 Hello everybody, my name is Christian Poe and I'm Alex Howard and today we are going to talk about an open source exploration of a semantic structure for climate accounting. So let's start with introduction. The work that we've been doing here we've we've been doing in the context of the hyperlegia climate action and accounting special interest group. And specifically of the standards working group of which I'm the chair. My name is Christian Poe I work for the Nova Institute we are are an independent not for profit company in Southern Africa. We work to promote the quality of life of low income households in Southern Africa. A lot of the work I've been doing relates to energy and air pollution in low income communities in South Africa. And since 2006, we have been active in the voluntary carbon market in developing projects and implementing them to generate voluntary emission reductions for sale to finance these projects. So one of the working group my interest is specifically to to work to find those deep ontological structures that will enable us to build a global impact accounting framework. I have been working with Christian for the greatest part of seven years now. For most of it I've actually been working alongside you at Nova Institute but nowadays I am an independent software developer. And it is in that capacity that I continue to be involved in the hyperlegia climate action accounting special interest group and specifically the standards working group. Yeah, that's me. I acknowledge our collaborators that work with us on this presentation. Carl Robinson from Braille Tech Consulting in Canada and Alfonso Guevella from Mexico from the hyperlegia Latino America. So in terms of the challenge that we are facing I don't have to really give you a lot of data I think all of you are painfully aware of the challenge of climate change. But here is just to set the context just to two graphs, both from Wikipedia in the spirit of open information. The one on the left shows temperature changes in the last 50 years. And you can see that especially in the northern hemisphere, those changes have been a dramatic. But also over the continent of Africa, Asia, and Europe. There have been substantial changes in temperature. And those changes you can see on the in the graph on the on the right correspond to humanly induced climate forces. So, as a global community we have to avoid catastrophic climate change. And that means we must dramatically reduce greenhouse gas emissions and increase common sense. And for this to be to be even doable to be even on the cards climate accounting must be become ubiquitous. We must, we must get a grip on who emits what and for what reason, and who is responsible for what and what can be done to avoid that. So you cannot manage what you do not know. And you know by accounting for for the processes that you are in control. There is no problem. There are in fact many different ways in which climate accounting takes place. There are many different practices, there are different protocols, there are different standards, and they are used for different purposes. And if we want to, to unite and manage the planet as a plan. These instances must at least be interoperable. Now, let's just think about what interoperability means. It means to be able to use information between different devices systems and frameworks. And there's in fact a hierarchy of interoperability in the context that we are working now and the first one is just description. Describe something in a way that it can be understood within another framework or in another context or another jurisdiction. You require stable terminology. Another level in the in the hierarchy of interoperability is is comparison to compete the things you need to express them in the same terms. You need a stable terminology to start with. And for comparison, you need now to express that which want to compare in the same terminology and aggregation is even further up the hierarchy. It means you must express, not only in the same terminology but in the same metric. And then you can aggregate whether that is is is something or whatever the case. These things are difficult. When you think of the global climate accounting space description and together with description goes discovery because if you can describe something you know what to look for the description and discovery is difficult because people don't use the same words to mean the same thing. There's a lack of a stable terminology. People use words like the environment, climate neutral, carbon neutral, but they don't always denote the same thing when they use that. People also are guided by different interests when they are speaking about impact. There are people that are in a situation where they want to play downplay the impact or exaggerate. If we are going to move to a system, a global system where machine where there are machine readable claims about environmental impact that cannot be done without a shared language. Because it will either just compound the confusion that they currently is by aggregating things that do not belong together or comparing things that that are really not the same. Or it will fragment into into just a large variety of incomparable systems and then no one can have a grip on the global situation. So, that is just that description so comparing is difficult in this in the circumstances because you have to. There are differences in what is accounted for and how accounting takes place. When you think about what is accounted for, there are differences in boundaries, for example in organizational values or in activity values, or even in the greenhouse gases that accounting takes place for. And even when those things are the same, even if you account for the same organizational boundaries, the same activity boundaries, the same greenhouse, greenhouse agents, there are different calculation methods. And those things are not always made explicit. Just a brief example from my field of work, so that to highlight this problem, even things that look the same are not the same. If you compare two projects that can be labeled in the same way. I show them on the screen there, there's a voluntary carbon market, a voluntary carbon standard project VCS 2505. Its title is fuel efficient cooking in South Africa. And there's another project in the gold standard GS 4536. Brickstyle Woodstove Maslava area, I'm involved in that one. So if you just look at them, they are, they can be labeled the same. They are voluntary carbon market projects with improved cook stoves in South Africa. But they differ quite significant. The one encompasses the whole country and the one just a specific region. The one estimates the baseline from into my mind in the Eve calculation using generic country level data. And the other one is based on empirical observations and household surveys in the area of application. And they have different methodologies which derive more or less from the same context, but they are not exactly the same. And it is actually very difficult to take these things like this and to understand what that result that resulting the or the voluntary emission reduction, what that actually needs. And this takes this takes place across across the field of of climate. And it really follows that when we description and comparison is difficult thing, then that aggregation is also a problem. Because you need for aggregation, the stable terminology expression in the same terms in the same metrics, and then with the same assumptions, things like emission factors or counting here. So when you don't have that you have this this this concept of blurring and semantic zoom out. In other words, if you if you have to, if you compare two periods but they're not exactly the same bit or you don't know if they're exactly the same. You can say, you can say over the past year over the year 2019 you can say over the period preceding 2020 or in the recent past, and you become even all the more imprecise to the point where, like they said Woody Allen, when he said he read speed speed read one piece in 20 minutes he said it was about Russia. So, what do we do about this maybe you say okay, maybe we must standardize this, let's standardize everything and then everything will be. That's a, that's a very strong idea and a lot of people have, have, have pursued that. And they were right in doing so, but they just think what, what, what are standards. In fact, we are from the standards working group so let's just think what what standards are is this the solution. Standards aggregate norms that help shape interest, constrain behavior, prescribe action and support a logic of appropriate appropriateness and consequences. And so what are norms. Norms are social constructs that emerge from persuasion, cascade to acceptance and internalize compliant behavior. So it stabilizes people's behavior. It standards make things happen in the same way. So that's that constrain behavior, prescribe action, and it lets consequences follow from compliance and non-compliance of those standards. So, maybe this will help. And there are in fact a lot of staff in the field of life accounting. This is not exhaustive that this is just an example of a few that's out there and that that we've looked at and they vary quite a bit. There is, there are standards for specific jurisdictions like the from in the US in the US or Canada that's here on the screen. There are standards driven by business, the business community for investors. Standards have different aims. And let's just look at a few examples of those aims and then you'll understand the diversity and in a certain sense the problem of it. So some of the standards work from the perspective of the investor like the ISP. The ISP aims to provide investors information about sustainability related risks and opportunities. It's focused on the investor, it's focused on risk and the opportunity. In other words, the investor as an investor looking at a company or some investment. And it helps that person understand the sustainability related risk components of that, of that decision. And this shapes the way in which that stand in front. If you think about, there's another perspective, there's a regulatory perspective. There are compliance to international treaties and that cascades down in regulation in individual countries, like the Parisian. Then there are standards for voluntary climate action. One of the most well-known is the Gold Standard Foundation and the VCX. And they represent, I would say two different views on voluntary climate action, the Gold Standard, a broader view. In other words, embedding that in what it calls environmental integrity and sustainable development. And the VCX is much more focused. And those are both, it can be argued, legitimate ways of proceeding. And then you can also have a project focus or you can have institutional focus. And this is something like the Greenhouse gas protocol. The Greenhouse gas protocol has a corporate accounting standard and a project protocol. And the corporate accounting standard is for businesses or for corporate, it is for their business as usual. And the project standard is for those things that you do extra in order to benefit the environment. So that is a project. That's not your business as usual, that's something. And there are different things that come into play when you take each one of these perspectives. So standards are different interests. But standards also use different metrics. You can use a performance metric. Or there are standards focusing on emission reductions or emission avoidance. And then there is a focus on net emissions. So the difference between those is a performance standard is about efficiency. And that is the acknowledgement that there are sometimes more than one goal. And that for the efficient allocation of resources, you have to balance those. And that's why efficiency is an important goal. And that's a legitimate way of looking at the problem. Emission avoidance compares what happened against what would have happened. So a counterfactual basis. And that aims to create a situation better than it would have been. And then net emissions is something different as well. That just balances really emissions and sinks. And so I want to sink more than I. Without reference to what would have happened. So emission avoidance, you can still emit. It would be a net emitter but you can say well it's less than it would have been. There's a move towards net emissions as we proceed into the 21st century. But there are different metrics and it's not so easy to compare things expressed in these different metrics. So is standardisation a solution? Well, yes and no. We cannot go back to these people and say you know what, me and some other people at a Zoom call the other day and we decide you should really standardise your things and then we can all work together. Because they all represent specific interests, they've developed these standards for specific purpose, it serves a specific community. And it's not for us to say that it's not a legitimacy. So there are different needs and there are different interests and there are different ways of looking at something. And it is to be prescriptive is inappropriate and it will not work anymore. So making just another standard. Just trying to convince people that are well established in what they do to do something different for the sake of harmonising standards is not going to do. It's not going to work. So what must we do? Our suggestion is that we must find the conceptual bedrock on which these standards are built. And that we must make that explicit and create a way in which these standards can be expressed in terms of that conceptual bedrock, that underlying ontology. And that we can build from that ontology to show how these things interrelate. And to build tools with which we can do comparisons and aggregation. And one of the examples that we've been thinking about is the origin of HTML of the World Wide Web. And that demonstrates the power of consensus on a profoundly simple opportunity generating enabling frame. So that's what we want to try to do on earth. We want to say it's not a standard. Let's find that ontology that lies underneath that. Let's make that explicit. And let's build on that tools with which we can take that which exists, which exists for specific purpose and a legitimate need. Let's make that translatable through going back to the core concepts. And, and make that interoperable in that, in that way. I give hand over to Alex to show you what we've done so. Thank you, Christian. Before I start, I would just like to mention two things. First of all, the ontology is a work in progress. It is by no means complete yet. It's an open collaborative effort. So if at any point in time during this section of the presentation, you have questions or ideas or suggestions, please feel free to post it in the Q&A section. And we'll try to get back to your comments during the Q&A session. If we don't, we would very much love to engage with you afterwards while some other platform means. The second thing I want to mention is that we'll be using the terms classes and axioms during this part of the presentation. By classes, we mean representations of the things that reality is made up of from the perspective of impact accounting. And with axioms, we mean statements that define the relationships between those entities, between those costs. Right. And the very basic statement that the entire ontology is built upon is that an agent engages in an activity that impacts an environment. In other words, somebody does something and it affects their surroundings. With this premise, we can see that the ontology must at least have the following three classes, one for agents, one for activities, and one for environments. And axioms that go with them are firstly, that an agent engages in an activity, and secondly, that an activity impacts an environment. This is obviously not enough. So now let's look at each of these three. And ask what more do we need to add to the ontology to be able to understand and describe these entities better. We'll start with environments. When we are dealing with impact accounting, the things that we typically want to know about environment is where is this environment. So what are its physical boundaries? When are we looking at this environment? It doesn't change all the time. So if we want to say something about environment, we also have to say, at what point in time, I was saying this about the environment or for which period in time do our statements apply to the environment. We also want to know other properties of the environment and their dimensions, like what is in the environment and which of them are of interest to us. So to capture these things in the ontology, we need a parameter clause. These things are all parameters that we want to know about the environment. There are things that we can either measure or quantify and gauge in some way. The axiom that goes with the parameter clause is that an environment is defined by parameters. We choose which aspects of an environment are important to us in our specific accounting scenario and that choice of parameters define the environment of interest to us. But there's an important distinction between parameters and their values. While parameters remain fixed in terms of their definition, the values that they can take on change almost instantly from time to time, continuously, I mean. And these values that a parameter can take on is what we can call the state of the parameter or the state of the environment. So we need to also add the state clause to our ontology and the axiom that state is the value of a parameter p at time t. So just a quick word here, you will see that all these diagrams in this section of the presentation, start with the owl thing class at the very top. This is because we are developing this ontology as a web ontology. And according to our specific, our specification, the very top class of ones ontology must be the thing class. So all the classes that we've added up to now agents activity environment parameter and state or all subclasses of the top class thing. Now that we've done with environment, what do we need to know further about activities. There are two things. Firstly, activities, the same activity a specific activity, for example traveling them performed in different ways. And while most mostly using different instruments or using different procedures. The activity of traveling, for example, can be conducted using a car or using a bicycle using an airplane. And these different ways of performing an activity will have different impacts on the environment. So the distinguish between them, we need to add two classes to the ontology. Firstly, that instrument and secondly that procedure. And the axioms that go with them is that procedure guides and activity is like the recipe, the steps that one follows the set of actions that must be conducted to perform the activity. An instrument is well an activity on the land is performed with an instrument, you always use something to do something. The second thing about activities that the ontology must capture is that activities have inputs and outputs. It is exactly via these inputs and outputs that the activities impact their environments, they take something from the environment and they change it or do something with it and then put it back into the environment either in a different form or whatever. Now, to capture these inputs and outputs in ontology we do not need different classes, we can use the parameter class for that. We do have to add at least two more axioms, namely the following, an activity has at least one input and activity has at least one output. On the diagram on the right you can see we have instrument and is linked to activity by the axiom that an activity is performed with an instrument and procedure is linked to activity by the axiom that the procedure guides an activity. An activity has parameters in forms of inputs and outputs. Lastly, what do we need to know more about agents? Well, up to now we have already covered the relationship between agents and instruments, between agents and activities and so forth. But we have not yet covered the fact that agents can enact different roles. One of those roles in role of impact accounting while the four of the most notable roles in the discipline of impact accounting is that of owner, operator, claimant and auditor. So, we add the role class to the ontology, next to agent and all those other classes, and we say that an agent enacts a role, that is the first axiom that binds the role class to the agent class. And then we add owner, operator, claimant and auditor as subclasses to the role class and we can put down axioms for each of them. An owner owns a thing, literally anything, it can be a facility, it can be an instrument, it can be a company, whatever. An operator operates an instrument, whether it be a measurement instrument, a transportation instrument, a production instrument, whatever. For the claimant and auditor, we add the axioms that a claimant as a role is an agent who makes the claim, and an auditor is an agent who audits a claim. Then you may ask, wait, where do claims fit into this ontology? That's a good question. If we look at our premise, an agent engages in an activity that impacts an environment, there is no mentioning of claims. But impact accounting is, as a discipline, deals at its very core with claims about agents, environments, activities, and their impacts. It makes statements about it. So we can add the claim class to the ontology with the axiom that a claim is a statement about some specific thing, literally anything that you state is a claim. We can say, my name is Alex, I am making a claim, and that claim can be verified in some way. This brings us to the next question, possibly the most important, where do standards fit into this ontology? Standards guide claims. You will remember that we said activities are guided by procedures or procedure guides and activity. What is guided is, in a certain sense, a very special procedure. It's a specific type of procedure. What it guides is the activity of making a claim about something. And the concept of impact accounting that will be making a claim about one's carbon footprint or water footprint, whatever that case may be. Standards do not only guide the procedure or the action of making a claim, but also the structure of the claim itself. So therefore, we add the standard class as a subclass to the procedure in ontology, and we add the axiom that a standard guides a claim. Apart from the fact that as a subclass of procedure, it also guides the action of making a claim. With this, we have reached, well, we have covered the full ontology as it stands at this point in time. We have 15 classes and 15 axioms that are all displayed on screen right now. And if we depict our ontology as a kind of knowledge graph, we give this diagram that you see on the right with thing as a top level in orange, the top level class in orange. And the yellow circles indicating the first level subclasses of the ontology and the green indicating the second level subclasses of the ontology and these bendy arrows indicating the axioms or the relationships between these classes. You may wonder what an implementation of the ontology may look like. Let me quickly show you a UML implementation. Okay, not a UML implementation, a protobuf implementation, but here is a UML diagram showing a possible implementation of the ontology in a protobuf based system. You have the parameter class there and date and time and location as special types of parameters. And then you have, well, in protobuf terms, a message for agents, a message for instruments, a message for environments, parameters. We already said parameters, but the values, which is constitute states or whichever message they, then we have the activity message and it all wraps up into an impact claim. This is a brief note. As we have said, this ontology is still a work in progress and there remains much to be done. One of the things that we hope to do soon is to embed our ontology in the ISO, IEC basic formal ontology. That will be quite a process, but there are a myriad of benefits of doing that. Most notably because the basic formal ontology is an ISO standard in itself and it's being used by literally hundreds of endeavors and people on the internet. So if we can embed our ontology into the basic formal ontology and some of its extensions, we will be able to maximize or should be able to maximize its usefulness to the world of impact accounting. That's all for me. Thank you, Christian. So how do we envision that this ontology can be used to solve the problems that we've been talking about? Well, the first problem we talked about was description. And we think that the ontology can help to clarify the meaning of terms. And that in itself can bring some interoperability. We also see the possibility that we can create tools that function like the Rosetta standard, where translation between standards are possible through going back to the core concepts and finding the points of comparison. And then we can have something that in this standard, the result would be this and in another standard, the result from the same data would be something else. And that can actually help people to navigate between these differences. And then also through making explicit the assumptions and having some common basis on which to express them, we think we can also solve the problem of aggregation. So, there's another aspect to this in that is there are emerging standards like the IWA building standards actively and we think our ontology can add some value to that. I have a diagram here of an ecological project and its properties from the IWA. And I want to draw your attention to the left, but on the left bottom there is a property called a modular benefit project. The idea is there that an ecological project can have some projects that create specific benefits. For example, one can create an emission reduction, another one can create an emission sink. It's even conceivable that they may be in different domains. Now, that sounds fine. But if you apply the ontology and say an agent engages in activity that impacts an environment, you come to the realization that you should not call something a modular benefit project. You maybe you should rather call it a project benefit. In other words, the activity leads to an impact in an environment. Well, if you if you think about it a little bit deeper, you shouldn't call it just a benefit because it can also be a negative impact. So, you can say it's a it's an impact of the project the project is the activity on aggregate level on the on the on the level of the individual activity. Then, what you what you realize is that, well, you can go even deeper and more generic. You know, it doesn't have to apply only to an ecological project any activity, whether that is business as usual for your company leads to impacts in your violence. So you can use the same structure. You can you can use the ontology to construct a generic structure that can be applied to project accounting like this, as well as corporate accounting, using the same framework that can be used for climate accounting but also for water impact also for biodiversity impact also for for any other impact in any other environment. And in fact, when you have that generic frame based on the realization that agents engage in activities that impact environment. And the accounting framework and tools that can be used across the mates, and that makes them worthwhile building. And that makes the impact that you can have through that process of standardization and the scope of the stands so much bigger. What do we want to do next, while we want to expand the the ontology to deeper levels, there are nuances to these things that we need to work out we are aware of them we have done some work in that regard, but there is quite a bit of expansion expansion needed. We want to embed this into the basic formal ontology and his extensions. And that is that process we think will add a lot of value but we are aware of that that is quite a bit of work. And then we want to develop tools to be able to use this. And for that we invite you to join us to participate to give us your your feedback your criticism, your, your, your, your suggestions. You can use the, the status working group wiki, there's the address there, as well as the hyperlegia climate action accounting mailing list. You can contact me at christian.po at Nova dot org dot zeta. These are the sources that we use the the interwork alliance orange ecological markets overview document, as well as the oboe foundry. And now this time for questions and comments. Thank you.