Skip to Content

Sustainability Videos & Lecture Series

Integrating Sustainability as a Key Driver within the U.S. EPA

The EPA has shifted from an agency focused on risk and hazard management, to one addressing sustainability challenges with regulatory responsibility. This talk, featuring a panel of biologists and researchers from the EPA, discusses the current strategic initiatives the organization is taking. The EPA's research focuses on evaluating scenarios comparing values, using story-telling approaches, and team strategies that are utilized as best practices when involved in sustainability issues.

Related Events: Integrating Sustainability as a Key Driver within the U.S. EPA


Kenneth Galluppi: My name is Ken Galluppi. I am recently here to ASU, back starting in April. The Decision Theater used to be part of GIOS, till my very first day. I was told that it no longer belonged to GIOS. I was moved over to OKED, which I responded and went to OKED. I’ve been figuring it out ever since. I am an environmental scientist, spent 25 years at North Carolina, 15 at UNC before coming here; worked a lot with the EPA Research Laboratories in Research Triangle Park, which is right down the street. I was with the National Exposure Research lab doing a lot of work with those folks. The group that are with me today were the sister laboratory, the National Health Effects and Exposure Research laboratory. I always get that kind of backwards.

They’re looking as an agency, for what does it mean to move into sustainability as a framework. I was back visiting in North Caroline a month or so ago and met with the speakers today. It was quite interesting. I think there’s a lot of overlap to a lot of the thinking that goes on here at ASU, so I was just able to get them here to—they’re doing fact-finding. They would love to get in discussions with some of you either today, or exchange cards in the future, either through video conference, phone calls, or whatever it may be.

I think what you’re going to find interesting is where the agency is trying to move to, and really through a partnership, I think we can really be a lot of help. I’d like to introduce the three speakers—four speakers. One is on the phone, Laura Jackson. The main speaker here in the room is Bill LeFew. He’s a mathematician, who is I guess, the team lead for this pilot program that they’re involved with. Rory Connolly is a toxicologist with the laboratory. Stephen Edwards is a systems biologist by training, but is doing more of the OMICS type research in the large databases. They’ll explain what they do here shortly. Laura Jackson is on the phone. With no more introduction from me, let me turn it over to Bill LeFew to kind of walk you through. I guess Laura is going to speak as part.

Bill LeFew: She will

Kenneth Galluppi: Okay, great, thank you.

Bill LeFew: Great. Thanks very much, Ken, and we’re very grateful that you were able to fit us in the schedules. Let me welcome you back from fall break. I recall in my graduate student days, fall break was something that I often missed and would sometimes even go to class and try to teach and there were no students there. We’re very happy to have you with us today. Also, thanks to the media and tech folks to kind of get all of this going. Laura Jackson’s on the phone. She’ll be speaking. She’s a world expert in one of the things that we’ll be talking about, and also wants to chat with you about that. One of the interesting things about this presentation is that—well, I’ll talk a little bit about what the EPA has done and is doing, and progressing as we continue to integrate sustainability across what we do.

When I get to the point to talk about the sort of things that we’re interested in doing on our project, the—what I would tell you next week is very different than what I’ll tell you today. We’ve been here a couple days now, and have had the opportunity to speak with a multitude of individuals. What we’ve learned is—has done a great deal to kind of transform some of the directions that we look forward to, to take in the future. Nonetheless, I’ll tell you what I would have told you three days ago, and then maybe we can discuss what we should do going forward. Sustainability has a long history.

In EPA though, as many of you know, we probably didn't always use the word sustainability, and maybe the eco folks were more adept at it than some of the human health folks. Very recently, the person who led ORD, Paul Anastas came and decided to start what he called the path forward. This was an initiative that centered around sustainability that was the true north, he said, to how one as an organization needed to move forward in the future, what the Environmental Protection Agency needed to really be about. Broadly, to my mind—and let me just emphasize something that is probably too small for those in the back to read.

This presentation—this is an abstract, a proposed presentation. This does not necessarily reflect EPA policy. The mention of trade names or commercial products does not constitute endorsement or recommendation for use. If there are members of the media here, or people who want to say the EPA said this, I said this, and EPA probably thinks something else officially. Nonetheless, I’m going to give you some intuition on the sort of work that we’re doing, and it’s our stuff.

Okay, so that said, we have this path forward. To do so, it was a great deal of organizational transformation, and one of the things that they did was restructured how we did research into six groups, which we’ll discuss briefly—research action plans, all of which centered around how our research integrated itself with sustainability.

One of the major initiatives was reaching out to the National Academies, and the National Research Council to ask them, “Well, we’ve been doing sustainability for a while in certain contexts, but we really need some information and some intuition on how to do it better, how to move forward, what it means in the context of a modern organization to do so.” This is exactly what they came up with, the so-called Green Book, Sustainability and the US EPA. Some of you have probably read it.

Some of you were probably involved in the making of it, and probably most are aware of it. I don’t want to do a slide karaoke, but you can see what they kind of said, and what they said about risk-based methods. They’ve had some success, but we want to challenge ourselves to go beyond risk management. They talk about the idea of sustainability assessment, and they walk through a program of how one might approach those problems.

I just wanted to mention that briefly as one of the things that EPA is looking at. While we haven’t—so here are the four questions that we posed to the National Academies. We said well, what sort of framework—how should we really be thinking about these problems? Where do we need to come from? What sort of expertise do we need to have? What sort of tools shall we, as an organization, be utilizing and building? What does it really mean to change from this sort of paradigm—there was a red book in 1983 that looked at risk assessment and risk management, and guided us strongly since then—to this sort of sustainability framework, which is really what this—what the green book is aiming to do.

I think the National Research Council answered very strongly. They answered very directly, and they came out with some ways in which we want to look at it. They said, “Well, there’s a bunch of different ways you can go about it, but let us just show you a certain framework.” This is—I won’t look at this picture. I’m going to break it into two different sections. As you can see, there’s a level one and a level two, and this center box, sustainability assessment and management, is where they spent most of the time looking at it. That’s really what they wanted to think about. This level one was just as important. We start on the back side. This is completely unreadable, but it’s a triple value thing. It’s economics, environment and social, okay.

For us, social kind of means in the context of human health, but there are, of course, other social issues with which we want to be involved. Those, combined with EPA sustainability principles, and whatever the law and legal mandates tell us, and that—you can just throw Congress among other things right there—that goes into what we constitute as our EPA sustainability vision. These three things are what’s going to drive whatever we do in the context going forward. From there—from the vision, we want to generate goals indicated in metrics, and metrics, as I’m sure everyone in the room realizes is a challenging thing because we have to figure out how one values and how one discusses, and what sort of people to bring to the table, so we have the metrics that we care about. Organization and culture is also something that’s very, very important.

If we have a risk-based metric and a risk-based paradigm, how do we convince and persuade and involve the sorts of science that’s generated from that? How do we make that into a culture that supports these sorts of sustainability assessments and management? Finally reach here, and as you all know, it’s a loop. It’s something that goes back and forth. We have periodic evaluation, public reporting, but anytime you have it in science, even from the modeling perspective, you have experiments and you have model, and you go back and forth and try to iterate and understand how the system really works.

On level two, and this again is dropping down from sustainability assessment and management, we have kind of very particular ways. This was their suggestion. I think for me, this was their entry into saying, “Well, how do we present it to the folks in ORD who do science? How do you apply your science to various points along this process? Where do you fall?” They say, “Well, we have to understand the decision.” We do some sort of sustainability screening and evaluation, which mimics something which we might do in risk assessment. Then you just go down in a very natural way. They have scoping and options identification, scoping application, trade-offs, results and decision makers, decision taken and implemented, and then evaluation and we move back.

This is sort of thing that the National Academies threw out to us and said, “You really need—this is the way—one of the ways in which you could think about this sort of problem.” While EPA hasn’t gone and adopted this officially yet, we’ve—each and every one of us has been instructed to go and read it, and see how our science applies to it, and understand exactly what’s going on. As the result, we developed the four ORD program areas, and there are two additional RAPs—research action plans—which come out of this, but these are the ones I’ll mention today: ACE, ERA, climate, and energy; CSS, which is chemicals, safety for sustainability—they change the S’s occasionally on me; and sustainable health communities; and sustainable water resources. Got them, all right.

We developed these four program areas, which is where each and every one of our research programs lies, and so essentially, all science is done in ORD right now. It falls in one of these programs, and it’s done in a new matrix management infrastructure, so we have folks who control money and folks who control people. They’re trying to interact and figure out how this research gets done. It’s interesting to say the least, how it’s happening over the next two year.

Finally, this is the challenge. We’re looking to identify sustainability metrics and indicators, which actually integrate human health because this is something, on the eco side; they have kind of a very nice intuition for, and have the long history of. On human health, we’ve been challenged a bit. This was recommendation 4.1 in the sustainability toolbox. They mentioned some other things, which I know folks around here are very involved with.

I know some folks here do life cycle assessment, benefit cost analysis. They went through each one of these and said, “You should think about applying these tools that the EPA does,” and in a broader sense, to chemicals and human health, and involving this sort of triple value idea. Each one of these was discussed rather thoroughly. Now I’m an applied mathematician, and so for me, I said, “Okay, well, if you want to do a systems model, what does it really mean in the context of sustainability?” Dr. Joseph Fiksel, who many of you are probably aware, said, “Well, let’s look at this triple value sense.”

We have environmental models, human health models, economic models, all of which are, from a systems model perspective, integrated and involved with each other. You have kind of a tiered approach, but he’ll say, very directly, there’s no single model that can address all the needs of decision makers and stakeholders with multiple skills. You kind of have to figure out how each one of these balances with each other and moves up, in a systems model sense, and then still informs the sorts of decisions makers and experts in each of the individual fields. That was kind of the philosophical basis for the project that we’re working on, and how we started. The context of bridging science policy and human values, there’s a kind of a thought process that I think is very difficult to convey to folks who don’t work in this sort of thing every day.

This center bubble, how should we proceed, given the uncertainties and ambiguities? Even the concept of uncertainty, I think doesn't necessarily transfer well to public interaction. If you say something is uncertain, and this is something that Ken’s mentioned over and over again, if you say something is uncertain, people may say, “Wow, why are you talking about something that’s uncertain? Give me the definitive, give me the answer, you know, what’s going on.” If you’re a scientist, you say, “Well, this is just—I’m trying to tell you where the real stuff is. It’s in this range and there’s an uncertainty, but that’s where it is.”

You’re trying to make a statement about certainty in your bounds involving it, where those people may be hearing, “They don’t know what they’re talking about.” How do you progress from a very complex thing? Beginning, what do we know today, and what are the unknowns? Well, systems involves very, very complex things, lots of unknowns, lot of things you know, but how one integrates them, it’s a very, very challenging thing. Finally, something that you might want to ask at the beginning, to get that base line agreement between the parties, what do you care about the most? Again, I’m a mathematician. All models are wrong, but some models are useful. This—you know, one of my personal favorite tongue-in-cheek statements, but—and useful for the modeling crowd, and again, one that can be misinterpreted by folks who don’t work with models every day.

If I start by saying all models are wrong, I’m giving the impression that whatever model I give you, I kind of don’t know what I’m talking about. Whereas, when I’m talking to modelers, the point is very clear, and says, “Well, this doesn't cover every last little thing you could possibly put in it, but maybe there’s something useful that I can extract.” I do some mathematics and I extract something back that, in a non-mathematical context, make sense, gives intuition, and pushes the science and intuition forward. Even that simple statement really conveys to me the sort of challenges that are involved when looking at sustainability problems. This is—for our team, this might be the picture that started it all.

Roy Conolly was involved in a white paper that was looking at the triple value model, and how we might integrate it across EPA. This is one of the graphics they produced. Fiksel was, of course, very much involved in this. He said, “Well, this is very interesting, and it tells a very nice story, but maybe we should consider what we can do computationally to tell a similar story, and really integrate some of the computational modeling sciences that we do in the EPA for human health, and integrate them into this sort of structure, and see what stories we can tell with that. This is where we kind of begin.

In the classical version of this, it just says community, and maybe social capital underneath, but human health, for our laboratories is one of the key elements, and so one which we look to integrated. I think everyone’s probably been in this circumstance, where you go out and you do research, and you say, “Well, let’s just—you know, we’re going to keep learning. Let’s go out and gather some data. Let’s look around and see what people are doing.” Maybe in the back or forefront of your head, we say, “Well, we’ll just—we’ll think about use for it later on.” This is one of the things that—it’s a good and bad thing, but it also comes back to the idea of producing insights versus producing answers.

I think one of the key thought shifts that we have to make as scientists looking at sustainability at EPA, is going from—I’m going to give you the answer, then I’m going to give you a toxicity value or a number that I’m going to shoot out. Those may not be the most useful thing. I instead want to give you scenarios and stories that tell you, if you apply this value system, maybe this is how your system evolves. Maybe this is how your system changes, and work from that. Work from a place of value systems, and give intuition as to what might happen, rather than giving you a particular number, answer or value. To do so, we would like to utilize some of the information that we actually have and think of a use for it now.

On to the project that we’re actually interested in. I’m going to talk through three tools, though actually Laura Jackson is going to talk to this EPA Enviro Atlas, and we’re just going to chat about how we might go and construct a computational tool that integrates some things, and tells a story to people, and hopefully gives the ability for multiple people of different disciplines to be brought to the table in an integrated way, such that we can all talk about a very complex problem. The first bit involves this EPA Enviro Atlas. As I click over and happily click through slides as Laura tells me, she’ll start speaking.

Laura Jackson: Hi, yes, this is Laura Jackson, and I’m the ecologist on our little team. I’m a landscape ecologist and do a lot of research right in my desk, unfortunately not doing field work anymore, but for many years, looking at satellite imagery, aerial photography, and indicators of environmental conditions, from more of a broad scale standpoint. We are—I and a very large team across ORD—are engaged in a really exciting tool development effort that we are now calling the Enviro Atlas. It started out being an atlas of ecosystem services because that was our last research program before we had a charge to expand to sustainability, which includes ecosystem services and many other things.

We are—we’ve got most of our efforts and most of our progress on looking at how to quantify the provision of nature’s benefits across the United States, and in selected communities, which are—that effort is discussed in this slide. The community component of our Enviro Atlas—it is a national product, so it’s at multiple scales. The community component is—we are looking at a selection of cities, based on various characteristics. Hopefully, we can transfer what we learn about ecosystem services and other urban aspects of community assets, so green and grey infrastructure for example, the livability of places, which includes the natural environment, and aspects of the built environment that contribute to health and well-being.

We have selected s few places to be representative pilot communities. We’re working with a lot of other partners, including extensively the Forest Service, that is helping us take land cover imagery that we’re classifying at one meter resolution, so we have very fine level land cover classification for our cities. We’re using tools such as the Forest Service’s Eye-Tree tool to translate those land cover classes into certain kinds of benefits that land cover has been modeled to provide, such as air filtration services, water storage, shading, and cooling effects of tree cover, and other things that you see here in the top part of this slide, and where also, we have one model that the EPA Office of Air has developed and used for regulatory purposes, to translate the benefits of this urban air filtration of trees into actual health benefits of health conditions averted because of ozone and particulate matter and other things captured by the urban vegetation, like days that are not missed from work or school, and also translation into dollar values of those costs, adverse conditions averted.

In addition to using existing tools, we’re also developing in-house, other kinds of indicators. We’re looking at trees and their potential to buffer roadway air pollution. We’re using—we have an extensive literature review. We are doing sort of meta-analyses to try to find if there’s anything quantitative we can say about those physical and mental health benefits of viewing nature, say through a window, depending on the design of the built environment, if there’s visual access to green space from homes and schools, and physical access to green space, both for recreational purposes, physical activity; also for a place for neighbors and community members to meet and relax, which provides a lot of individual and community benefits, and other benefits that we are drawing out of this very growing field across a lot of different disciplines about the—how nature actually increases longevity and reduces adverse birth effects, and improves school performance, and many different kinds of things of that nature.

We are—we have an environment justice focus, where the agency now, since the new Obama administration has come in, is paying a lot more attention to trying to improve inequities in environmental contamination. We look at ecosystem services also as—can be a form of inequity, where low-income and overburdened populations often have a deficit of natural infrastructure that can help them buffer pollutants, buffer natural hazards, and also serve a health promotional capacity in communities.

This is all at a very high resolution, as I said, with summarizing at the census block group scale, but we—the data, in many cases, go down to one meter. Just as an example, in the next slide, we take this aerial photography from the USDA, very current day, and we characterize the—from six to ten different kind of land covers. This is just an example of tree cover. We have it at the spatially explicit scale, and then we summarize at the block group scale. This is for Durham, North Carolina, which is our first pilot city. We don’t really know how much tree cover is enough, so there’s plenty of unknowns here in this research.

Getting to thresholds is probably going to be a long-term research issue, you know, how much should you be looking out a window to improve your chances of a health newborn? Well, that’s—probably not in my lifetime will we know the answer to that question, but we can do things like compare neighborhoods, and see at least who seems to be particularly impoverished when it comes to tree cover. We do know for some of the Eye-Tree metrics, exactly how much runoff is prevented with tree cover, given impervious surface in an area, how much the ambient temperature can be reduced and things like that. We of from very quantitative to some pretty qualitative assessments at this point with these indicators.

Now the next slide shows our work on the near road, where we take tree buffers alongside major roadways, and we make a lot of estimate about you know, maybe ten meters is good, given some field work that we’ve done here. We map what parts of the city have what we are estimating as being a sufficient roadway buffer to reduce air pollution, particulate matter, for example, coming from the roadway. You could use this to highlight areas where, if there’s a lot of population living alongside, say a red colored roadway, that might be an area where mitigation could take place with tree planting if possible, or maybe you’d want to keep a new school out of those kinds of areas.

There are local uses for this kind of information, even though we haven’t nailed down exactly how much trees is going to be enough to mitigate roadway air pollution. We’re trying to show people there’s a suite of benefits that our green infrastructure provides, so not just mitigating air pollution, but all the many things, so that you get a lot of benefit for the effort that you put into carefully designing how you can have nature’s infrastructure work for a community and for their health. The next slide shows what we’re doing with census data. This is something that Oakridge National Laboratory actually pioneered, and they call it dasymmetric mapping, really very simple.

You make some rules about where you don’t think people are likely to be. You take the census data at the finest resolution available, which is the block level, and then you figure well, the people aren’t going to be in agricultural fields, you know, if you’re working in a rural area. They’re not going to be in water. They’re not going to be on steep slopes. That way, you get a better idea of how those people are really distributed within a census block, which helps us make some assumptions about how many people are within a 300-meter buffer of the road for example, which is determined to be the high exposure zone for near roadway populations. The Enviro Atlas is providing the metrics for the ecological part of this triple value tool that we are developing in our project. Phoenix, by the way, slide 18, is one of our six pilot cities, and this is very stretched out. I apologize. I told Bill to put it back to the right scale.

Bill LeFew: That’s my fault, sorry.

Laura Jackson: This is hard to recognize, but hopefully the labels will convince you that indeed, this is Phoenix. We are really looking forward to working with the Phoenix urban LTR site, and hopefully additional folks locally at ASU and in the municipal government, to help us particularly with social data, with health data, that we can—as we’re doing in our other pilot cities, trying to validate some of those relationships that are showing up in the literature, and CFA really do hold in different places, different kinds of populations, so that maybe we can develop some models to say like well, you know, this much shade does seem to—you can sort of assume that there will be X number of heat-related deaths averted, you know, given high temperature days, as an example.

We are hoping to do basic research, as well as provide, hopefully, useful information to improve decision making in our pilot cities and elsewhere with similar conditions. Now the slide 19 is—features the literature review in a very user-friendly way, so think Bill was saying he wanted to talk about, so I can turn it back over to Bill. I just want to say that this is one of our roots of getting at the health component of our trio tool, using available knowledge. Bill, did you want to take it from here?

Bill LeFew: Yeah, absolutely. Thank you very much, Laura. Laura will be available for further questions on that tool, after the talk of course. This is actually my absolute favorite tool that we’re looking to link in our search for computational sustainability things, and also brings us to the audience participation part of the show. When I click on this, actually, I’m going to bring up the website, and maybe I can get Steve to come over and click through and off. I’ll point on the screen. This is active. This is live. Hopefully, it will come up right there and easily. There it is.

What you can do with this tool is you can actually choose something to click on, and it will delve deeper into the relationships that might extend from it, either—and you can see these arrows are directional, so it’s going, too. These places will actually have information. What sort of struck me so strongly about this initially is, as a mathematician, I really—I don’t know basically anything about any of these things. Nonetheless, when you have urban systems here, I can immediately take in the information that heat mitigation must be related to them, promotional physical activity must related to it, water filtration, air filtration, all these sorts of things, and it very intuitively and easily tells me a story about what’s going on.

I think those are the sorts of tools that we certainly want to integrate into whatever computational stuff that we do, such that we can reach out to people who might not be so comfortable with say, differential equations, or very complicated graphs, or 3D immersive exercises or whatever. This sort of thing, I think is very attractable for a wide variety of people. I said it was audience participation. Which one would you like to go into first? You, Ma’am. Air filtration, that’s my favorite one. Yep, we’re going to come back to that later, so I’m very happy about it. You can see air filtration, a whole bunch of things just pop up, which is—it’s kind of remarkable isn’t it? Where do you want to go?

Audience Member 1: Oh, another place?

Bill LeFew: Yeah.

Audience Member 1: Oh, water.

Bill LeFew: Okay, before we go to water ,and so where—

Audience Member 1: You mean one of those in there.

Bill LeFew: Yes.

Audience Member 1: Oh.

Bill LeFew: Forests, okay, so let’s click on the plus between forests first, Steve. You can just click on air filtration and you can go right back. The plus—and this is directional—so it gives you a description of kind of what we pull from our database, and so here it says, for people in the back who can’t see, forested environments are an important process in natural air filtration. Forests have been shown to produce ambient concentrations of the harmful air particles, PM10 and PM2.5, as well as carbon monoxide, sulfur dioxide, nitrogen dioxide and ozone. Then it gives the reference for what’s going on. One study found that air pollutant removal—found this air pollutant removal to result in an average monthly air quality improvement of 1.5 percent. Reference, the efficiency of trees to remove air pollutants varies by species. Conifer dominated forests have been found especially efficient—to be especially efficient in capturing significant quantities, references, etc. Yes, Sir?

Audience Member 2: Happiness.

Bill LeFew: Okay, we’re going to go to happiness, and so that will be next. Let me just point out one of the things that I like very much about this is that it has some things that I think you could pull into models and data-driven computational things, along with the references, and it states it in a very concise and nice way, that you will gain some information. Let’s go to happiness. Air filtration and happiness, evidently, on a scale of one to four, happiness is increased in Belgium by .0—this must be real data, right—by .043, when nitrogen and lead air pollution decreased by 14 percent and 40 percent respectively.

Let’s see, Steve, can we scroll down and see if there’s a reference or a joke attached at the end. There we go, 2006 study in Belgium, Denmark, France, Germany, Greece, Luxembourg, Netherlands, Portugal, Spain and the UK. There you have happiness and air filtration, something that I would have never expected to be related to each other, but there you go. This is very nice, I think, that you can kind of pull up this sort of tool and walk through it, and have a story. You could imagine that if I had maybe a plot over here, that I might be able to select air filtration and happiness, and air filtration and forests, and start building up my own personal tree. From those trees and branches, build up a model, in which you might be able to associate these various things in a very reasonable way.

Let’s just go into happiness, and see what else is connected to happiness, since that’s probably a fan favorite. Engagement with nature and promotion of physical activity, both of them we have, and we see nods across the audience for that. This is my favorite tool. This is what I really look for to passing data and information through it ,and I think this is a very reasonable way to go and tell a story to a broader audience. Steve, let’s go back to the presentation, unless anyone wants to walk through, a burning desire. Yes, Sir?

Audience Member 3: How do you find that?

Bill LeFew: Yes, so I’m happy to distribute the presentation, and there’s a link at the bottom, but if we go back to the slide, the web address will be at the bottom of that slide, and there it is. It’s at Okay, and so I’m so glad that you chose air filtration because I also chose air filtration, as we were looking between urban ecosystems and air filtration, and just noted again that there’s some information in there that you could actually use to build some sort of model that would say something useful.

As an applied mathematician, I just said, “Well, okay, what would I extract from that?” Here, I would say the trees are related to ozone, and it eventually pushes it to zero if you have enough vegetative coverage, and it tells you something about the rate, similar things about vegetative hedge, green spaces. Evidently 2000 square meters of uncut grass takes a -4000 kilograms of PM on some sort of time scale, which wasn’t references in the short bit of information, but there’s a paper associated with it. These are the sorts of things that I would extract just from that edge of the tree. We go deeper. I didn't choose happiness, I chose hospital admissions.

Once you go from air filtration to hospital admissions, you also get a great deal of information which you could then go and look and say, “Well, what sort of information might we want to plug into a model?” Increase in PM, sulfur dioxide, increase in admissions. That’s pretty intuitive, but there’s something about the rates that are associated with them. Rise in smoke, rise in SO2, all these things—I didn't know anything about London fog, but evidently five to nineteen times more London fog than regulatory standards gives hospital admissions for respiratory disease by increases of large percentages.

These are the sorts of things that you could contain in a relatively straightforward and simple interface, and then extract into models that would then give the persons interested in those trees some further data, some further information. Again, I’m not really concerned with giving them answers, but I’m concerned in looking at the scenarios those folks are interested in, and so that’s kind of a natural way to do so. At the end, if you go to hospital admissions, we see two of the things that are connected, and maybe if we were building that sort of side tree that I was talking about, they might recommend that you also include these, if your endpoint was hospital admissions.

This brings me to the tool, so okay, fine. We have all this stuff. What sort of tool might you use to integrate this knowledge and do simulations? Well, there’s a bunch of them out there. The one I just chose to look at today is Vensim. Some people in the room have probably used it. Some people are probably experts in this sort of thing. Really, it’s something that’s not so different than if you’ve used SimuLink and Metlab, or you know, any sort of variety of things, where you have stocks and flows and information and differential equations that are actually covered up by boxes and arrows, that sort of thing.

This is an example from this website, but I just chose it because it has stocks and flows, and you can kind of see how things might be related. Nonetheless, I don’t think this necessarily tells a story to a non-engineer, a non-technical person, someone in particular, if he just hasn’t used Vensim, or anything like it a lot. You might want to use a wrapper. This is not also what I think you should use, but this is a more complicated system. Narragansett Bay Pilot, Joseph Fiksel and his team looked at this, and this is coming out of the same triple value philosophy that he has before.

I only throw this out because it’s kind of big and scary because the next slide I’m going to show you is something that’s nice and less scary. You may take these and all these relations, and you could do this in Vensim. In fact, they did. They make a nice little wrapper on it. They say, “Well, here, how about this tool?” You have economy, society, environment, and you have these little tabs of different plots that you might look at. If you had something as cool as a decision feeder, where you could throw it on seven different screens, you might have those little plots all over the place, but here, they’re just tabs. On the left, you have the ability to just slide things along.

Look what happens if heavy rainfall starts happening, how those systems—how the watersheds interact with each other. What happens when a wastewater treatment gets—you buff it up? I think it allows you to tell a kind of a nice story. What I’d really want to do, so this is kind of my current goal on the technical side of the project, is to integrate the type of information that you find in the Enviro Atlas, put in this sort of storytelling framework that you see in the Eco-Health Browser, and then be able to output it to the sort of computational, integrative tool that you see in front of you there. That’s kind of the—that’s kind of the idea, and that’s what I would have told you three days ago.

Today, that’s actually still true. I still think this is very interesting. We’ve heard a great deal about the component that’s been neglected, that is bring the sort of stakeholders and interested parties to the table beforehand, so you can have a common basis on which to talk. If I come to my interpretation—and Ken can correct me after the presentation, if I haven’t really heard what he’s telling me—is that if I just come to the folks who are concerned about this, and I put this on the table in front of them, we haven’t built the sort of trust that’s necessary in order to move forward with a complicated problem. I haven’t really given them a great reason to both believe me, and both believe the process which brought us to this point.

That’s the sort of thing we’re going to work on, I think when we get back to RTP, but the technical side will still be there. To do what we want to do with the previous example, we’d just adjust the Narragansett tool, and so maybe for instance, we would look at PM and ozone setups, instead of what we had there. The economy tab is a little light right now, but we’ll certainly get on that, and you know, environment tab is very natural. You’d have something like air filtration versus trees, air filtration versus vegetative cover, air filtration versus grass. You would imagine those to be—in this environment, there would just be tabs here, and you would follow the plots along. There’s more. You can always do I think a little better.

One of the things that we certainly want to plug in is natural language references. That is, while it’s interesting to look at—backing up twice—interesting to look at these sorts of plots, I have no base expectation that someone is going to intuit what I want them to take from the plot by just looking at it. I want to be able to have this plot, to put something that’s verbal and—or just natural language English, and they can—it’ll just tell them what I want the plot to say. This will say between these years—if they click on something between these years, while there’s a variation between this—this value and this value, it’s raising, it’s lowering, you know, some sort of story, what information I want to convey.

I don’t necessarily want to trust the graph to actually say it itself. I’m a mathematician, so I’m—I like dynamical analysis, and so we’re interested in resilience and tipping points. We want to be able to extract those from the sorts of systems that might occur, and make sure that those are brought front and center, and general stuff. I mentioned the addition of economic things to the eco health relationship browser, and sub-diagrams, that idea of being able to build your own separate tree as you’re trying to construct the model. Of course, we have the simple interface, simple save and load, and some other details that are associated with making a good computer package.

What brought us here is the search for best practices, and so this is just kind of a quick thing. We’re looking for the best visualization tools, the best sort of value system friendly interfaces. Again, we want to be able to extract information for people that we want to work with in such a way that we can produce a computational product, having things that are useful to them, and finally, this idea of technical versus useful balance. You can do some very, very cool things technically, that you may not be able to express in a very useful way, and so maintaining that balance as you’re approaching the problems of sustainability in a computational way is something that we’re still wrestling with.

This is one of my favorite slides. This is the team which I’m very honored to work with and be a part of. This over here is actually the beginning of kind of how we got this idea on the computational modeling. It’s an example case study involving water bottles, city water and money, that I actually came up with on my honeymoon. We were in the Greek Isles, and one of the things that I noticed was that on some of the islands, the water wasn’t potable. You could just get bottled everywhere, and I wondered to myself, and then to my wife, and she wasn’t as interested in this—but also, what would you have to do in order to generate an infrastructure, and what would be the costs and those sorts of things.

When I got back, I presented this to the team, and kind of gave them the sense of what we were talking about in a computational sense when you’re involved in all of these things. I have fond memories of both this and the honeymoon, and so I thought I’d put that up at the end. At this point, we’d be very pleased to discuss, answer any questions that you have, and I thank Laura on the phone, too, and I’m sure she’d be willing to answer as well. Yes?

Audience Member 4: Laura was discussing how she was planting these trees along the road. I was wondering if they could also serve as corridors for habitat fragmentation as well. I was just wondering how wide these tree-lined streets were.

Bill LeFew: Laura?

Laura Jackson: Yeah, exactly what we tell people, that there’s a suite of benefits that it can provide, wildlife corridors, improve property values, could be an opportunity for a greenway. Then if you’re being completely honest, you have to think about the dis-benefits. Maybe those would be really scraggly corridors and you would just get invasive species. It could be a way of bringing fire into town, or maybe criminal activity, so you know, really trying to look at the pluses and minuses, and see, again, what the bottom line is, with investing efforts into planting trees, maybe primarily to protect near-road air pollution exposure. You know, you hopefully would be getting additional benefits, too, but possibly it would be a bad idea if you really looked at the full picture.

Bill LeFew: Yeah, please. Steve, Rory, you want to join me up here? Yes, Sir?

Audience Member 5: On the EPA heat island website—there’s a heat island website underneath EPA, and that’s what they call the MIST—

Bill LeFew: MIST tool.

Audience Member 5: MIST tool, mitigating against the urban heat island, and it considers whole city scales, but now I’m hearing that it sounds like you’re doing research, or about to, at very fine detail across the whole country. You said 25 to 250 cities or something like that?

Bill LeFew: Laura?

Laura Jackson: Yeah, in fact now, while I’m participating in this call and looking at some results from some Durham high school students—we had folks go out in the field and measure the ambient temperature and the surface temperature of dark pavement for example, under trees, and out in the open on a sunny day versus light colored pavement, versus grass. I’m actually look at a [inaudible 00:42:56] right now of what the differences are. We are hoping to use this as sort of a ground truthing from the Eye-Tree data. We do have a mismatch of scales though. We have people going out in the field and measuring points on the ground.

The Eye-Tree data is derived from field work as well, but it’s summarized at a whole block group level. We might see a difference of like ten degrees between ambient temperature, say on—above a dark impervious surface out in the sunlight, and from a grassy area out in the sunlight. That might be at a single point, but when you summarize how much the trees are doing across the whole block group, depending on how many trees there are, you know, maybe it would just be like three degrees or something. It is certainly dependent upon scale, and you can have buildings—you can have energy savings from the placement of trees on the south side of a building for example. That would be much more beneficial than some other placement of trees.

I’m not familiar—I mean I know the heat island website on EPA’s webpage, and we have directed students to it, but I’m not a part of the projects that are described or the information that was developed for EPA’s site in general. We’re doing something—this is fairly small, considering what all may be going on at EPA. Yeah, this is going to be at a finer scale for the community component of the atlas.

Bill LeFew: Yes, please.

Audience Member 6: I have a kind of bizarre conceptual question, but bear with me. It strikes me—I mean your sort of comparison of red book to green book, the shifting of—this shifting of orientation from getting that metric, getting the number, getting the relevant knowledge and then applying it to judgments about what we should do in the world, to asking first what do we care about and what should we do in the world as the basis for metrics. Sorry. It strikes me as an interesting and important shift.

The red book was in a lot of respects, very explicit about the dimensions of judgment that were entailed in the processes of risk assessment and management, and was quite concerned to make space for sufficient—I don't know what you want to call it, institutional heterogeneity, such that robust deliberation would happen. When knowledge was made and got stabilized and was implemented, it was processed sufficiently that in effect, values were present—tacitly present in it. I was, you might say, politically robust, not in a kind of cynical sense, but in a serious sense. It responded—it tried to respond to the needs of the world as best as possible, in the work of knowledge-making.

This reversal, in a sense—so my question is, how are you—what are the mechanisms for approaching the assessment of relevant values in a similarly robust way. The idea that you would bring the relevant stakeholders to the table to say what you care about in the construction of the modeling enterprise, thank you very much. We made a list, we’ll incorporate this as best as we can, and then we’ll build very baroque and clearly very useful structure that you were showing to us. It stabilizes those values in a sense, and it gives a kind of characterization of the notion that stakeholders—relevant stakeholders should be self-evident up front and won’t change in a sense.

We know what we care about, and we can act on what we care about in the construction of these. There isn’t—I didn't hear a mechanism for iterated dimension of the what we care about question. I’m wondering whether an attention to that kind of politics fits in here, and if so, how. Moving—the parsing of people, the parsing of health and well-being as a kind of modeling problem is simultaneously a sort of problem of asking if we’ve kind of got it right in terms of what we care about, in terms of what we’re interrogating, and sort of abstracting from the level of life as lived. I’m wondering what’s the—where’s the space for this kind of thing in this shift towards sustainability?

Rory Conolly: Yeah, sure, right. I could just say yes or no to your question, I guess. If I understand what you were saying, I think it could be restated maybe as different people would have different ideas of what a sustainable solution might be for any given issue in our world, and I mean clearly, that could be across a political dimension, or probably any dimension you want to imagine.

We’re here because we have a small internal project that the agency funded for a year, for us to think about the sustainability issues, and to develop—hopefully to develop a computational tool that—at least as I understand it—would address the issue you’re raising, not by indentifying any particular solution to a sustainability issue, but by facilitating the evaluation of options in the process of decision making, so that in other words, different people have different values or ideas about the value of some parameter in a model.

You would have a slider possibly on a webpage, and you could slide the slider and that would reflect different weightings of the value for a certain parameter. You could explore how changing the value of this parameter would influence some metric of sustainability downstream in the model, okay. Different individuals could all play with the model and come up with their own optimal solution very quickly because this computational tool would facilitate their doing that. What we’d come up with wouldn’t be a solution to a problem, but maybe a suite of solutions that would reflect different social, political values that individuals would have. That’s kind of how I envision the project that Lauren and Bill, and Steve and I, and some others are working on. Does that address your question a little bit?

Bill LeFew: Let me also follow up with that. On the structural problem-solving front, you know, the idea of what sort of robust framework does a sustainability assessment take, relative to the conventional robust framework of a risk assessment. In the green book, they have a pretty detailed discussion with that, and I wish I actually had a nice PDF copy because they have a nice chart that compares what they mean by a sustainability assessment, in a kind of piece-by-piece manner, to what they say when they say risk assessment. Do you recall that chart? I think they’re struggling with the same sort of question that you’re concerned with.

In the red book, there was a very robust framework for risk assessment. I think one of the reasons—and this I my own personal speculation—that EPA reached out to the National Academies, was that they were concerned with that very question. If we’re going to go to a sustainability paradigm, how do we have the same sort of robust assessment. I think that’s something that they’re struggling with at very high levels of EPA administration right now, and might be a potential reason that the green book hasn’t yet been adopted wholeheartedly yet because they’re struggling with exactly the sort of framework that they want to implement, such that the decision making process is a robust one.

Audience Member 7: You were referring to tipping points. Where do you feel like we stand on those at this present time?

Bill LeFew: Well, I guess—when you ask me that, I want to answer from a mathematical perspective, in which case I would say I’m for them, and I’m very interested in how they occur. Where I think we stand on them is—you need a—I guess I don’t know—so can you expand on your question a bit?

Audience Member 7: Well, I mean how far along are we progressing towards it? Some people say we’ve already hit the tipping point, some people say we haven’t.

Bill LeFew: Relative to?

Audience Member 7: Well, you could talk about greenhouse gas emissions for example, or whatever.

Bill LeFew: Got you, so the tipping points that we were talking about ,or at least that I was referencing, were much smaller tipping points. They were saying things like if you have an aquifer, how much can you suck from it before it does not—it won’t sustain the population over time, you know, maximal bounds on parameters, such that a behavior changes in the system. Rather than global, is the earth, you know, toast, sort of tipping points.

Steve Edwards: Yeah, so the project is more about how we approach the problem rather than solving the problem. While we’re looking to build a solution that could go to a global scale, you obviously can’t start with that, and so one of the reasons he showed Durham is that we’re working—there was this ongoing pilot that Laura Jackson and Rochelle Araujo, who was also on the team list, were already involved in. We’re trying to piggyback on that. It’s really an urban redevelopment type issue, which is much more tractable and kind of we can use a tractable solution to build the tool, and then see can the tool kind of expand out to some of these broader issues that, even if we got a solution, nobody’s going to agree that we got the right now.

Bill LeFew: Let me give you a particular example though, just very briefly, using the air filtration that we have up there. For instance, let’s take—let’s think about air filtration and use the Enviro Atlas to look at how the trees are integrating with how we place them spatially around the city ,and the PM readings that we’re getting. You can further integrate that with toxicity data and know how much PM it takes to actually kill a tree. There’s certainly a—there may be a threshold, at which it doesn't make actually sense to throw trees in as filtration devices because you’re just going to kill them.

That would be the sort of tipping point that we want our model to be able to integrate and to be concerned about. As you imagine going through that Eco-Health Browser, you might be able to click on several notes and construct an integrated model that will actually produce that sort of behavior, and inform you as to it, in a way that might be counterintuitive, since you may be coming with the perspective that, well, if we have so much PM, well, we just need to put more trees there. Then you’ve taken the tox data, and say well, your trees are going to die. That would be a tipping point which you would want to be able to display out of the model, and extract them and represent in a way in which someone who is concerned with those kinds of issues would get the point.

Audience Member 8: Phoenix is one of the six pilots. How much information is there on Phoenix, and how could we access that? Where are you at?

Bill LeFew: Laura, would you like to talk about your wish.

Laura Jackson: Yes, you’re saying how much information is currently available for Phoenix?

Audience Member 8: Yes. We have made—we’ve reached out to the Phoenix LTR—urban LTR site PI. The first one is—I think it’s Xiao Xiao is her first name. I’d have to look in my email to get her last name, but she’s a remote sensing technician. She has been working with her colleagues to do one-meter land cover classification for the greater Phoenix area, and I think we’re going to be able to use those data in exchange for delivering to their team all of our metrics. We’re just really happy because Phoenix is huge, and we’ve discovered so far, in working in Durham, and in Portland, Maine, and then Tampa, that Tampa’s huge.

As soon as we get into larger cities we are just really done for because it’s so much work. We have access to data that you probably have access to also, and that’s this one-meter land cover data from the late ‘00s, I imagine is the timeframe for that. From that, you know, we’ll be looking into various things that we should, consistently from city to city. We’re just beginning in trying to make contacts in Phoenix that we hope that we can make some contacts. The city has a Public Health Department for example, or perhaps we would work with the counties involved, so that we can hopefully do some modeling, as I say, to validate some of the connections, some of the associations that the browser displays from the literature.

We’re very early on is the short answer. There’s just various things that come up when you talk about this out west. There’s issues—it’s hard to distinguish rooftops from bare dirt. I don’t know how much trees are really an issue in Phoenix. I’m from the south. I don’t know the west very well. I imagine Phoenix is fairly planted up in town, but there’s going to be probably other ecosystem services that we’re going to looking at, I would imagine, more heavily than tree cover if it’s more of a desert environment. That will be the first place that we go to that’s not pretty green. It’ll be sort of a new—it’ll be a new environment for us. I’m rambling, but we are just beginning.

Steve Edwards: To reiterate a plug that Laura mentioned in the talk, we’re looking for collaborators. If anyone here is interested, let one of the three of us know and we’ll make sure your information gets back to Laura.

Audience: You mentioned that this is kind of not geared towards a particular solution, but really about developing a process. Now there are other groups in EPA that are also working on a similar process developments, and also incorporating like decision, like stakeholder solicitation, things like that. How do all of these different groups come together, or do they ever come together, and what is that kind of process? How do you find like the model or the process, and how does that kind of come back to us and whatnot?

Bill LeFew: You want that one, Steve?

Steve Edwards: Of course. The short answer is they come together as we figure out what other people do. The EPA is just like every other large organization, and unfortunately, even though we try to kind of stay aware, we don’t always. Now the nice thing—so Bill didn't get into as much about this—the specifics of our project, but this is the second year that they’ve had it. They’re called Pathfinder Innovation Projects. Kind of the idea behind them is to take a step back, so everything kind of ratchets out of the day-to-day activities. At the end of our year, we had no, quote, EPA deliverables that we have to address. Now we’re hoping we can have an impact from the Durham pilot.

We’re not part of any other critical paths of that particular aspect. We kind of have the luxury within this project, while managing all of our other projects, to explore this space. We—all three of us actually just finished one of the innovation projects from the first year. What we found there is that as we worked our way through the year, we found out about a lot of these projects. Literally, just having that year to kind of explore the space helped us to coordinate and get in contact with a lot of the people.

We think that we’re—the current trajectory with this project is very similar, so the Narragansett based study that Bill talked about, a lot of that came into being when we started the proposal for this project, and mentioned it to Joseph Fiksel, and he said, “Oh, we’ve got this similar project going on.” Even though I was—I’m two doors down from Laura, I know something about what she’s done, but even my appreciation for a lot of her work has grown over the course of this project. I guess right now, I wouldn’t guarantee we know about everything. Hopefully, six to twelve months from now, we’ll be a lot closer.

Audience Member 8: Okay, I was just asking because—

Steve Edwards: If you know of anything, let us know.

Audience Member 10: Are you familiar with GED? There, we were in loose affiliation with the Citizen Support Framework Team, and I know their project was on [inaudible 01:00:32], but you seem like you’re already familiar with that one. That’s just what I was thinking about.

Bill LeFew: Gentlemen, any closing remarks on things that I left off? Laura, anything that I left off that we should have mentioned?

Laura Jackson: Well, I’m hoping that’ll be a teaser for more conversation.

Bill LeFew: Yes, in the back.

Audience Member 11: You had—right as I tried to look it up. What exactly was it EPA—

Bill LeFew: The browser, yeah. Laura, do you have the email address? I’ll give it to you right after. We’ll do it right after.

Kenneth Galluppi: Well—is this turned off? I turned it off, okay. Well, again, thank you for the short notice, for coming. This is the beginning of a dialogue that we’ll be having with EPA along these lines, so these sort of things, if you’d like to know about those, they’re here for fact-finding. Like I said, will be the first of many interactions of some sort. If you have some follow-up dialogue that you’d like to add, just contact me over at Decision Theater, and I’ll get it up to these folks, or you can talk to them directly. You don’t need to go through me, for certain. This is something we want to try to get into. I have a long history with EPA, as these folks do. This is some serious business, and I really, really believe that what you have to offer would really help the agency out, so I appreciate that. With that, I appreciate your time, and we’ll talk to you again.