complete shut down and we can see a lot of different impact of this. All our data is still being recording. I can only assume it’s the same for everyone, for economists, for technologists. It’s tragic but within that there’s an opportunity for research.