Monitoring Water, Air, and Earth – Remote Sensing In the Fourth Industrial Revolution

Monitoring Water, Air, and Earth – Remote Sensing In the Fourth Industrial Revolution

Show Video

- [Narrator] Welcome to MITRE's Grand Challenge Power Hour. Here's our MITRE vice-president of engineering and prototyping, Doug Robbins. - Good afternoon, everyone and thank you for joining us to kick off our 2022 MITRE Grand Challenges Power Hour series. Big data, machine learning, AI, these are hallmarks of the fourth industrial revolution, which is rapidly evolving and so is our climate crisis. And at the intersection of these is an important topic like remote sensing, where data can provide a window into what's happening in the earth, coupled with big data analytics and this can offer scalable solution.

So that's what we're here to talk about today, about the technology, the applications and the implications for, of this very important good. But before I introduce our moderator and get going a few housekeeping rules to think about. First, the event is public and open to the press. So please keep that in mind when commenting and posing questions in the chat for our speakers. Also, if you're interested in helping us solve this big challenge or any of the other challenges MITRE's tackling, we'd love to hear from you, we are hiring, you can reach to recruitinghelp@mitre.org

for more information. And last but not least, we will have a link to more information about each of the speakers in the chat as we go along with the program. And we do want a robust conversation, so please don't hesitate to put questions you have for them in the Q and A section.

And we will have time for Q and A during the panel discussion. But with that, it's my great pleasure to introduce today's moderator Anne Hale Milgarese. Anne is an environmental science and geospatial solutions executive with a focus on bringing new technologies to market. She presently serves as the impact science Executive Program Officer for Saildrone but has a long history of this. She was the founder and CEO of PlanetiQ of Fugro EarthData and most recently, the Radiant Earth Foundation.

She was the first chairperson of the National Geospatial Advisory Committee. And among many other honors, has in 2019 received the NOAA Vision Award for her lifetime body of work. She has three professional passions, the first, very much related to today or all of them, embracing new technologies that enhance mankind's ability to map, quantify and understand the earth and society's impact on the earth. Second is building truly all organizations that span all sectors of society to answer the questions of where and why and last but certainly not least, mentoring young professionals. And she has a BS and MS in Geography from the University of South Carolina and it is my pleasure to welcome Anne today. - Thank you, Doug.

Welcome all for what I'm sure it will be an interesting and thoughtful 90 minutes of conversation. I'm honored to participate in this webinar along with three outstanding thought leaders and most importantly doers in the earth observation and climate change space. Joining me today will be Stewart Collis of the Bill and Melinda Gates Foundation, Dr. Stephanie Schollaert Utz from NASA Goddard and Dr. Steven Hamburg from the Environmental Defense Fund. The design of our conversation is each speaker will present for 15 minutes.

At the end of those three presentations, we will have a round table discussion and then take questions from the audience. Please use the Q and A button at the bottom of the Zoom interface to submit your questions for review. So with that, let's get started. Stewart Collis is a Senior Program Officer for Digital Agricultural Solutions at the Bill and Melinda Gates Foundation.

Where he focuses on digital farmer services, smart farming and digital support systems for small scale crop and livestock producers in low and middle income countries. Mr. Collis has over 20 years of experience in the agricultural information technology and was co-founder and CTO of aWhere where we originally met at. aWhere's a benefits corporation providing global ag-weather content, agricultural modeling and analytics to private and public sector partners. Previously, Stewart worked on climate-driven spatial crops, simulation modeling and ag-data management tools at Texas A&M University and the International Center for Research and Agroforestry.

Stewart has a Master's of Engineering Science and Geomatics Engineering from the University of New South Wales in Sydney, Australia. So Stewart, thank you for joining us today, what do you have to share with us? - Thanks a lot Anne, I really appreciate being invited to present to you today about Earth Observation in Agricultural Development. And I'd like to start my session just with a brief video that Bill produced for Gates notes, if we could bring that up please, to give a quick overview on background on agriculture and small-scale producer challenges in the face of climate change. (upbeat music) - Tragically, billions of people who are least responsible for greenhouse gas emissions will suffer the most from climate change. Imagine you're a farmer in Kenya, you work a two acre farm to feed your family.

You produce approximately 55 times less carbon than the average American. But now suddenly you're dealing with drought or flash floods or locusts. Even if there isn't a catastrophic event, the land you rely on is producing less and you were already on the edge of survival.

Worldwide, there are 500 million small holder farms like this. In many low income countries, more than half the population works in agriculture. So when small farms fail, whole economies fail and that means more poverty, hunger and hardship. That's why it's so important that we make agriculture more resilient. We need innovations like crops that are resistant to drought, disease, pests and weeds, digital tools to get farmers the information they need to adjust to changing conditions. Financial services from insurance to credit to protect farmers from catastrophic loss.

If we're serious about addressing climate change, we have to invest aggressively in two goals. First, advancing the technology and policies to get to net zero emissions by 2050. Second, protecting the livelihoods of families in danger of losing them to an already rapidly changing climate. That's how we avoid a climate disaster.

(upbeat music) - Thank you very much. So kicking off, if we can move to the next slide please. Is, I just like to give a little background on the agriculture development program that the Gates' Foundation, of course we're mostly focused on the second point that Bill made there at the end of the video on protecting livelihoods for small-scale producers.

That means that we're working on helping farmers move from subsistence agriculture to a small commercial farming. Farmers in the commercial world are then contributes to economic growth and we're focused on several impact goals. We're focused on agricultural productivity for small-scale producers, increasing small-scale producer, household incomes, increasing equitable consumption of safe and affordable nutritious diets year round and increasing women's empowerment in agriculture. So these are the guidelines under which we work and we develop our investments. I just gonna run through a few slides very quickly.

Bill's already mentioned a lot of this in the video but if we go to the next slide, as he spoke about you were starting to see incredible amounts of climate change. And if you could click again please, we see increasing, extreme events, whether they be droughts or floods that we saw in Zimbabwe last year and droughts in Kenya in the last couple of years in Northern Kenya. Next slide, please. We also see more gradual changes around the amount of rainfall occurring, changing seasons, seasons maybe getting shorter or the starting date is changing. And these obviously having dramatic effects on small-holder farming systems that don't have the tools to adapt to these types of changes. Next slide, please.

As we see the environment changing, we're starting to see occurrence of pest and disease that we haven't seen before. You may have seen it in the news or some rather large locust outbreaks in Northeast and part of Eastern Kenya this last year and the previous year. And we also had some fall army worm issues in East Africa. So as these types of issues increase, we've gotta have systems that allow us to predict those and have more visibility to where those issues are occurring and where they're likely to occur. Next slide, please. At the same time, farmers are faced with trying to maintain soil health, the African continent, especially is a very old continent with very old soils.

And so keeping your nutrients sustained for sustainable agricultural systems is critical to the long-term sustainability of these farming systems. So that's a difficult challenge when you've got these other issues to address, next slide please. And one of those issues is lack of relevant and timely information.

Farmers often don't have the chance to speak with extension agents, they might see once a year. If that and women in particular are usually not, don't have any access at all. So obtaining relevant information that's up-to-date and the most recent best practices of farming is a real challenge, that's information asymmetry that exists. And next slide please. And then supply chain and market access is also a challenge, not just locally but for accessing global markets and the efficiencies of those supply chains in farmers getting, small-scale producers getting appropriate pricing for their produce and then next slide. And we've all seen the effects of COVID, I mean, we've seen that here domestically in the U.S.

with inflation effecting the economy and food prices being affected. Of course, that's exacerbated when you're talking about small scale producers that are earning just $2 a day where we've seen almost 20% increase in food inflation. And it's projected that there'll be an estimated 840 million people hungry in 2030, which is 210 million more than was projected in 2015. So huge challenges ahead as we face these types of shocks, whether they be climate or COVID like shocks. Next slide, please.

Now I'm in the digital agriculture space and digital farmers services specifically and there's a myriad of solutions out there that are trying to solve a lot of these challenges. Looking at supply chain issues and advisory and finance and so forth, over 400 in Africa alone and many more in India. If we move to the next slide, please. And we see an impact here, we see increases in income, increases in productivity for each of these types of solutions, especially when you bundle those solutions together.

So we see a multiplier effect on income and productivity increases. When you say, for example, include an advisory service with a financial service, increasing the likelihood that that credit or that loan would be repaid using best practices. So they're reinforcing one another.

Next slide, please. And now just to talk a little bit about how does earth observation or remote sensing fit into this landscape, next slide please. And I'll just run through a few examples of that. This is broader than digital pharma services specifically, but there's a lot of use of remote sensing for food security analytics and also for these food price indices that are being developed.

A lot of governments today don't have visibility to what the total production is going to be in a country. And so if you're looking at a shortfall in maize production and a rice production, this of course affects exports and imports and prices. So there's a lot of efforts to try to leverage remote sensing more in a collective way, in a regional way to monitor this type of crop production and using crop analytics and satellite derived analytics.

Next slide, please. We're also using remote sensing for crop research. So we spend a lot of time and investment around upstream crop drought or development, for example. And so understanding agro-ecological zones and how those zones are shifting, how versatile regions are changing and growing seasons are changing. It's really important for us to be able to understand what types of varieties need to be in the pipeline to adapt to climate change. And so having high resolution data from products like IMERG and the GPM platform is really critical for us to be able to do that type of market analysis on those new variety developments, next slide please.

And another click please. And then, sorry, previous slide. The build is not working.

If we look at also soil property mapping, I mean, this is, we have the SSURGO in the United States, but in Africa, for example, we didn't for a long time have much on soil properties. So having visibility to soil properties allows us to provide appropriate agronomy and advisory for farmers, which is something we've recently produced for the investment called Easter, where we have both 250 meter and 30 meter resolution soil property maps, next slide, please. And then of course, early warning systems, having visibility to seasonal changes, seasonal forecasting, climate risk analysis, to allow us to assess when a food crisis might be merging or when we need to advise farmers to take different actions plan to earlier, consider different varieties and so forth is critical. Next slide, please. The pest and disease I talked about already, and there's applications out there to collect data in the field and provide immediate advisory on what to do about that issue.

But if you aggregate that information and extrapolate out using remote sensing, we can start to get visibility into the spatial distribution of those issues and enables mobilization of support for small scale producers to tackle some of these new problems that are occurring that not maybe ever seen before. Next slide, please. And then field boundary detection is a interesting area where I think AI and ML can really play a role. A lot of farmers we work with don't have land tenure.

We don't actually know where they are and we might have their phone number, but we don't have the boundary within which they are operating. And so, with that boundary information we have can of course, apply satellite imagery with that geospatial area and provide things like advice on crop health and timing of certain messages that might be relevant for that farmer at that particular time. Next slide, please.

And then localized advisory, that's really gonna be driven by everything I showed earlier. So having some of these satellite drive crop health maps, having these harvest yield predictions, harvest timing, the implanting timing, tailored agronomy based on those soil maps I talked about. And as I mentioned, farmers often don't have access to extension agents. So getting that information very easily, very directly through a feature phone is really what we're talking about. So we're talking about very sophisticated cloud-based solutions but then at the end of the day, when we're delivering that information, it's gonna be very easy to use and very easy to access. Next slide, please.

And then finally, just one of the key components here. And I think, this has already been using satellite remote sensing for insurance products and finance. So insurance people are using it for those agro-ecological zone mappings so that they can define the right policies that match those farmers' needs and the environment in which they're working. So that a payout is, kicks in at the right time, depending on the crop loss they most likely experience due to monitoring the climate there.

But then also on finance, a lot of farmers as I said, don't have land tenure, they also don't have collateral. And so looking at ways that we can develop alternative credit scoring systems that allow banks and financial institutions to have a better understanding of the types of risk that they're undertaking to provide credit to small scale producers. So having that historical information of crop health at a particular location for the last 10 years or the weather for the last 10 or 20 years that can help inform those types of models. And next slide, please. So just to finish on a project that we're working on that you may be interested to have a look at is a project called Enabling Crop Analytics at Scale, managed by our partner the Tetra Tech.

And what we're looking at is one of the challenges we see in remote sensing specifically across all of those applications, the lack of ground truth and training data. And so a lot of groups are collecting that type of data. They maybe have some of that data, but it's not particularly accessible to others.

A lot of people have small bits here and there, but we don't have a comprehensive view across the geography. And so we're looking at different data models and data sharing models to incentivize actors to share and collaborate on that type of data and information. And there may be a marketplace option, there could be some global public good components to that.

We're very reliant on crop cuts at the moment. And a lot of people collect that data, but we're also talking about innovations in the space. So how do we lower the cost and increase the quality and volume of ground-truth and training data? And we're looking at new technologies like ground LIDAR or LIDAR on a phone or a tablet or just simple photographs.

Low-flying drones, we're providing some funds to innovate in this space to improve the quantity and quality of ground truth data. And then trying to pair that with the data sharing model. If you're interested in more, please go to the website listed here. We actually have a stakeholder database that we're plating. So if you navigate to the stakeholder page, we'd love to get you to input your work, if this is an area you're interested in.

And you can reach out to the team on the website as well. And then I'll just finish with a final slide on innovation specifically is, this was a quote from Bill's book on the climate disaster is you're thinking about innovation, not just in terms of inventing a new product or service, but how do we then scale that service? And that's a lot of what we're working on. We need that pipeline of new inventions, but we've got to really scale those out to millions of farmers. One of the problems we have is that many of the solutions out there are reaching a small number of farmers and we've got 500 million farmers to reach, and we're facing a climate to challenge here in the future, as well as food availability.

So innovation is one way in which we can address those issues. And that, with that I'll close, thank you very much. - Perfect, thank you, Stewart. Just a couple of questions.

I don't worry so much about industrial agriculture. I'm sure they will adapt. They have access to finances, they have the science, they have the supply chains. Though I'm sure we'll all pay more for what we eat. What I worry about is how do we really support those 500 million small lease holder farmers across the globe? We have much of the data, sure we need more. We're certainly developing the algorithms and the training data sets for ML at scale across earth observations.

I was really encouraged to see that market map for Africa that you showed. As you know, I've followed this field for a while, but scaling this industry is all about that last mile, right? And what obstacles do we have to overcome to deliver the crop intelligence at that last mile to make it actionable information for those small farmers in developing nations? - Yeah, thank you for that question. It's certainly a challenge. I mean, farmers, small scale producers in low middle-income countries absolutely need these services and finance insurance, subsidy programs, advisory market linkages, and governments are really challenged to provide these services on their own. They often don't have the digital infrastructure or the digital tools to really scale these solutions.

And, so we see as that landscape map showed there's an awful lot of solutions out there. But as I just mentioned, a lot of them really challenge to reach scale with over 500 million farmers to reach. We've got very few of those solutions have reached more than a million farmers. And I think that's for multiple reasons, agriculture is a very tough market. It's very lean even in the United States. A lot of Ag tech companies are challenged to sustain themselves.

And especially in low and middle income countries, it's a riskier investment for, even for impact investors. So I think, the cost of these services need to be very affordable, small scale producers only earning $2 a day can't afford to pay much for these types of services. So how do you distribute the cost of those services across the financial institutions? Input providers off-takers is one way.

I think also there's barriers around language literacy and digital skills. So, and access to handsets. One of the big costs for farmers is their mobile data plan to even access these types of information services. So I think lowering those costs is another area. And then, that data infrastructure is very challenging. There's not a lot of global public goods, like the soil map I showed.

So how do we improve on those types of elements so that these Ag tech innovations and digital solutions don't have to build the whole stack themselves. So a lot of issues, but we're working to address as many of those as we can in the development community. - Okay, great. Thank you, Stewart.

I'm sure we'll have many more questions for you as we come back around to the round table session. So next I wanna introduce Dr. Stephanie Schollaert Utz. Stephanie is an applied science manager at NASA Goddard Space Flight Center, where she leads activities in advanced, to advance the practical applications of NASA's data and science, connecting researchers across the Earth science division with end-users in developing external partnerships and fostering innovation, innovative uses of earth observations for societal benefit. As a part of this effort, she leads a team that convened six working groups with scientists and stakeholders around food security, air quality and health, climate and environmental health, disasters, mission applications in the Chesapeake Bay. Her research focuses on the response of Marine and aquatic ecosystems to physical forcing through the use of satellite data, Institute measurements, model outputs and statistical reconstruction. Stephanie has a PhD in atmospheric and ocean sciences from the University of Maryland, a master's of science and physical oceanography from the graduate school of oceanography at the University of Rhode Island and a BS in oceanography from the U.S. Naval Academy.

Stephanie, thank you for joining and we look forward to what you have to share with us today. - Thank you, Anne. And thanks very much to Victoria and MITRE for inviting me to this really interesting discussion. So I'm going to start, my slides will come up.

I'm gonna start with an overview of how NASA studies earth as a system. And if you could click on this as, yeah, this is an animation, exactly. So this demonstrates all of the different variables. NASA is, we've got about 20 satellites monitoring different geo-physical variables from vegetation on land, in the ocean, soil, moisture, precipitation, salinity in the ocean, temperatures, winds and clouds obviously. And all of these, we can see through satellite looking at earth, all of these are interrelated.

So the precipitation leads to soil moisture, which leads to vegetation. And then later on that can lead to fires. And by studying earth as a system like this, we have really gained a lot of insights into how the water cycle, the carbon cycle and the energy cycle are interrelated. But the biggest challenge for us for NASA is we have this understanding globally and we have the combination of the satellite data. And then we also have a lot of Institute data that is essential for calibrating and validating the satellite data and then assimilating into our models where we can do predictions and projections.

We have these at a global scale. And so our big challenge is how can we scale it down to deliver it locally at the resolution and scale that people need for decisions. So next, please. So this is our current fleet and our future fleet is shown in the orange at the bottom there.

So we have approximately 20 earth observing satellite systems with, in partnership, many of them in partnership with other agencies and international agencies as well. So for example, the most recent launch was Landsat 9 with our USGS partners. And that's extending the legacy of the sustained land imaging program, which is been going for over five decades. So that's really important as a climate data record dataset. And I'll mention now that from my role as applications, what we're doing with missions in the past, because NASA is a research agency, we would launch missions, data would be collected.

And then we would work with NOAA and other agencies, USDA to figure out how to apply the data to practical decisions. But now we're thinking about this early on in the mission study and design phases, we're addressing applications, we're starting working groups with our stakeholders and we're getting their feedback early on. Consider this with the engineers in mission design process. And the hope is that we can, by doing that, we can benefit society more and have if there's any kind of adjustments that need to be made that that could help get more value from these, next please. So specifically on the topic of climate change, I just wanna mention a couple of different activities. I'm gonna run through a lot of examples of work done by other people across the agency and across Goddard.

And at the end, I'll talk about my own project and research in water quality. But on the topic of climate change, we've got a number of new initiatives including this week, our new chief scientist and senior climate advisor was announced Dr. Kate Calvin. We're also working with other federal agencies to integrate our science and data products into a climate resilience information system. And then, we've got the disasters program, and this is an example, one of the other activities NASA is doing right now is called climate adaptation science investigators working group. Which is looking very closely at all of the NASA centers and facilities, the idea being if we can use all of our science information and downscale it and be heard within our own agency and have our own facilities adapt and be more resilient because of that, then what we learn in working with those facilities managers, we can then scale it to the communities around the centers and then hopefully to the larger population. Cause it's not an easy task to figure out how to do this and to figure out what needs to be downscaled.

The project right now is specifically focused on extremes of temperature and precipitation, fires and air quality, coastal and inland flooding and sea level rise. And this is an example from 2017, when Hurricane Harvey hit Houston, that climate adaptation science investigators group had been meeting and advising Johnson Space Center. And because of that, the impact of all that extreme rainfall was less and the stormwater runoff was somewhat mitigated, next.

And this, I'll mention this activity is being led by Cynthia Rosenzweig at NASA GISS in New York, which is part of NASA Goddard. So just today, you probably heard the press release by NASA and NOAA about the eight warmest years on record have happened in this decade. And I think 2021 in particular was the sixth warmest for air temperature. But earlier this week, there was the news put out that it was the warmest in the ocean, which is extremely concerning because the ocean has a lot of thermal inertia.

And so a warming ocean, the ocean has been a big buffer against climate change and a warming ocean is very, very serious. And we don't even completely understand all the possible ramifications of warming our oceans. Okay, next, I'm gonna fly through a lot of examples. The 2021 disasters season had, NASA was monitoring a lot of the different disasters from the tornadoes in the Midwest, Hurricane Ida and typhoons.

And we, so we provide a lot of information. We have a disastrous portal to support risk reduction, response and recovery. We're working very closely with FEMA and other agencies, and then internationally with aid organizations and others, next. We could also see in our satellite record the cryosphere. And for example, the Arctic sea ice has been declining 13% per decade, next.

That gets released every September. Between our satellites and our in-situ observations that are assimilated into our models. We can see that carbon dioxide levels are the highest they've been in 650,000 years. This is an animation you can find on our website where it shows CO2 emissions and then transport in the atmosphere.

And you'll notice that most of it is in the Northern hemisphere where the industrial emissions are, next. And then we can see vegetation. So here's an example of a researcher who does, looks at deforestation in the Amazon and with the Landsat record that's almost five decades now, we can see land use changes.

So this is critical for understanding the terrestrial carbon sink, next. And then working together with some of the commercial satellite providers, a recent study by both Goddard and university scientists counted using artificial intelligence and the planet data and other commercial satellite data. They were able to discover a large number of previously uncounted trees in the African Sahel. And they had a nature paper about that. And then NASA satellites pioneered the use of ocean color remote sensing to see where there is, where the Marine ecosystem is and its changes. And now, using the benefit of that synoptic view, combined with ships and in-situ sensors to get at the vertical distribution of features in the ocean, we're undertaking to understand the role of the Marine ecosystem in the ocean's carbon pump.

So that whether the ocean is acting as a source or a sink, and whether the biology that sinks out of the surface ocean, how quickly, how deep it's sinking and this is part of the exports field campaign. That's this multi-year multi-mission. So for sea level rise, this is altimetry and that's been collected since the early 1990s. And this record shows us that sea level has been rising most due to the melting cryosphere and thermal expansion of the ocean. And all combined the average over the past 30 years is approximately three millimeters per year of sea level rise in some places more than others. And this map shows the spatial distribution of that, next.

And then we have active tools near real-time tools to monitor fires with the forest service, both in the U.S. and Canada. And we also have a lot of research ongoing to look at pre and post-fire vegetation, fuel availability and whatnot, next. And then this leads to forecasting global air quality.

Our modeling, GMAO is our Global Modeling and Assimilation group. They have these, they put out these forecasts of global air quality. This one is showing ozone. And so at ground level, that's a pollutant. And then in the stratosphere, it protects us from UV radiation from the sun, next.

And then NDVI, the Normalized Vegetation Difference Index informs global agricultural outlook. So we've been in partnership with the USDA for an agricultural service for many years to help with the global crop monitoring. And this helps to prepare for, there's also efforts with USAID on Famine Early Warning System. So using NASA satellite data, they can anticipate where there's gonna be humanitarian crises and plan food aid, next. And then soil moisture is a newer, more recent input into our products with USDA. So working with the National Agricultural Statistics Service is a downscale product to one kilometer.

So using the in-situ measurements to down sample the resolution of the satellite, this is extremely useful product here. It shows the areas that are green and blue are the, have higher soil moisture. And then areas that are red obviously have lower soil moisture.

And so this is very beneficial for again, understanding the soil moisture in the land can, gives information about drought, next. And then altimetry can show us the heights of major lakes and reservoirs around the world for agricultural and regional security. Also, this is for an ag service, next. And our projection group at NASA GISS, they do the agriculture model in a comparison project, and this has helped, helping to inform some of these small holder farms that Stewart talked about in countries like India, Nepal and Pakistan, where they're dependent on the hydrological cycle. The long-term projections from AgMIP can help with risk planning and helping them to make better choices about what to plant for example, next.

Our models are also used for understanding renewable energy, for example, the, for solar radiation and solar panel placement or efficiency and wind, next. And then here's a specific example of working directly with stakeholders and having an impact on policy. In 2020, our colleagues at NASA GISS worked, working with New York city, they passed heat equity legislation using NASA analysis of hotspots. So they can see where there are urban heat islands and they were able to, in fact, you can see the nice blue rectangle in there is a cool area for Central Park. But they use that to pass heat equity legislation, next. Okay, and then finally, this is my project.

So I work with partners in the Chesapeake Bay to monitor water quality. Specifically, I work very closely with NOAA who has operational jurisdiction over the water quality products in the bay, but we're preparing for our upcoming hyperspectral missions. And we're working with Maryland Department in the environment and the Shellfish Monitoring Division.

There's a growing aquaculture industry in the Chesapeake at the same time that we've got more runoff from land during these extreme rainfall events. And we've got aging septic systems around much of the bay. And we recently did have a runoff event that led to fecal coliform poisoning because there was a shellfish farmer who didn't hear that there was this problem with the runoff.

So this is a serious concern, and we're hoping that by using satellite data, we can, and by using hyperspectral data, we can move beyond just looking at chlorophyll and water clarity, but start to detect some of these other features in the water and proxies for them, next. And so this is, partly, this is getting ready for some of our upcoming hyperspectral missions. One of which is surface biology and geology. And the other missions that are coming are clouds, convection, precipitation and aerosols, mass change and surface defamation change.

These missions are coming in the next 10 years and, next, this is gonna be all together up on the cloud. Having an open science framework, there was a press release about this last summer. And the goal is to better inform decision makers, how our planet is changing on previously unimaginable scales. So we're just starting to get up onto the cloud and do open science, there's challenges. NASA lawyers haven't caught up to this. They're not letting us put our code on the cloud as easily as we would like, but it's a work in progress.

So with that, I will end my talk and take any questions. Thank you. - Right, thank you, Stephanie. Well, that's a lot of interesting work that's evolved in earth science enterprise over the last 10, 20 years. I guess one of my questions is around NASA, clearly NASA, ISA, JAXA and the other space agencies of globe have so many satellite programs and instruments, but still the oceans are terribly under sampled for a myriad of factors but perhaps most importantly, dissolved carbon. And the science community has long assumed that the oceans are sink for carbon.

A recent study in 2019 showed in fact, the Southern ocean is a source of carbon for a significant portion of the year. Currently there's a mission that is funded by google.org but conducted by your Alma mater at the University of Rhode Island and European Commission on meeting wrench, weather forecasting and Saildrone to evaluate the functions of the Gulf Stream. One of the great boundary currents, as it relates to being a source or sink for carbon.

If this basic assumption is wrong, what do we need to do to build a more accurate global carbon budget and model? - So getting back to Stewart's point about the importance of ground-truthing and training data and I'll just, I'll mention, I believe the research you're talking about is by professor Jamie Palter at URI, which did have National Science Foundation Funding. So I'll put in a little plug for the federal government there that so important for our science enterprise. But one of the great things that's happened over the past 20 years for these BioGeo, well Argo floats.

So Argo floats are profiling floats, they're autonomous they're, the last I heard there were maybe more than 6,000, I haven't kept up. But those are really filling in our understanding of what's happening down to, hundreds of meters in the ocean. And now they have probes with BioGeo chemical sensors on them called BioGeo Argo floats. And those are really critical for filling in the missing gaps and understanding the ocean. - That's great, thank you.

We have a question online, "How do researchers connect to these datasets "programmatically, what would be my first stop "to incorporating this (mumbles)." - Yeah, so the, all of our datasets are publicly accessible. The climate data are, a lot of them are online and for example, we have like the FuseNet and the hydrological modeling, the LDES and the FLDES and the GLDES. Those are publicly accessible. Some of the models, some of what we're talking to potential partners about is having them actually run the models for their, down-scaling it to their particular area. That's something that we currently do on our super computers, so we get people access to our supercomputers when they need to do that.

That's one of the things we're exploring, the possibility of putting that on the cloud. It's a lot of compute to run those models and so it's not trivial, but we're exploring that. - Okay, right, thank you.

And I'm sure we'll have many more questions in the round table discussion, thank you, Stephanie. All right, let's move on to Dr. Steven Hamburg. Steven received his undergraduate degree from Vassar College and has MFS and PhD from Yale University in Ecosystem Ecology. He was a post-doctoral fellow at Stanford University and a Bullard fellow at Harvard University. Prior to joining EDF, he spent 25 years on faculty of Brown University and the University of Kansas where he served in numerous administrative positions in addition to his teaching in his research.

He has served as lead author for the IPCC and was recognized as one of the scientists contributing to the award of the 2007 Nobel Peace Prize. He's also been award the U.S. EPA Agency Merit Award, Region one, twice. And he's published more than a hundred scholarly articles and papers on biogeochemistry, climate change impacts on forest, carbon accounting and methane emissions. Welcome Steven, I can't wait to hear more about MethaneSAT. - Thanks so much, it's great to be with everyone.

And it's really great to follow the last two speakers, I'll note that the first three letters of all of our first names are the same. So first for me and I really would like to, if I get the first slide, really build on what was said before. So I wanna talk a Stewart talked about the impacts on small-land holders across the globe really provides the why for my talk. Why am I gonna look at methane emissions using remote sensing.

And then Stephanie really talked about remote sensing in a broad strokes. And I'm gonna sort of take all of that and to give you a specific example, how we can take advantage of the sensing revolution. And really make some things happen at a scale and a level of impact that we probably never could have thought of before. Next slide, please. So as we think about methane and we generally don't recognize the important role can play within reducing the rate of warming over the next few decades and thus the impacts that Stewart so eloquently described.

So we talk about methane emissions, I'm only talking about anthropogenic emissions, there are natural emissions that have been around for millions of years if not billions. And we can think about those sources through this waterfall diagram, I'm not gonna get into all the details but we can think about agriculture, livestock, rice, fossil fuel production and then waste both from solid waste and landfills and waste water. I'm gonna focus on oil and gas today, but we really need to address emissions across all of these sources. And the good news is we really have momentum now in the, at the conference of the parties, in Glasgow we had for the first time, a coming together of more than 100 companies to recognize the importance of methane as a greenhouse gas. And in particular recognizing how critical it is to reducing the rate of warming and thus the damages and impacts that Stewart so eloquently talked about, next slide please. And as we think about that, this is just some modeling that the team that I work with at EDF has done to really look at, what are the real temperature impacts of these kinds of different methane mitigation opportunities.

And so the top line is the business as usual, what we anticipate the climate impacts of methane will be moving forward. The yellow line is what we can do right now with basically minimal economic impact. And then what's technologically feasible is the green line and then of course, if we can fully mitigate all anthropogenic emissions is the bottom line. But you'll see for those of you familiar with climate change, these are fairly big numbers.

They may look small they're tenths, from a 10th of degrees Celsius to half a degree Celsius. And if you eliminate them totally to half a degree Celsius to one degree Celsius, those are huge numbers in the climate change world. And then below you see the different sources of that mitigation, oil and gas being the largest chunk of it but they cross the different sectors. So we have an enormously powerful lever that we need to push down on to reduce the rate of warming. And we can do it quickly, if we can do it quickly, we're going to reduce the rate of warming quite significantly from an impact standpoint, both from human society and natural systems in the next few decades. Next slide please.

And just from the fossil fuel side, as I said, I'm gonna focus on that, it's just illustrative because it is the easiest sector, the biggest sector where we can have the most impact in the immediate near term. So if we just model, so some people would say, "Well, we shouldn't worry about it because "we're gonna get rid of fossil fuels." We need to address climate change and we absolutely do. But timing matters, so the graph on the left-hand side shows what the trajectory of a missions would be from the oil and gas sector with no action. What would they look like with a decarbonisation on this structure and the timeframe it's a little faint there, goes out through 2050, you could argue that that's not as aggressive.

These are data from the International Energy Agency, but the key here is that we need to decarbonize and reduce methane emissions from the fossil fuel sector at the same time. Because when we do, it really matters for the temperature impacts over the next couple of decades. Because methane is the dominant determinant of the climate warming we're gonna experience over the next couple of decades more so than CO2 from the emissions over that period of time. So methane is really critically important to the damages that I'm gonna experience in my lifetime. And we can reduce oil and gas methane emissions by 70%, according to the IEA and we can, roughly half of them will pay for themselves. So literally in the net they're free and we can do much of it at really limited costs, next slide please.

So what do we need to do to address greenhouse gases to emit the climate crisis? We need to know what's being emitted, where it's being emitted, how much is being emitted and how are those emissions changing over time? And this is the role for the International Methane Emissions Observatory, I'll come back to that. But this is the challenge we have across greenhouse gases, across the globe. Currently, we do not have a mechanism by which we actually collect empirical data to answer these questions and present that openly to the world.

We're beginning to see a revolution using both remote sensing and integrated systems that Stephanie talked about to be able to bring this data to the world, open for everybody to look at. And we're gonna, we believe at EDF and I believe personally that this revolution will change the rate in which we can start to ameliorate the emissions and thus the impacts, next slide please. This is just an illustration of some data from Alberta, Canada, just the kind of problems we have with data quality with respect to greenhouse gas emissions.

On the right hand side you'll see from Alberta, the, on the left bar, the green one is what the industry reported emissions are. The official inventory is a bit higher, but not a lot higher and then measured, directly measured data is multiples of that. This is not ubiquitous this pattern but it is the norm, most cases, the reported emissions are less than the actual emissions. So the key is how do we create accurate data that characterizes emissions across sources and across the globe? So I've been working for the last decade with a host of hundreds of scientists from around the world to begin to build this picture of what is being emitted and how much is being emitted and how it's changing. Next slide, please. So these are data from a synthesis that we published in science a few years ago, taking all the data we generated over the proceeding seven years.

Again, couple hundred scientists across at least 50 different institutions. And you'll notice on the blue bar and this is going across the natural gas supply chain from drilling and production to gathering, gathering being bringing all the gas together, processing it to produce a uniform product. Transmission and storage making sure it's available to you and my house here, sitting in Providence, Rhode Island, I have a natural gas boiler in the basement that I can turn it on and off as its cold, it's been cold recently and it will get cold again. And then of course the local distribution to get it into my house. And we see of course the drilling and production is where most of the emissions occur, but the emissions are basically double what the EPA suggested.

And since we published this paper, actually the estimates for EPA have gone down, emissions have actually empirically shown to go up. So there's a major gap between what we say the estimate is and what the empirical data is. Now, this took lots of people, working lots of scientists, pulling together all the data to create this global, this U.S. national estimate which had never been produced with empirical data before.

But we need to be able to do this basically on a routine basis. So if we can think about this as a snapshot of what's going on, we need a motion picture, we need to get this continuously, we need to know how it's changing and how it varies spatially. Next slide, please.

So to do the work that I described here, we have top-down methods from aircraft on the left, on the top left and a tower, a tall tower on the right, top right. And we can use these approaches by measuring the methane using different types of transport models, to inversions, we can calculate with the mass of methane's being admitted from a region or an area. We also do it from bottom up, we go out in the field, you see the picture on the left with some researchers working at the University of Texas. Study we did some few years ago, looking at individual sources and quantifying them. Or a project we did jointly with Google, where we'd use the street car with instruments in the rear and we were quantifying methane emissions from local distribution leaks over tens of thousands of kilometers across 13 cities in the U.S. So we can do these methods, we can do it in site-base making sure we have accurate estimates, but again, we need to scale this.

And this was a point that was made, we really need to scale the data. We have to have lots of data and we need to be able to do it in an integrated way, next slide please. So that's where we come to using a satellite technology, I'm showing you my favorite satellite 'cause I'm co-directing the construction of MethaneSAT. This is a philanthropic only funded satellite and I'll talk about it more later.

But we really need, as I said to have this motion picture being, collecting data regularly, being able to process it. But we have to do it in a way as I said, that produces what's being emitted, where it's being emitted, how much is being emitted and how it's changing over time. The satellites now flying to measure greenhouse gases don't do that.

And part of that is you get concentration data but it doesn't take it all the way to the policy relevance. The key is building that entire ecosystem, automating it and in the case of MethaneSAT as Stephanie talked about, everything's gonna be done in the cloud. We're already in the cloud, we're already building the algorithms, testing them with aircraft, pulling measurements. So it's a cloud-based system to provide global data on oil and gas methane emissions with a greatly increased level of precision. Next slide please.

And what we can do with these kinds of data of using existing satellites, this is from TROPOMI which is a Sentinel-5P based instrument, this is a European space agencies. Once the satellite went up, we started using the data, took 11 months in quantifying methane emissions, in this case for the Permian Basin, which is in the U.S. the largest oil and gas producing region in the country and one of the largest in the world. Now, this gives you concentration data but again, that's not policy-relevant. We need to know how much is being emitted, where it's coming from, so next slide please. We can then use an inversion process to be able to generate on the right-hand side, the actual emissions and quantify them to the total emissions.

Which as you'll see relative to the prior, which is the inventory, is more than two times higher. And now the challenge we have with TROPOMI as you'll see, the pixel sizes are quite large. So we can't really get to a policy relevant in this, who should do what we don't know, we know what the total is we know how it varies over space in a general way, but we don't have the level of granularity that we need to really make the data actionable so we can reduce emissions effectively and assign those emissions to specific facilities. Next slide please. So these data are just showing you the different types of satellites, at the top MethaneSAT which will go into operation in 2023. It's got a grid size of roughly one by one kilometer with actually sub pixel or the native pixels of 100 meters by 400 meters, a swath of 200 kilometers and a precision of two to three parts per billion.

Now, this is important because we wanna be able to characterize all emissions from fossil fuel production, particularly oil and gas. And we wanna be able to ensure that we can do it around the globe and see all of the emissions that are occurring. Now, there are other satellites, no one single satellite is going to be able to provide all the data. We have some of them already flying, I've mentioned TROPOMI in the middle of the list, which is a European Space Agency. We also have GHGSat which is a target mode instrument, so you're seeing small areas, TROPOMI very broad swaths, a higher precision. So it helps give us a background as I showed you from the Permian but it doesn't give us the data that we need to fully go carb, to have the data necessary for implementation.

Next slide please. So what I'd like to do is I'm gonna show you a video of comparing and contrasting three different satellites and their complementarity. So the first is TROPOMI that I talked about, sort of giving these global mapping of what's going on. And then we're gonna see GHGSat, a target mode that takes sub-spots that direct it and then followed by MethaneSAT which gonna do large areas with high degree of precision, where we can see and quantify all the emissions and may keep maps totally understanding where the emissions are coming from. So this is a contrast of the kind of remote sensing data and what we're doing is fully automating the inversion process so that the data that we generate is what's called level four.

Which is FluxData that will give us exactly what you see here, will give us a heat map at a very fine scale, so we can do source attribution, we'll see large points towards us and we'll see the total emissions from these large geographies around the globe. Now we'll have for the first time of any greenhouse gas, detailed quantitative data come in near real time and having with repeat measures from year in and year out. Where we can start to understand the patterns of emissions, who's responsible and what actions are being taken and how effectively are they mitigating emissions. Next slide please. So if we could hit it again and so we move on, thanks. So the Methane Mission as I described the goal, the primary goal in the near term is to help catalyze a 45% reduction in methane emissions from oil and gas and production by 2025.

And as I mentioned, this is fully funded philanthropically, putting together a team who is building it from commercial space and academia and people who've worked in government and bringing everyone together to do this. We're gonna do it for all the oil and gas production regions and minimum of 80%. We're gonna get area missions and point source emissions and it's gonna all be with a FluxData product that really allows us to be policy relevant from day one. All the data is publicly available, it all be and the scientists can get into the details and the rest of the world can see it. Next slide please.

And then, but that's an integral piece of it but it's not sufficient. I described different sources of data. So announced just little in 2021, is an effort called the International Methane Emissions Observatory where we wanna bring all these data together, all the remote sensing data, the science data, data from the companies voluntarily reporting their data as well as their inventories. And this is under the UN Environment Program, we're gonna have the science, it'll be transparent and allow for implementation, providing solid data integrated on a global basis by an important global multi-lateral that is trusted. Next slide please.

And to do that, we've got to do as we described, we've gotta use a whole host of machine learning and AI techniques, we've gotta reconcile the data across these different platforms. We've got to then generate final products that people have confidence in with transparency so that then decisions can get made in the political arena. And at the country level and globally and even at the local level to affect strong mitigation options and in turn reduce emissions and address the climate crisis at a scale and at a pace that up til now we've not been able to do. That's the vision, we very much hope we can realize it, I knew it was off to a strong start.

As I put on the beginning, I chair the science oversight committee that's about taking, realizing this vision on the data side and last slide and thank you very much. - Great, thank you, thank you Steven that's fascinating. I've watched the formation of MethaneSAT LLC for years and the weaving of your suppliers from your partnership with New Zealand and SpaceX and Blue Canyon Technologies and Rocket Labs, you certainly have threaded that geopolitical science needle, well done. I'm interested in your business model.

So you talked about the data being open for the science community. I assume you've gotta generate some revenue, is the idea that you will serve as commercial consumers and government regulators on a fee for service approach or. - All the datas, the L four, L five, L four data, so everybody free of charge, no matter who you are and where the world will get that product free of charge. If you want to, you have, everyone will have to sign a license agreement which says you will not commercialize it, that's the only restriction. If you wanna commercialize it, then you have to get in an agreement with us. But we have funded this philanthropically and so the only costs that we have are operating.

And as you said, we have a partnership which I failed to mention, thank you for raising it with the New Zealand government. And we are exploring what the income might be from that commercialization. But the overwhelming majority in our costs are done philanthropically and we in no way are going to restrict access to the data. We believe transparency is the most effective tool for affecting rapid mitigation, which is what we all need to address the climate crisis. - Right, okay. So in your data streams, if there are methane emissions detected at coal mines or natural gas facilities, what process is MethaneSAT and EDF gonna use to ground-truth these anomalies.

How's that going or not anomalies but signatures. - It's observation. - Yeah, how's that gonna work? - So of course the remote sensing itself is with the T-Cons to just make sure we're getting accurate data from the sensor, the spectrometers. But we also, because we're doing ground-based work, the science that I des

2022-02-03 23:32

Show Video

Other news