 My name is Dr. Barry Colfer and I am the Director of Research here at the IIEA. I am delighted to welcome you to today's events, our first of 2023, and the third presentation of our Environmental Resilience Lecture Series, which the IIEA is delighted to organize in conjunction with the Environmental Protection Agency, EPA. I'd like to take the opportunity, as ever, to thank the EPA for their sponsorship of this series and for their ongoing support of the Institute. Today we are joined by Professor Emily Schuchberg, Director of Cambridge Zero and Professor of Environmental Data Science at the University of Cambridge and Fellow of Darwin College. I'd like to thank Professor Schuchberg for being able to take the time to speak with us today. By way of introduction, Professor Schuchberg is a climate scientist and mathematician, a Fellow of the Cambridge Institute for Sustainability Leadership, an Associate Fellow of the Centre for Science and Policy, and a Fellow of the British Antarctic Survey. Professor Schuchberg is also Professor of Environmental Data Science at the Department of Computer Science and Technology at the University of Cambridge. Professor Schuchberg leads the UKRI Oakree Centre for Doctoral Training on the application of AI to the study of environmental risks. A polar expert, Professor Schuchberg previously led a UK National Research Program on the Southern Ocean and its role in climate sounds super cool. In 2016, Professor Schuchberg was awarded an OBE for services to science and the public communication of science. Professor Schuchberg is co-author with H.R.H. Prince of Wales and Tony Juniperp of the Ladyberg book on climate change. The title of Professor Schuchberg's presentation today is Artificial Intelligence in Climate Change Research and the work of Cambridge Zero. Professor Schuchberg will speak for approximately 20 minutes and she does have slides. After the presentation, we'll move as ever to the Q&A session with you, our audience, and thank you for being with us. You'll be able to join the discussion using the Q&A function on Zoom, which you should see on your screens, and please feel free to send your questions in throughout the session as they occur to you and we'll get to as many of them as we can. We encourage you to identify yourself and any affiliation you may have when asking a question. Finally, from me, a reminder that today's presentation and Q&A are both on the record and please feel free to join to join the discussion on Twitter using the handle at IIEA. First, however, please let me hand over to Mary Gurry, Water Program Manager at EPA. We'll offer some opening remarks before Professor Schuchberg's intervention. Thank you, Professor, and over to you, Mary. Thank you, Barry, and also on behalf of the EPA, I'd like to extend a very warm welcome to Professor Schuchberg and thank her for her time today. I'm sure it's going to be a fascinating talk. I'm certainly looking forward to it. Data, of course, is at the heart of what we do in the EPA, whether for our work on climate, water quality, air quality, or waste. And it's an ongoing challenge for us and indeed all scientific organizations to get the best out of our data, you know, in terms of understanding and communicating the problems and finding solutions. And I'm particularly interested in hearing about the role artificial intelligence can play in meeting that challenge. In this keynote address, Professor Schuchberg will give an insight into how the latest in artificial intelligence and computer science can assess and monitor climate risks and contribute to sustainable solutions in adaptation, mitigation, and policy. Professor Schuchberg will offer examples of the water-related challenges in our recent work in the fleet land surrounding Cambridge. And water and climate change are, of course, inextricably linked, so she will be discussing interconnected issues such as water management, risks to water systems, and examples of using artificial intelligence and understanding problems with water systems. Finally, Professor Schuchberg will give an overview of Cambridge Zero's facilitation of decarbonisation, research, education, and great engagement across sectors. So with that, I'll now hand over to Professor Schuchberg to deliver her address. Thank you. Well, thank you very much. And it's a pleasure to speak to you all today. So if I just share my presentation. So what I want to do today is just give you a little bit of insight into some of the work that we're doing in Cambridge, but probably more importantly, how we're going about doing that work in Cambridge. And if there's one overarching theme, I think, of what we've been trying to implement over recent years, it's radical collaboration. And what I mean by that is radical collaboration collaborations across interdisciplinary research areas within the university in Cambridge, creating collaborations that previously simply wouldn't have been thought of between different disciplines, but equally radical collaboration beyond the university. So in particular, so that we are connecting together in a really effective way the research cycle with the policy cycle, so that there's continual dialogue between the two, but that we're equally integrating into those two cycles, the needs and the ideas from all the different stakeholders that are relevant to different environmental problems. So I'm going to give you some examples where we've been trying to implement that viewpoint, both in terms of some of the key problems that we're looking at addressing, but also very specifically on that interface between how we start to use the new tools and technologies that are available, particularly through artificial intelligence and associated techniques, and apply those into the environmental science domain. So let me just dive in the three particular examples that I'm going to just give you a little bit of flavor onto what I've just described. The first is I'm going to, as was said in the introduction, give you an example of some of the work that we're undertaking, looking in the first instance at the landscape coincidentally around Cambridge, which is a very heavily peatland landscape and has significant environmental as well as actually social challenges associated with it, and how we're adopting a whole system approach to addressing that challenge, and I'll explain what I mean by that. I'll then give you a couple of examples of the work that we're doing, applying AI to specific environmental problems, and the two problems that I'm going to describe are associated with flooding. And then finally, I'll very briefly explain what we're doing in terms of providing integrated solutions to environmental challenges. And if I have an opportunity, if I have time at the end, I'll tell you a little bit more broadly of what we're doing with Cambridge Zero across the university. So let me dive in, first of all, with the whole system approach. We have recently in the last, just over, but this time last year, in fact, received significant funding from UK research councils to set up a new center that's focused on looking at how we can develop optimal solutions for landscape regeneration. And critical to this is looking in a very holistic way at that challenge. So looking at potential alternative land management practices from the perspective of both climate change related to challenges, biodiversity related challenges, but then critically also those social challenges, recognizing that solutions need to work for the people who live and work in those landscapes and for the people who those landscapes end up serving. And so there's a set of, if you like, root challenges associated with that that we're looking at in terms of understanding how outcomes vary with alternative land management practices, what optimal solutions might be and optimal will be very much dependent on who the different stakeholders might be and then how those solutions can be in some centralized and supported both from a policy framework perspective, but also more broadly from financial instruments, for example, and how we can work with businesses to help support these solutions. The particular projects that we're working on, we're working in the University in Cambridge in partnership with other organizations and some of those are rather research organizations, but also critically some of the NGO organizations. And the particular project that I'm going to describe to you, the first landscape that we're considering is the Finland around Cambridge, but then there are other landscapes in the UK and eventually internationally that we're also going to be looking at with this approach. So this is a view of the landscape, the Finland landscape around Cambridge. It's heavily peatland landscape. Most of it was reclaimed land from the drainage of the Fens in the 17th century. You can see here the drainage ditches, the waterways that crisscross the landscape. And you can also see the rich peatland heavily agriculturally formed landscape. And the challenges are really from an environmental perspective are two-fold. First of all, there's really significant greenhouse gas emissions from the peatlands and the peatland itself is wasting away at a very significant rate, which causes its own challenges. But then additionally, there are significant water challenges in the region and water challenges from two different perspectives. On the one hand too much water because this is a region that's largely below sea level. And so there are really significant flooding risks associated with the region, especially in the context of future climate change. But additionally, it's a region where there's actually significant water shortages. It's one of the driest parts of the UK. And there's significant development, housing development in particular occurring in the region, which is putting significant strain on the water resources. And the challenges, the emissions from the peatlands and the challenges of water in the region are connected together, because one of the ways in which you might look to limit the amount of greenhouse gas emissions from the peatlands is to reflug to rewet those peatlands. So there's a connection. These are not independent challenges, although historically they've often been perceived to be independent challenges. So what we're looking at is whether or not there are innovative solutions that we could look to to the future of this Finland landscape in a way that is beneficial to climate, beneficial to nature and to the biodiversity. You can see this isn't a particularly rich, biodiverse landscape at the moment, the way that it's managed, but crucially also that works for the people who live and work in this landscape. And the Finland's are one of the most socially deprived parts of the UK. So that latter aspect is critically important. It's also critically important that this is a hugely important agricultural region and a very large percentage of many of the UK salar drops are grown in this region. So recognising the economic importance is also critically important. So we brought together this collaboration, which really is an incredibly diverse group of different academic disciplines across the university, most of whom have never worked together before. So we're bringing together people on the climate and on the conservation side of things who haven't traditionally worked together. We're bringing in people on the technology side of things to bring in both new technologies to do measurements of all sorts of different kinds. And I can give you one example of that in a second. We're bringing in the computer scientists to make sure we're using the latest data science machine learning and AI to help support our monitoring frameworks. We're bringing in the social scientists and the economists and the business school said that we've got that broader holistic view together as well. And then there's a very wide range of external partners that we brought in additionally from across the piece, people from the policy side of things, people from the NGO community, people from many different aspects of the business community, but also people from community groups themselves. And one of the things that we found most powerful is being able to connect together to existing community groups across many different levels, many of whom themselves are membership groups covering other organisations so that we're really starting to create this in a very much a co-designed way. And that's been critically important. So to give you just one example of some of the innovations that we're trying to bring into this project, critical to understanding what optimal solutions to alternative land management practices might be is monitoring what the current state of the landscape is in terms of different climate and biodiversity metrics and understanding how that might change under different possible alternative land management practices. One of the ways in which we're trying to introduce innovation is looking at how we can adopt much broader monitoring of those landscapes. And we've, particularly technology that we started to adopt is implementing low cost sensor networks. And in particular, we're taking, we're transferring knowledge from the monitoring of urban air quality and using very similar sensors that are used in urban air quality context in the field, first of its kind, so that we can create a much clearer map of different greenhouse gas emissions. And those sensors, you can see photographs here, we installed them at a farm level at the end of last year and the first data we're expecting to be coming in over the next few weeks. And it's just one example where we've linked together and transferred knowledge from across a different discipline into this particular discipline. So now I'd like to move on to a specific example of where we're using AI to start to tackle environmental problems. And again, as was mentioned in the introduction, we're essentially looking at using AI in two different ways. The first isn't related to what I was just describing in terms of assessing and monitoring risks. And the second is how we can use AI in terms of helping to contribute to developing sustainable solutions. And I'm going to give you an example, first of all, on how we're using AI in terms of actually monitoring risks. Flooding is an important concern in terms of future climate change, both in terms of coastal flooding and also in terms of river flooding. In both instances in different ways, we're looking at how we can create new ways of producing future forecasts of flood events using machine learning and AI based approaches. So in terms of coastal flooding, what we've been doing is developing new models of coastal flooding where we are using past records of impactful coastal flooding and training models to classify whether or not there is or isn't likely to be a flooding event based on different climate variables. And then we're using that model that we've, classification model that we've trained to predict the risk of future coastal flood events based on climate model output. And this is an entirely new way of creating coastal flood forecasts. We've also been looking at river flow forecasts, looking at peak flow rates in terms of rivers. And similarly, looking to train a model based on climate variables and then use that to predict out into the future. One of the things that we've been trying to do in terms of developing that model is produce a method of producing forecasts that would work anywhere in the world. And so for the UK in particular, we have very good river gauge measurements that we can train models, machine learning models on. But for many other parts of the world, we don't have that record. So what we've been looking at is how can we develop a model that's trained on UK data but could be used, applied elsewhere on data where we don't have that river gauge measurements. And so what we've done in order to do that is include not just the information about the river gauge rivers themselves, but also information about the catchment because that is information that we conceivably could have for places where we don't have the gauged measurements. You can see from the, these are actually actual results for different rivers in the UK. Most of them are projections and the actual observations are doing very well. Where you see the data points align with that solid line, diagonal solid line is where we've got a good model. But you see one instance where we don't have a good model and that one instance where we don't have a good model of the peak flow rates is where there's large amounts of abstraction of water from that river and that's resulted in difficulties in our prediction system. But we've been able to account for that by including descriptions that give an indication of the likelihood of the being, for example, large amounts of abstraction from those rivers and so being able to create a system that is actually really quite powerful in terms of producing peak flow rates that we can apply elsewhere in the world. It does depend on information about the catchment and we're also looking at how we can use the latest machine learning AI approaches to automatically determine from various different satellite remote sensing data critical information about the catchments that can then be fed into those machine learning models. So all together this is really transforming our abilities to be able to make projections into the future about some of these critical risks. Now the last example that I want to just speak briefly about is how we're starting to create integrated solutions to environmental challenges, bringing together many of the things that I've just described and the example that I was going to give is a relatively new center that we've launched in the university focused around carbon credits. This is bringing together that expertise from the environmental sciences with computer science expertise and indeed expertise from the business and finance side of things. What we have been this actually initially started from the university itself wanting to understand how from an operational perspective we could be looking to implement robust offsetting schemes for some of the emissions that we currently don't know how to eliminate and so we wanted to look at the different potential offsetting schemes that we might be able to invest in and assess how robust they might be. We brought in our own academics from the conservation side of things to help us make those assessments and then we started to think well actually we could be looking at how we can bring in our computer scientists and others to actually create an entire as it says here decentralized marketplace around carbon credits and the idea here is that we have created a framework for robustly analyzing different offsetting schemes both in terms of their climate impact but also broader impacts associated with their impacts on biodiversity or on social challenges as well as critical issues associated with for example the permanence of the offsetting schemes. We've implemented based on the input from the computer science side of things mechanisms for monitoring and verifying in a robust and indeed tradable way those different offsetting schemes and thereby creating this idea of a decentralized marketplace where through a form of blockchain technology we're able to keep a traceability of exactly that robustness of the different carbon credits that could be onwardly traded. So I hope that gives a little bit of a sense of some of the things that we're looking at as I say the critical element to all of these is radical collaboration internally within the university but then critically also outside of the university and then the very final thing that I was going to describe is just Cambridge Zero itself if you're interested in finding more you can visit our website Cambridge Zero is the university's umbrella for all the activities that we're undertaking associated with climate and that includes all the research activities that I've you know some of which I've just been describing but also all the activities we're undertaking in terms of education both of our own students internals Cambridge but also the education we can provide beyond the university at all stages of the lifelong learning journey so looking at school education and particularly working with our partners in Cambridge University Press and Assessment as well as beyond university including executive education and engaging with education on the policy side of things and then looking at how we can put in place processes whether that's through industrial collaboration or entrepreneurship for example to ensure that ideas and innovations happening within the university are rapidly moved out into real world deployment and scale up and then finally as I was just describing with the offsetting side of things we're looking at how we can both look at ambitious decarbonisation and supporting sustainability within the university's operations but critically how we can ensure that we're bringing in our own academic and other expertise to help inform that ambitious policies so I hope that's given you a little bit of a flavour of what we're doing in Cambridge I'm going to be really delighted to take questions that's absolutely brilliant professor thank you so much really really riveting really interesting I have a couple of questions myself just to kick things off but we're joined by a fairly sizeable group here hello everybody so I look forward to your questions flooding in a really practical one just on what you were saying momentarily about Cambridge zero can you just talk about where it originated like was it something that there was you know particular pioneers you know was it because of the work of a person or a group of people or could you say yeah so in a sense it was bought an off and top down within the university so the bottom upside of things came very much from the students in the university wanting the university to be doing more particularly around climate so there was a strong impetus from that direction and the top down perspective is that overall the university's mission is to contribute to society through its education and research and broader engagement and there was a sense that to be true to the university's mission we really ought to be doing something significant on this on this challenge and so those two things came together and in 2019 November 2019 was when we officially launched Cambridge zero and we're still very much spinning it up but you know we've achieved a huge amount over the last couple of years in terms of really starting to make a step change difference I think how we how we respond Cambridge is a place I know well I love very much and I would always think if I think the university began in 1209 so if it can change then surely we all can you know yeah um I've one very specific question I'll put to you then I've others I hold and reserve as I want to privilege those who are asking questions online when you had the lovely image of the tree and when you were talking about the the whole systems approach without being too hard-nosed about it I noticed one of the roots made reference to incentives can can you say anything more about what sorts of incentives you envisage yeah well I mean I think we were in large parts um thinking about through the lens it very particularly in the UK context of the you know post-eu potential land management incentive is the most obvious direct sense you know slightly in that slightly broader sense where we're looking very much at multiple stakeholders who are using these landscapes then it's probably in a slightly broader way than just the pure financial incentives that might be associated with different future subsidy regimes which is no which is one of the critical elements but it's just as much about in in a much broader sense understanding how to encourage different approaches so I'll give you a specific a specific example what I've quickly discovered we've quickly discovered that there are a very diverse range of different users of landscapes who can if they're not involved at an early stage and helping to think through different solutions can create significant barriers to them ever happening and so one of the one of the potential solutions to to the challenges as I was sort of alluding to is around rewetting some of the landscape we're looking at for example how we might be able to rather than we rewet the entire landscape which has all sorts of potential implications for different users of that landscape do something that's more nuanced and maybe rewet use the drainage ditches that exist in the landscape to enable us to rewet almost on a rotation basis different you know field by field where there've been issues in the past with different proposals they've often come from very diverse different groups so the anglers get upset if you talk about doing something different to some of the waterways in the region or the archaeologists get upset if you thought you know so it's understanding where all those different potential users of landscapes how they're using those landscapes and how you can collectively incentivize in a in a much more broader sense of the word the involvement and and by and of all those different groups seems to be critical to the success of this extremely interesting that's kind of part of the whole systems approach is understanding where those different perspectives coming from and therefore where the barriers are but more importantly where the opportunities are yeah that's I think that's lovely language that I've heard you use before the idea of removing these barriers and trying breakdown silos which is what we're trying to do as an institute as well to break down silos between different forms of thinking I'm going to turn to some of our questions now professor so the first one I'll put to you is from peter mclune peter mclune is a trade unionist and a board member at the institute hello peter I'll read at some small length although we seem to have the tools to anticipate catastrophic events resulting from the climate crisis we are with that we are currently experiencing we seem to be light years away from developing technological solutions to impending disaster uh excuse me surely changes to human behavior at every level in every part of the globe remains the only solution that will reverse emissions I'll put the question part to you again because I kind of stumbled there forgive me surely changes to human behavior at every level in every part of the globe remains the only solution that will reverse emissions yeah so this is you know can it's technology is just going to solve this for us what's that or can we not wait on that for technology to solve it or do we not you know do we not really need to be encouraging significant behavior change and there's a sense in which I mean we need both and actually both and also the interaction between the two um so take for example transport we know that in the UK currently transport accounts for the largest share of our of our emissions um since we've uh significantly decarbonized our energy um supply and uh well you know on one level we know that there's somewhat of a technology solution to the to the transport challenge in the sense of electrification but we also know that behavioral change is also going to play a significant role in that we know that investment in public services is critically important but it's actually again taking that whole systems approach um there's also an element that is um more fundamental which is around which is around land use which is around land use planning which is around looking at how we can design neighborhoods in a way that eliminates the need for so much um mechanized transport so that you can walk or cycle to critical services um and so it's about taking that much more holistic approach to generating solutions that does look from the perspective of multiple different users so that we can integrate technology but also integrate and support the behavioral changes that are required. I think there's a a natural follow on questionnaire from Francis Jacobs who's the former director of the EPA's on office in Dublin and a member of the institute hello Francis. Francis says it makes so much sense to have integrated solutions to environmental problems. Has there been much resistance from individual disciplines and stakeholders to working together? I'm sure this is something you probably have a whole unit at Cambridge Zero puzzling over this. What can you tell us about resistance? So actually I would say exactly the opposite. If this is done right then I've then it's been really inspiring because people like it's really critical to create a trusted environment to allow that sort of dialogues to occur but if you do then I have discovered both within the university and also outside the university that it becomes a hugely exciting forum for people getting energised by the opportunity for creating new solutions so in the research environment we often say that the biggest advances are made at the interfaces between different disciplines because that's where you can really you know in a sense it's some of the example I gave on the local sensors where we've taken something that was done in the urban air quality arena and taken it into the field. Those might sound like they're not very different but they're very different disciplinary silos. People normally doing urban air quality don't normally have anything to do with the people doing conservation and rural work and so bringing those two together has been the way in which we've been able to create a significant step change in terms of how we've done things. That happens and it creates excitement in the research environment and we've found similarly that if we bring groups together in a trusted way so that they're actually able to talk constructively together then we find exactly that same level of excitement by bringing together if you like state user groups different user groups. And I was trying to divine it from your presentation I'm from my knowledge of Cambridge zero but Cambridge is it's not unique but one of the special features of the town is how integrated the university is with the town right that the university is spread throughout it but there's obviously Chinese walls and barriers of various kinds but where it does work can you talk about the engagement between the university and the town specifically whether it's the schools or local authority or businesses or anything like that. Yeah so what I would say is that Cambridge is is spread out across the the city of Cambridge but actually historically we've not really integrated in the broader region and so we're very much forging away in terms of doing that and we're doing it through different mechanisms significantly integrating with the various levels of local government that exist in the region but also looking at leveraging any networks that we can to create those trusted communications channels and one of those networks is the schools network because schools are central to every community and so we're one of the unique things about Cambridge is that associated with our department for education we have a state primary school that is the only exemplar teaching connected as in sorry you know it's connected to a research environment teaching school in the country but then you know we've been working closely with that primary school and then the network of primary schools in the region to really look to see how we can get deeply embedded in different communities so that's just one example of where we're trying to utilize the the networks and channels that we have to make sure that we're really embedded in a very in a very deep way and to understanding the challenges at a root cause level of different communities and also the opportunities and the potential solutions. My connection with Cambridge began at Hamilton a long time ago and obviously some work at Hamilton for those on the call have a strong tradition of teaching education and I can vouch for that connection between the schools and the university it's really it's a lovely thing and something that there are similar ventures in Ireland as well with things involving schools and the environment but we can certainly learn from Cambridge University experience I'm sure. To follow up on Francis's question before taking a different tack Francis just said the previous question I put to you professor for example how is the farming community responded to your ideas of rewetting for increasing nature protection if you can just give that as a specific example relations with the farming community. Well I can be kind of candid that you know historically the farming community has often been at war with the conservation community because they feel they felt very much threatened by the idea that you know somebody or and indeed with the policy community I have to say as well the idea that somebody a Westminster might come out into the fence and tell the farmers in the fence what they might want to do doesn't go down too well as you might as you might imagine and and so that's where you know we've been working really hard to bring to help support constructive dialogue between the farming communities and the conservation communities in particular and and where we have you know many of the innovative solutions that we have been starting to surface have been coming from the farming community themselves and what we've discovered is that there's nobody who better understands the land than the farmers who for many instances have been farming that land for generations and therefore they are they have critical knowledge in terms of helping to develop the solutions and the one thing that I have learned from talking to many of them is that more than many other groups that I talk to in terms of the climate related challenges they really understand the long term and then and the need to leave land in a state that is going to be sustainable for generations and so they're very committed to helping us think through what the potential solutions are but it's about doing that that trust is critical trust and fairness I think of the two most important words that I've that I've understood associated with generating solutions to climate related challenges that work for different communities. Okay in that spirit Professor Schuchberg there is a ton of questions that are coming in and I'm trying to organize them both in my mind and also loop some together when they're thematically related so I'm going to put one further question to you now that regard it's building on your discussion of farming then have a question about carbon carbon credits and then a question about international collaboration so we're moving around a little bit and thanks very much for your for your counter in doing so Alan Matthews says a criticism of trying to integrate thousands of farms into greenhouse gas accounting is that the transaction costs of measuring and monitoring emissions will be very high looking at your small inexpensive sensor how do you see the balance in the future between actual measurement of emissions and sequestration on farms and the modeling of emissions and sequestration using your AI algorithms take a moment if you wish yeah no well actually I mean that's one of the really the things that we're looking at in terms of the AI algorithms because one of the things that you can do with AI is integrate data sets of different levels of fidelity so in terms of looking at global greenhouse gas inventories then we can that we are looking at how we can integrate together the data sets we have from remote sensing on the one hand the national country level data sets that we have and then different you know even down to field level data sets how we can integrate them them together and it is possible to use machine learning approaches in order to do that and and then it's also possible and you know another another key question that's also asked about this is in the moment we don't have good ways of being able to do those inventories at anything other than a national level but it would be interesting and useful to get a much greater level of granularity and so that is another key area of research that we're that we're looking at and and to the international partnerships question absolutely not Cambridge alone I mean this is very much an international collaboration that's required to do that sort of good that sort of approach can I ask just a really a real ladybird question it's just for myself when you were doing you mentioned during the presentation and this is before I get on to Mike and Daryl's questions you said that the sensors were installed in November 22 and data is expected shortly just in like the most basic terms what sort of data are you actually gathering I mean I can guess but could you just explain it for an expert oh this I mean those particular sensors yeah that so it is just carbon well primarily carbon flux measurements so literally the amount of carbon dioxide that's coming out of the fields and we're also looking at methane as well so carbon dioxide and methane fluxes and one of the so normally those are measured by very expensive flux towers and in order to be able to take the measurements from a particular sensor a small local sensor without the fancy equipment on a flux tower we have to combine together the direct measurements that are made by the sensor with some modeling assumptions to account for you know aspects to do with the wind for example that can have implications for the fluxes but getting back to I think was Peter McLoone's question it does sound like relatively affordable by kind of you know as compared to what some technology costs I mean it sounds like a relatively affordable way to do what you do right well so a typical annual cost of a flux tower can be you know 50,000 or 100,000 a year they're they're expensive whereas these sensors really you know it's a few pounds amazing the big difference but it's a different they give you different information and so it's like having the combination the two and being able to integrate them it's what's really powerful yeah very good a bit of triangulation and for Mike Brennan who is Dr. Mike Brennan senior researcher environmental science at the house of the Eroctus so the Irish House of Parliament hello Mike could you speak a bit more about the carbon credits and what measures are in place to safeguard let's go question against large cash-rich polluters coming in and monopolizing these offsets? Well yes good question so the the aspects that we have been looking at are as I briefly described associated with the robustness of the the credits themselves so we've been looking at the leakage the permanence and the additionality in particular so that I mean often this is easiest thought of in offsetting schemes are not only about tree planting in fact very definitely not only about tree planting but often it's easiest to understand what those concepts are in terms of tree planting so the permanence issue is around you know you might plant a tree and either it dies or it gets cut down so it's not necessarily permanent right so that's the permanence issue and you can monitor for that using remote sensing satellite information over time and the leakage aspect is that you might do reforestation or avoided deforestation in one area and all that's happening is that you're displacing that from somewhere from somewhere else so either you're stopping trees being cut down in one region and actually just the same number of trees are being cut down somewhere else instead or you know a similar form of leakage so you want to make sure you're accounting for that so that's another aspect that we're taking account of and quantifying and also monitoring and then the third aspect is the additionality so you might say well you know whatever scheme it is is not you know does have some leakage associated with it is not entirely permanent but you still want to invest in it because there are all sorts of other benefits that you want to take into account of in some way as well so those aspects are the aspects that we are currently in the current format of what we're looking at focused on you know it is also true that there are all sorts of other ways in which the schemes could be exploited and overarching in terms of offsetting and carbon credits that certainly what the university's policy is but what I would encourage everyone else's policy should be is that they are something of the last resort everything that you could do to eliminate your emissions should be done first before looking at investing in offsetting schemes because otherwise and I think this is you know where a lot of the credibility of offsetting schemes at the moment is under scrutiny for you know effectively greenwashing and worse is that there simply isn't enough nature to go around or there isn't enough opportunities to go around for everyone to offset all their emissions so there's a huge opportunity to double counting you know and all sorts of other challenges. Thanks professor there's a question now from Darrell Gunning and this was the international collaboration question I promised so to quote apologizing for raising the b words the UK's withdrawal from the EU but have you found that international collaboration has been made more difficult following brexit and how are you dealing with this considering climate change is an international crisis and that's Darrell Gunning marine biologist at the EPA. Well I mean you know clearly in many ways yes yeah the there were many many particularly EU projects that we were significantly involved in we still have an involvement in those but obviously on different terms which is disappointing but at the same time you know pretty much everything that we do on the environmental side of things on one level has to be done at a global level and so that global collaboration is absolutely continuing so you know it's a sort of mixed story really. Okay that's a very different matter to you professor and a handful more I'll go to when I loop two questions together now I'll start with Philip Beck. Philip Beck who is an IIA member hello Philip asks it's rather specific are you looking at modeling other ecosystem services accruing due to afforestation for example soil erosion reduction biodiversity increases and water bodies due to overshading by native trees etc thanks for the great presentation. Yes so particularly in the fennlands there's not very many trees as you saw in the photographs that I showed you but the other landscapes that we're going on to look at in the Kangolms and in Cumbria trees have a much more prominent part of the landscape so those some of those other aspects are absolutely things that we'll be looking at in some of the other landscapes and in terms more generally of you know some of the research initiatives that are taking part in the university they're very broad and encompassed many of those different things. Okay very good thank you very much and then I'll go to a question from Francis Nagade on thank you for being with us Francis asks hi on the use of AI to forecast flooding is there any data yet on how accurate this has been that's Francis McGaion who's from the Department of Agriculture Food and the Marine. Yeah so the two examples that I gave of coastal flooding and the river flow flooding so the one that's more progressed in terms of the research that we've been undertaking is the is the river flow as I think I was you know I showed you some of the results that are looking very promising in terms of that approach. The coastal flooding example is somewhat earlier stage in terms of the approach in part we're looking at slightly different things in the two contexts so the work that we've done to date on the river flooding was the thing we were trying to project was peak flow rates in the river on the coastal flooding thing side of things we what we're actually trying to predict is slightly different and in fact one of the things we discovered early on was when we looked at the records of there's a visit for the somebody there's been some research done on the UK context looking at the most impactful as in the most amount of damage done by different coastal flood events and when we looked at the dataset interestingly that didn't exactly correspond with the highest tidal levels and so understand so it's not so so there's two slightly different things that you can project one is peak flow and or highest tidal levels and the other is the amount of damage done in different contexts and so what we're trying to do at the moment with the coastal flooding side of things is really understand what exactly the different criteria are for creating maximum damage associated with floods because as I say it doesn't seem to be exactly that it's the highest tidal levels and that's the other thing that we're that that's a real benefit of these more data-driven approaches compared to the more traditional physics-based approaches is that we can you we can look at data-driven approaches to create that bridge between the physically defined quantities and the things that you know the things that actually impact whether or not that is economic damage or whether it's damage to health for example from from different environmental environmental extremes extremely interesting this is all it's really interesting I'm going to tack on a preface to a question we have I noticed when you when you had the river flow forecast part of your presentation and the map obviously and understandably covers the four nations of the UK but obviously this country is physically integrated with the United Kingdom and I'm just wondering is there anything and perhaps you don't but is there anything that you know about the all the stuff you said about coastal erosion and peatlands and stuff absolutely the same in in Ireland matters of central political importance that's my preface to Dara Lawler's question do you think your very interesting framework could be applied within Ireland and if so how and is there anything kind of that you find curious about the case in Ireland that would make it different or similar to the case of the Fendlands I think I mean I'm sure we could exactly transfer the things that we've been doing into the Irish context or indeed anywhere else the I mean so for all the machine learning AI type approaches that we've been developing the most critical thing is to have the underlying data sets in order to be able to train models on as I was describing with the river flow example we're also looking at approaches where we can then do transfer learning from an environment where we do have data to an environment where we have less data and so that is also possible but in terms of looking at similar issues for in the context of Ireland then I'm I mean I'm sure that much of it is directly transferable at very least in terms of methodologies excellent and actually I should say you know if anybody wants to reach out if we're we'd be delighted to collaborate and share any of the work that we've done today so please do reach out there is at least one person who's made contact to ask if they can make contact with you in fact professor so we'll share their contact details with with you after the after the call I was going to put one more kind of specific question from from an attendee to you and then I'll have one final kind of general question but indeed to the audience I think we've gotten through a great group of questions and thank you to professor Shookbook for going through them so carefully but if you have any final ones now is your time there is one I'd like to put to you from Keelan O'Sullivan who is our climate and energy researcher currently at the IIA Keelan says it is clear from your remarks that AI is crucial to mitigating and responding to climate change what are some of the barriers that exist in operationalizing AI generally to its full potential in this regard and are there areas where key investment is still needed I guess that's about how can we persuade people that AI our artificial intelligence is is good and people needn't be afraid of its usage um well actually I think there's two things there actually one is one is the last point you made about how do we you know make set AI in a in a responsible setting actually it was the critical bit and it comes maybe in some ways it comes back to that trust word that I was using in a different context earlier um the other area is actually just from the practicalities side of it so what we have found at least that you know this might change over time but at least at the current time to really maximize the use of machine learning and AI technologies you have to have a an environment where you can bring together those people with the computer science background you understand where the cutting edge and opportunities for in terms of new AI technologies and approaches together with people with the domain expertise on the environmental science side of things or whichever other application domain you're wanting to apply the AI to um and really um create an environment where they can work very collaboratively together and understand that that both sides um because there's still you know it's getting that that um interaction between those two communities working effectively that's critically important and so investment in bringing those two together I think is absolutely critical and that I mean in the UK there is a lot of investment currently being made exactly on that um interface so I'm hopeful that actually if you if we roll forward a couple of years then that situation might be quite different because we'll have made those bridges and we'll have created if you like a whole community of people who already you know who are trained in understanding both areas and then had that you know we're not it isn't two different communities but at the moment it still is and that's where the greatest barrier currently is but doing all of that crucially in a context of understanding the broader responsible innovation aspects is is clearly very important especially on anything that involves vast amount of data much of which can end up even if you think about it you know even if you think compared to health data for example much of the um environmental data isn't so personalized but there are still when you get below the surface there are many aspects which do have those sort of ethical questions that need to be considered very much it's remarkable how your work is so it's at the cutting edge of so much new stuff you know AI and climate and various technologies but it still boils down to that kind of fundamental notion of trust right and of kind of people knowing how to work together in an ethical fashion we're going to wrap up pretty presently I'm going to throw down a challenge to any of my colleagues I'm looking for the recording of our event from before Christmas with Kathy O'Neill the weapons of math destruction which is a brilliant session that we had I think it was in November and if anyone can find it for shared in the chat for those attending it's on the website but it's really good and it was getting to the kind of the similar kernel professor what you were talking about about the kind of no ability of artificial intelligence and kind of understanding her decisions are made and stuff and it's a it's a puzzle that's going to remain with us I think a very general final question to professor and then you by all means you'll have a minute to wrap up with anything else you want to say but Cambridge Zero is there a is there a shelf life is there is there an end game what are your medium to long-term plans for this excellent initiative if you could share it yeah so I mean our our mission with Cambridge Zero is really to support ambitious climate action within the university and and everything we can do to facilitate that beyond the university and we I mean is there a shelf life it's it's very much a challenge oriented initiative so whilst climate change remains a significant global challenge then we are wanting to ensure that the university is bringing all its assets and resources towards helping to support the response to that I you know in a the shelf life one hopes is short because we hope that the global community comes together rapidly and addresses the climate challenge unfortunately we don't appear to be on that track at the moment but as a global community but the university is absolutely committed to to to really maximizing its its response to that challenge fantastic thank you I thought I'd been provocative with that question actually and good to know okay I um I think we'll call it a day there I wasn't able to find the link it's gonna be on our youtube channel for Kathy O'Neill as will this event as well one of you one of the questioners asked if recording will be available it'll be both on the IIA website and indeed on youtube as well so just to thank professor Schockberg really both for your excellent presentation but also for your brilliant work it's highly inspiring there's lots of people on the call working in various parts of public policy mainly in Ireland but also elsewhere and I'm sure you've provided loads of food for thought so thanks and good luck with your important work thank you to the EPA for this sponsoring the series and looking forward to continuing this great collaboration and thank you to my team at the IIA for pulling it all together and finally thanks of course to our audience and for your excellent questions I wish you all a great afternoon thanks all