 Good morning everyone. Thank you for joining us this morning. We're joined today by Nigel Cizer, the Global Director of World Resources Institutes Forest Program and the WRI team here in Indonesia. So Nigel here will demonstrate the Global Forest Watch which is an online forest monitoring and alert system. He's going to talk about how GFW can be used in Indonesia as well as worldwide and what WRI's plans are to support Indonesian stakeholders with early detection and monitoring of fires and haze during the upcoming dry season. Before we begin, I'd like to just give a short introduction to Nigel. So as I mentioned, he's the Global Director of WRI's forest program. This includes the Global Forest Watch, the Forest Legality Alliance and the Global Restoration Initiative. Prior to WRI, Nigel served as the Vice President for Asia with Rare. He also served as the lead advisor on climate change and energy issues in Asia to form a president, Bill Clinton and the Clinton Global Initiative. And he's also worked with UNEP in Nairobi and he established the Nature Conservancy's Asia Pacific Forest Programme. So thank you very much Nigel for coming here and for the WRI Indonesia team for joining us. Thank you very much. It's a great pleasure to be here. I actually used to have an office here so it feels like coming home. When I was with the Nature Conservancy, I had an office for three years here at C4 when David Kymowitz was Director General and we did some wonderful collaboration together, particularly with Kristof Obzynski and his team. There's several of my colleagues here. I would like them to just stand up and say hello so that you know who they are. Maybe each of you could just introduce yourselves and just name and what you do very briefly. Good morning. My name is Paula Mastre. I recently joined the WRI. I was with the World Bank in Jakarta for four years with the environmental team and I'm the recently hired Indonesia Manager for the Global Forest Watch. So helping create partnerships and leading engagement with Indonesian stakeholders on the GFW and the sister websites as well. Good morning. Selamat pagi. My name is Hidayah Hamza. I work at the GIS with the WRI Indonesia just for this recent two months. Thank you. Hello, Selamat pagi. My name is Tanya Puspita-Firdausi. I'm with Hidayah as the GIS Analyst at WRI Indonesia. Thank you very much. So we actually have an evolving collaboration and partnership with C4. We were just having a meeting with your Director General. Very much hoping that that's going to deepen and expand significantly over the coming years and I think there's a good chance that will happen. We've also just joined as an implementing partner for Landscapes Day. I think we're the only NGO at that level in the organization of Landscapes Day for Lima. So close links, long historical links. I think we received one of the, when C4 was first established I was telling the team this morning when Jeff Sayer was setting things up here. We actually I think got probably the first grant that C4, at that time C4 was giving out some grants. I think we got the first grant for some work we were doing in Brazil at that time in 93, I think it was. So the connections between C4 and WRI go way, way back. For those of you who don't know, WRI is a global think tank and do tank. We've got about 500 staff now on the team and we're organized across six programs. Forests, Water, Food, Climate Change, Energy and Cities. Here in this region our most active program is the forest program. But our cities program is also becoming and our climate change program also becoming more and more engaged here as well. Here in Indonesia we have a team now of about 10 people working full time on forest issues. So what I'm going to talk about today is the Global Forest Watch initiative and how that has been evolving. I'm sure some of you have heard a little bit about it, seen some of the news coverage and so on. It's a major initiative at WRI, it's our largest effort on forests right now. It actually builds on work going back over 20 years. We launched the first Global Forest Watch in 1999 as part of something called the Forest Frontiers Initiative. And that was an effort that was very much based on reports like this. Publishing as up-to-date information as we could about what was happening to forests in key. At that time we focused on what we call the intact landscape. So Russia, the boreal, the Congo region, Indonesia, parts of the Amazon, the Guyana Shield and so on. So you can see here a sampling of some of the reports. We actually did the first interactive atlas of Indonesia's forests at that time in collaboration with the Ministry of Forestry. So this was the norm and what we were doing merrily until four or five years ago. But of course it's not good enough. This kind of information is typically several years out of date by the time it's actually published, read and policymakers are starting to look at it. If they read it and look at it at all, which I think my sense is that they increasingly do not want to look at these kinds of reports and read them. And at the same time, our friends who are in the business community, the investment community all take this kind of information completely for granted. Updated every second or almost every millisecond online. Thousands of companies, the world's economies. You can see what's happening almost instantly. It's free. It's in many languages and it's very easy to use. We all probably look at this kind of information from time to time. This is the BBC website and there are many you could choose from. So our sense, having worked on information about forests for many years, is that it's time for a new approach. That we need to get as quickly as we possibly can to near real-time, if not real-time information about what's happening to forests around the world and deliver that in a format that policymakers can use and understand almost instantly. So it's a hugely ambitious initiative, but that's what set the heart of Global Forest Watch. As you know here better than anybody, there is a huge amount of information about forests. There are all kinds of studies and data sets being produced constantly. Of course there are still some very big gaps, but most of that information does not find its way to those who need to be using it. So the basic mission of Global Forest Watch, the initiative, is to take that wealth of information and the complexity therein and make it as simple and accessible for policymakers and policy relevant as possible, whether they're in government, NGOs, business, communities, wherever they may be. And of course the good news here is that the technology now enables us to get much closer to realizing that vision. When we started this 15, 20 years ago, Landsat data was not available for free. You had to buy it from the USGS and it was very, very expensive. Almost no one was doing anything with Landsat data at that time. Now it's all freely available, the entire archive, let alone all the other systems. Cloud computing enables us to take terabytes and terabytes of data and run parallel processing and sophisticated algorithms with those huge data sets and come up with answers to some key questions or at least get closer to those answers. And people are used to sharing this kind of information with social media and online in various ways that we would never have imagined even three or four years ago, let alone 15 years ago. So a partnership of about 40 organizations has come together over the last two and a half years to create Global Forest Watch. And what you have here is, I mean you've got UN agencies, UNEP FAO is also a collaborator now, some very key technology providers, Google, Esri, Planet Labs who are high resolution satellite imagery, Janix who are processors of high resolution imagery in Russia, donors, Norway, USAID, the Dutch, the British, the GEF, and then various NGOs who have expertise on this around the world together with some key practitioners in different regions around the world. There's a number of other groups, not on this slide that are also involved, but these are sort of the key groups. And they've come together to create Global Forest Watch which we launched as you may know in February this year. This is the home page. And what I'm going to do now is I'm going to switch to a video, which is a five minute video, which just takes an animated walk through some of what's on the Global Forest Watch system. And I'll talk while that plays. It basically stops me talking too much because I have to keep up with the video. And then we'll come back to a few slides talking about what's going to happen next. So let's start the video here, the full screen. So here you have the home page, which is where we just were. And you click on here or on the map here and it opens up. This is meant to be a very simple and intuitive system. The resolution on the screen here isn't great, so you're not going to be able to appreciate this. But you've got a set of pull down menus. You can click on those tabs. There's information about what's in the data set for each of those tabs. So this is forest cover change. And what's selected here right now is the UMD Google data set of global tree cover change. There's near real-time, former, all the world's fires from the NASA Active Fires MODIS system. There's forest cover layers, forest use, logging, palm oil, plantations and so on. All of the world's protected areas from the WPDA database. The people layer, we're adding data on traditional and indigenous land into the system right now. And there's a news tab. We'll come back to that in a moment. So zooming in here right now into the western Amazon, what's up on the screen here right now is the Google UMD high resolution global data set that was published in Science last year. So they're core partners with us. And we've pulled up one of the protected areas here. This is the Surui indigenous area. And the pink areas are tree cover loss year by year. And you click on the time slider and the tree cover loss comes up year by year from the Google Hansen data set. This is 30 meter global resolution, a very, very powerful data set. Obviously some significant limitations as well with that, which we can talk about. We can change the backgrounds by clicking on here. So this is bringing up a satellite, a composite satellite image background. In some places it's very high resolution like this. This is actually some digital globe imagery that's within the Google database. So depending where you are in the world, switching to the satellite view, you actually can pick up an awful lot of context at surprising resolutions on what's going on. So zooming back out, you've got the zoom buttons over here. It's just like using a Google map, coming over to Indonesia. And we'll pull up some of the other data sets in the system. So from the forest use tab, here are the palm oil concessions for Indonesia. This is Ministry of Forestry data. Obviously limitations with that, but it's the best that's publicly available. Click on any of those and the name of the company and some basic information about the concession come up. And now we've turned on the near real-time former alerts, which are modus based at 250 meter resolution, updated month by month. And you can move the time slider around here to look at whichever time period you're interested in. So again this is tree cover change month by month at medium resolution. So zooming in here and you can see there's a protected area, palm oil concession, interesting patterns of tree cover change here, of tree cover loss. And this is now being very actively used in the industry and I'll talk more about that in a minute. Switching to the other layer, you can actually see clearly here is the area that's been cleared for the palm oil license. So those are all automated algorithms that are generating automated updates running on the Google Earth Engine cloud every month. So zooming out, the last place we'll go to is Ivory Coast and the protected area here. And with the protected areas database and the tree cover loss and gain information you can start to see some extraordinary dynamics around protected areas. This is Kijimpan's e-habitat which seems to have been virtually lost. You can see exactly when that happened, how that coincides with local political dynamics of course which have been dramatic in Ivory Coast. But then you have other protected areas which seem to have fared much better. By clicking up here we can draw on the map and on the fly it will give us an analysis for the polygon that we draw, how much was lost, how much was gained over that particular time period. So it's an analytic tool as well and you can subscribe to alerts for that area by typing in your email and it'll email you an update on what's happening to that area as new data becomes available. So that's some of the basic tree cover change data and how that can be overlaid with other data sets. The flip side of the system is the bottom up piece. So this is displaying stories that people have uploaded into the system. So there's a crowdsourcing component here which allows you to submit your own story. You go to the stories tab, you can bring up the various stories that people have submitted and you click here which says submit your own story and type in a title, you can browse around on the map here, zoom in, drop a pin anywhere on that map and then you can put in where and what happened, the dates and add multimedia and other information to that and that will be seen instantly. I think most important for you here is that all the data sets in here can also be downloaded directly from Global Forest Watch. So we are designing this, this is designed for the general non-technical user but then in the background there's also all the technical stuff. So you can actually download all of these data sets in various formats and feed them into your research program here as well. So that's the end of the video. Let me just get out of that and go back to the slides. So we launched this in February in Washington D.C. and we've been quite taken aback by the level of interest, roughly about ten times the level of interest that we were targeting. Literally within a couple of days about 200,000 people around the world had gone in and looked at the system. I think we're getting up to nearly a thousand media stories now. On the social media side, within the first week, we reached 15 million people through the tweeting and retweeting and Facebooking and all of that stuff. And globally the pattern of users is quite... Well, we were happy to see how dispersed the use of the system has been. I remember it works across several languages, so it's there in Mandarin, Bahasa Indonesia, French, Spanish, Portuguese, Arabic. So this is from Google Analytics here, which is amazing if you've never played with Google Analytics with a website, but it shows us that the U.S. has had nearly 100,000 users. But then across the world we're seeing nearly everywhere else people have been looking at it with the exception of Western Sahara and North Korea, over 200 countries and territories. I took this screenshot yesterday, but the map already was pretty colorful within the first few days. But as of yesterday, it's interesting to see who's in the top 15 here and how many people have been looking at this, compared with the kind of reach we would have with the normal kind of communications where we would do. So Russia is up there with 30,000 unique... 30,000 sessions, so about two-thirds of these would be unique visitors. And then you've got France, China's coming very high up on the list. Brazil, nearly 20,000. Indonesia's number 15 at 7,000. So this is... we were very happy to see this, happy to see people in countries where forests are a very big deal, paying attention to the system and starting to use it. We've now got about a... the current return rate of current users is about 40%. So people are coming back into the system, and they're staying in the system now for five or six minutes on average, whereas around the launch, the return rate was about 5%, with an average period of use for about two minutes. So we're starting to see use consolidate and deepen. So that's where we are right now, and I'm sure you've got some questions about that, but I quickly want to talk about where we're going next. So this is a very active ongoing program, and major investments are going to be made for several years to come to continue to expand the program. So the first area of development is country data. Right now in the system, you can select a country, and you can pull up some basic information about that country, quite a lot of which we draw from the FAO, Forest Resources Assessment. We want to tailor each country's piece of this much, much more to the priorities and interests in those countries. So most people coming into this are interested in that country. We want to give them what they're looking for for that country. That's a huge, huge amount of work, but a very high priority. And then across the system is developing more specialized apps for more specialized user groups. And the one that's first out of the box that was launched in beta version about a month ago is Global Forest Watch Commodities. We've been quite amazed at the level of corporate interest in this system, particularly from palm oil industry related companies, particularly consumer goods companies and major trading companies, folks like Unilever, Nestle, Wilmar, Golden Agri Resources, and then the broader Roundtable and Sustainable Palm Oil process. Our team here has been very active on palm oil issues for several years, and we're very active members of the RSPO. And so we've teamed up with RSPO and the other partners in Global Forest Watch to experiment with Global Forest Watch Commodities, which aims to take the information in Global Forest Watch, then add commodity-specific data sets to allow companies to start to make progress in addressing the traceability and transparency challenges that they face. As some of you know, they've made commitments to what they call deforestation-free palm oil and pulp and paper and so on without defining any of those terms. And I would say, broadly still, without having defined those terms a year or so later. So we're aiming to help them address those challenges here again and the partners here. So this time you've got RSPO added into the mix. Here's a few screenshots from Global Forest Watch Commodities. So this is more specific to this region right now on palm oil. It will expand for soy and beef and so on with Latin America becoming very involved. But you've got land cover and land use data added into the system over lane with tree cover loss. And you've got the RSPO member concession data. So if you're an RSPO member, company, you have committed to share your concession maps publicly, which is actually very precedent-setting. As you may know, the quality of concession maps available from the government here right now is variable. And so the companies actually have the most detailed and accurate maps and they are now forced to share them through a resolution that they actually signed onto at RSPO. And they're using Global Forest Watch Commodities as the place to disclose that information voluntarily. So this is what's in there right now. Each of these is an RSPO oil palm concession. These are the certified ones. And by November, all of the member's concessions will be in here a lot more than you can see here right now. So if you are sourcing from the SIPEV group, you can pull up the SIPEV group's group-level company data in Global Forest Watch Commodities and see what the various overlays are telling us about tree cover loss, what land uses those are over, how many fires there are currently in their areas, and other useful information. You can also look at it at the PT level, at the company level, rather than the group level for a more detailed look. And you can overlay this in various ways, so for fire monitoring and so on. So we are in this partnership now with these companies who don't provide any financial support. What they're providing is their expertise to help us understand how can the information that we have or information that we don't have back could potentially be generated with further investment. How can that inform their supply chain management decisions as they shift to what they call deforestation-free commodities? The second example is Global Forest Watch Fires. This is going to be hopefully launched next week in Jakarta. We've been working very fast on this just literally over the last month. Because of the fires issues here, as you know, some of you know we did a lot of work on that over the last year. Some of it, working quite closely with David Gavo here as well at C4 to basically publish very simple GIS analyses, overlaying fire alerts with concessions and protected areas and so on. Very controversial. And it generated a great deal of discussion amongst the companies and the governments and the enforcement agencies and so on. So we're now taking that and adding some additional information to it and creating a tailored fires version of Global Forest Watch. I'll just show you a few screenshots from that. This is actually the current Global Forest Watch, the general system which has fire alerts from the last week, the last three days, the last two days and the last day. So that's what's on the general system right now that's been out there since February. The more specialized system looks like this. So it's a similar configuration here, but you can also select only high confidence fires by clicking here. We're adding the NOAA fires data. And the maps is focused on this region. It comes up with the peat layer on it as well, which is not on the general map. And then you can click on any of these and bring up other fire-relevant information. So here are the wood fiber plantations, for example. So quite similar to the general system, but with some added bells and whistles. If you click on here, then you get an automatic summary analysis and figures of the fires over the past seven days. So this is analysis that we were periodically publishing and producing manually. So it now automated this. So it's updated every day on the system for everyone to see. There are about seven different figures like this, which... Ooh, I keep pushing the wrong button. Hold on. Which... So this is which districts have the highest number of alerts over the last seven days. I mean, here we're really hammering home the message that fire work in Indonesia needs to be concentrated in this area. Riau is half of the alerts, and three or four districts and sub-districts are where nearly all of those alerts are. That seems to have gotten through into the policymaker process, policymaking process, and various ways of presenting that information. Very, very simple and easy to understand for those who are working on this. We're also adding some very attractive graphics. This is actually on the system. This actually is an animated map of winds and wind speed. And you can see, particularly here, the patterns of wind and so on moving over Sumatra. So you can see, for example, here, that Singapore was getting wind coming up from the south. I looked at it yesterday, and actually the wind was coming that way. So you can see in real time this is updated every four hours. So these are useful tools for the situation room and real-time response from policymakers who are trying to figure out what's going on on the ground. There's also a Twitter conversations tab, which brings up any Twitter conversations, and you can see those automatically that refer to any specific hashtags related to the fires in English or Bahasa Indonesia. When the fires get serious, there's a massive amount of activity on Twitter about that, particularly here in Indonesia. So we're putting all that together, and this is going to be launched hopefully next Wednesday in Jakarta with BP Red and some involvement from some other ministries as well, hopefully including Ministry of Forestry. And you're all invited. We'll send the details over as soon as they're finalized in the next day or two. The system also includes an effort to increase the response time of the government. So we take the fire alerts. We prioritize them based on the data about each alert that's in the system and automatically can send SMSs out to the Kapaladesa, the Chamats, the Bupatis, the Mangala Agni, the various firefighting response crews and so on the ground. So we've put that together. We're now in discussion with the government about how they want to roll out an automated SMS communication process, which I believe would cut the fire response time by some agencies from about 35 to 40 hours, which is almost useless, of course, to something more like four or five hours for alerts coming into Jakarta and then communicated with orders to fight fires on the field. So that's quite a simple set of technologies that we can put together to do that. So that's already up and running. We're waiting for the government to look at how they want to integrate that into their management systems. And we will roll it out ourselves directly with some of the companies and so on, probably with the alerts coming directly from our system in the U.S. This is an example of a sample SMS alert that you might receive overnight with the long of a fire and so on. And we can obviously tailor those to say whatever would be most helpful. We're working with Google Earth Engine to look at how we can add information about how to access that particular fire and what water sources there might be nearby and automate that as well. So that's where some of the big data map stuff can potentially come into play for very rapid response across very, well, basically across the whole world potentially, but certainly across an area the scale of real at very low cost. The other piece, the final piece I'll talk about here, is beginning to work with ultra-high-resolution satellite data. This is extremely challenging because of the size of the files, because of the cost traditionally of this data, and we're just not used to handling this type of information, and most researchers are not, because it's usually the provenance of private companies and intelligence agencies. So we have a partnership, a new partnership with Digital Globe who owned four ultra-high-resolution satellites over Indonesia twice a day. And they are now sharing data with us, tasking those satellites as fires are detected and trying to collect images of the fires, and those are all going to be shared on Global Forest Watch fires as well. I'll show you what they look like in a second. The interesting thing about this partnership is that we've been able to bring down the cost to a level that we can afford, that's not still not cheap. And more importantly, the licensing arrangement we have allows us to share these at full resolution with you, with the government, with universities, with anyone who's working on these issues. So we're going to make all of this freely available. So David Gavo and his team and others here will hopefully be able to take this imagery and do some important research with this. So we have basically a perpetual license that allows us to share it with anyone who's working on these issues. And I think that's the first time that's happened. So here is what you see on the digital globe interface. They've captured this image, and you zoom in, and they've put it in false color here. Basically, you can start to see here at zoom level 13 an area that's burned and an active fire and smoke plumes. You can see which way the wind's blowing, and you can see precisely where the fires are burning. But we can go much better than that. So we can do them into 15, 16, zoom level 17 here, 18. And here you can start to see it's not so good on the screen. There you go. So this is the maximum zoom level, although this isn't full resolution because of the file size and the screen here, right? But you can see here individual palm oil trees, precise burn scars. And we also have access to their archive so we can start to try to see the history of land use in each of these areas. So this is hopefully a boon for speeding up some of the research on the dynamics around the fires and could have interesting applications in the law enforcement process because you've got a precise time stamp and a precise lat long for every point on these images. So this is already being made available to the government on a day-to-day basis as we collect it. Oh, sorry, one more zoom level there. Another image there. This is more real color. So you see here active fires, an area that's burned, again in a landscape of palm oil and scrub. And if you know the context on the ground as well or you go out there and look, you can learn a lot from this. And then we can render them in this type of format. So here's the fire, here's the scar, we locate the scar for the policy makers precisely on the map, provide them with this information and then it's over to law enforcement to go out and see whether that's helpful. Coming soon, we're working very actively right now with the Woods Hole Research Center at the University of Maryland on a global carbon flux and carbon stock layer for the maps at 30 meter resolution. We'll be launching the tropical piece of that at the Lima COP, so we may well include presentations on that at Landscapes Day. And the global, the full global coverage of that will be launched at the Paris COP next year. Global Forest Watch Biodiversity with Eric Dynastine who used to be at WWF, the WWF Chief Scientist in DC as one of the leading global experts on patterns of biodiversity and biodiversity data sets around the world. So he's now part of our team and we're designing and developing Global Forest Watch Biodiversity to take the biodiversity data sets and present those in a way that support improved decision making around infrastructure investment, link it in with the fire stuff, and so on as well. Go, no go zones for the commodities industry, for example. Terra Eye, which some of you may know was developed by SEAT, I think I mentioned that earlier, did I talk about Terra Eye at the beginning? No. So Terra Eye, SEAT Columbia have developed a wonderful near real-time alert system for vegetation change, not just forests, that we're now integrating into Global Forest Watch and we're funding them to take that to the pan-tropical scale and increase the frequency of updates through a major partnership with SEAT and the CG system. And I also showed you some, I just showed you some of the high-resolution work, so I think learning from our experience here with Digital Globe as that develops, we want to potentially apply that to many, many other problems around the world, such as detection of illegal logging across the Congo. Several governments there have indicated their interest in us supporting their efforts to enhance enforcement around detection of illegal logging roads, snaking out of concessions and cutting blocks and so on into areas where they shouldn't be. The high-resolution information allows us to do that kind of thing. It's not just tropical, it's global, we've actually got a piece coming out, maybe today on our website on the Tarsans and Canada and the Excel pipeline, which is a huge, huge story in North America, the biggest environmental story ever just about in North America, I think. Looking at deforestation, forest conversion linked to Tarsans' development in Canada's very special boreal forest, basically about 2 million acres of boreal forest has been cleared over the last few years. I can't remember the time period since 2000, I think, for Tarsans' oil development and some of its very precious caribou habitat and so on. So more and more work of that kind. I'll stop there, take some questions. There are lots of challenges and constraints and shortcomings with this work as well that we are quite familiar with and very interested to get your perspective on those. And we look forward to much more collaboration on this with C4 in the months and years ahead. Thank you. Thanks, Nigel. Does anyone have any questions? Can I ask people to stand up and state their name and nature of their research? Thank you, Nigel, with you. It's a very, very excellent presentation. Perhaps this is a great effort, really, that maybe scientists can take a lot of benefits from the existing database or information like this. Perhaps also important to have a real clear data dictionary. Data dictionary, because, for instance, the definition of forest is not necessary all people talk the same thing when we talk about forest. Inside the country also the different definitions for different countries. For instance, Malaysia defined forest quite differently from how Indonesia defined forest. I'm wondering how this kind of global database, if you want to say, accommodate different kind of definition of forest. To me it's like to have a clear and strong data dictionary is very, very important in this database. The second one is talking about the update because this also covers the dynamic of the data. It's very nice about the fire. But sometimes the dynamic of the data is also about the government. I mean, sometimes, for instance, government already cancelled this concession. It's no longer exist. You talk about the data, but actually government already, for instance, cancelled this concession. I don't know how often you actually update this kind of data. It's very, very challenging this kind of dynamic. Thank you. Thank you, Paheri. Paheri and I did a wonderful panel. When was that? It seems like a long time ago. I think it was last week in Jakarta on the fires at the Jakarta Forest Correspondents Club. And that was a great discussion there. In terms of the data and the definitions, of course, this is a huge issue. I think Ken McDickin at FAO, who used to be here a long time ago, likes to say he's cataloged something like 2000 or 3000 different definitions of forest and what forest means. So this is a huge issue. A couple of things. We're very, very transparent about what definitions we're using and how the systems are derived. So you click on the info bubbles, or go to the data page, and it takes you right back to the original citations and summaries of that research. So people can see if they want to what it is that we're working with here in great detail, and then work with the data directly themselves. So we aim to be very, very transparent about what we are using. There is a modification to the system which is in the process of being coded right now. I don't know if, Paul, you know when that's coming into effect that allows you to change the definition of forest in Global Forest Watch and that will change the analysis and what you're seeing on the screen. So we're able to adjust whether we define forest as 10% tree cover, 25% tree cover, 50%, 70%, 75%, something like that. So there's basically, like on an iPhone, you can change the settings with little sliders. We'll have a default setting which corresponds to the FAO for our definition of 10%. But if you know a lot about this and you want to start fiddling with the settings on the system, you can go in and start to do that and see how that changes the results. So that's, it's very constrained by what's in the data and the resolution of the data and how far we can go with that. But as time goes by, we'll be able to do more and more of that. So that hopefully allows more people to work with the system with a definition that's closer or closest to what they're currently working with. In terms of updates, the concession data here in Indonesia is of course extremely challenging and we're getting into that in quite astonishing detail with the government right now with various agencies as they recognize the value of putting this together with the data sets that they have, many of which are not yet public. And yes, the situation is very, very challenging. On the fires, we're starting to dig in deeper at certain parts, certain places on the map where there are repeated and very serious fires. So who is actually there on the ground? Which company is it? Is it Palmwell? Is it Polkwood? We're seeing areas that were on our map as Polkwood are actually clearly Palmwell. We can start to see that with the high resolution imagery as well. The maps in the Ministry of Forestry versus in the province are completely contradictory even on the basic type of licenses that are issued for some of these very significant areas where there are fires and a lot of conflict on the ground. So we're figuring out how to deal with that. For now, we'd like to continue to use the national government's published databases but are aware that there are some serious inaccuracies with that as they themselves have said. We may try to move towards showing the various different databases as we get access to those so that people can also see that this is very problematic and that will then create more impetus for resolution of these discrepancies by the government. And as you know, there are people in the government working very hard to try to resolve those discrepancies right now. So the more resources and the more support that they can get broadly, the more rapidly that work will advance. Hey, Nigel, thanks very much. Very enlightening talk. I just wonder whether you've been here for a long time and been working on this issue for a number of years. What would be new in this kind of exercise, especially on the approach? People keep on talking about bottom-up, bottom-up and then forget about top-down, which is still important. Is there anything new that we can really push this approach? As we are talking here, fire is raging somewhere else and a lot of new institution is in place and what would be the appropriate entry point to deliver this kind of information? Talking about Indonesia specifically. That's a tough question. I'd be interested in your answer to that more than mine. What's new here? So trying to get closer and closer to real-time, policy makers, particularly very high-level policy makers, ministerial level, they need to know what's going on. They need to know what's going on now. They're making decisions on the fly, very big decisions, and they're very constrained by poor information about what's happening out there on the ground. So the message we get from them is the more precision you can bring to this, obviously, the better. But the closer you can get to real-time, the value of that, the message is almost that sort of exponentially increasing value the closer we get to real-time. Data that's several years out of date by the time it's been published and peer-reviewed obviously has a value for long-term policy making, but not responding to the kinds of rates of change and dynamics that we see on the ground in Indonesia right now. So what's been the impact of the moratorium up to a month ago? That's what policy makers in Jakarta want to understand. Where and why are there so many fires again in early June? Why are they in the same places as they were in February and last June? And who actually is responsible on the ground for that? What is the response of the various agencies that are trying to respond to those fires in practice on the ground? Policy makers in Jakarta do not have that information to hand right now. So giving them up-to-date information that's closely tailored to what they need and what they understand is extremely important. In the business community, the confusion, the level of confusion right now is high. If any of you who are working on palm oil and the corporate standards around palm oil would know that you've got RSPO, ISPO, the Malaysian's involving standards, on the production side, RSPO+, and then on the consuming side, you've got the consumer goods forum, all the big consuming companies globally who buy this stuff saying, we want deforestation free palm oil. What does any of that mean and where is that going? And is that informed by any good information and data again? And they want that in near real time because they're making buying decisions. They're involved in business transactions day-to-day across literally a company like Unilever literally is sourcing from 6 or 700 palm oil mills across Indonesia and Malaysia day-to-day through Wilmar and Cargill and suppliers like that. And they want to shift those patterns of buying to areas where forests are being managed better rather than worse. So there's a very, very direct demand from extremely powerful and resourced stakeholders who can make changes very quickly but they do not have the information they need to do that. So we are racing to pull together the existing data sets and new data sets and task satellite systems and so on to see if we can start to answer those questions in a way that would inform those buying decisions and then further upstream inform the investment and banking decisions and the equity investment decisions that are being made around the world. So very direct impacts on Indonesia very direct market signals from Indonesia if that can be done effectively. But it's obviously extremely challenging but I think the only way we can do that is through these kinds of tools and systems and then we have to combine that with information from the field. So the crowdsourcing side of this is the other piece of the technology which I just barely touched upon. Indonesia has tremendously high cell phone and even smartphone penetration, even in rural areas. How can that be used to gather feedback from the field including the things that we can never see from space like labour violations, child labour, slave labour, bonded labour and so on which these companies are equally committed to eliminating from their supply chains. Clearly it's possible in theory to start to do that so we're looking at those aspects as well but that's what we need to try to bring together here and I think what we see in the evolution of the technology is that this will get easier and easier over the next two to three years and then the tremendously interesting part of this is what do the institutions here then do in response to the opportunity to use this information to make better decisions. That's where, that's where, that's your bed. Hi Nigel, nice to see you. I'm Celine, I'm an intern here with Kristoff working on palm oil and I was curious about the response of neighbouring country governments such as in Malaysia and Singapore and what has been their response and their desire to be part of this, thank you. Celine is from Singapore, right? Yeah. Well, the Singaporean response is the most interesting and the most active. Malaysia's response is also very interesting, I think. They're very different from each other. In the context of ASEAN the five governments that work together on the haze have biannual meetings and we've been party to some of those discussions sort of up close because of the work we've been doing on this. The significant development there last year was Singapore pushing the five governments to commit to share more accurate concession data with each other. Malaysia and Indonesia but Malaysia actually more than Indonesia I think are very strongly opposed to doing that publicly although Indonesia actually already has published a lot of information publicly to their credit Malaysia has not at all I don't think. So there's an interesting discussion going on about transparency. Those in other countries those in Singapore suffering from the haze want to know who's responsible on the ground so they've been pushing that piece of this on the transparency which obviously would be very helpful we're pushing that very strongly as well not primarily for the reasons Singaporeans are but because we believe that simply having more accurate information of this kind in the public realm is key for coordination of across agencies from the federal to the local level with Indonesian civil society with research institutes like C4 who want to do more detailed work on this and there are many in the government here who are sympathetic to that. The other piece of the response more clearly from Singapore is the New Transboundary Haze Pollution Bill which had its first or second reading in the Singaporean Parliament last week I think very recently which now imposes actually much higher penalties than the first version of the bill did several million dollars potentially and finds for companies who Singapore finds guilty in court in Singapore of generating haze which harms Singapore so they can take Singaporean based companies to court they can also take Indonesian companies to court this haze also coming down quite a bit from Peninsula Malaysia that affects Singapore as well depending on the wind perhaps most striking in the legislation though is it also allows for civil cases to be brought there are criminal cases but then there are civil cases so Celine as a Singaporean citizen could herself bring a case against a company that is polluting the air she is breathing in Singapore or hypothetical example a hotel owned by her sister or the national airline or any company like that could bring a civil case against companies that affect them and we saw last June very dramatic effects and the most dramatic effects on the Indonesian economy right 2 billion dollars in impacts more or less is estimated possibly hundreds of millions of dollars in impacts on the Singaporean economy as well due to fewer tourists conventions being cancelled and so on the situation on the ground there was extremely serious even more so in real but so Singapore is basically in my opinion in a very tough situation on this and is doing whatever they possibly can to try to bring pressure to bear on the situation they are also very active behind the scenes through diplomatic channels but the diplomatic relationship between Singapore and Indonesia is extremely complex as you all know so it is not clear that there is very much progress being made there Singapore has offered funding and so on for work on the ground how much funding would potentially available is very unclear on the one hand and Indonesia I don't think is particularly enthusiastic about more Singaporean funding to do work related to the fires on the ground there has been an ongoing effort in Jambi with Singapore research funding maybe C4 has been involved I'm not sure on these issues so that's kind of where it stands I can say right now that Singapore is extremely extraordinarily nervous about a repeat or an even worse haze crisis than the one that happened last June because of the high likelihood that we'll have an El Nino extended dry season and worse fires this year than last year and they and then it's just a question of wind direction it's really just a question of wind direction of how hard they hit in Singapore but as we always say the real issue is here on the ground in Indonesia where many more people are suffering and paying a much higher price for this than the Singaporeans but it's bad enough there I don't know if that answers the question Sleen is there anything else you are curious about there involved in our IGFW we are we are in we have a very good discussion going on with the Singaporean government about this they're very very interested in the work that we're doing I'll be so whenever I go through Singapore now it always involves meetings with the Ministry of Environment the National Environment Agency and other agencies there to update them on our work which is all publicly available as well they have some quite good technical capabilities there as well with their various research institutes so they've also been able to enlighten us on some of their research findings which generally don't seem to be made public and some of the modelling and haze projection work that they've been doing they have an interesting partnership with the IBM co-laboratory on haze modelling and haze projection they've been doing some really impressive work at the Singapore Management University on analysis of Twitter and Twitter sentiment including related to the fires as well as in Singapore on the ground in Singapore they're monitoring about 300,000 Singaporean and Indonesian Twitter accounts to help monitor what's happening to fires to try to give them more real-time information so we're very intrigued by how that can help inform a real-time response to the situation on the ground because people are tweeting as it's happening the last thing I'll say on that is for example we have no air quality data at all from Sumatra none is collected I don't think any is collected if it is none of it's made public but we have very extensive air quality data from the Peninsula, Malaysia and Singapore so that's all going to be on global forest watch fires could Twitter and monitoring Twitter actually be a proxy for air quality monitoring is something that is an interesting piece it's not as good but it will tell you that there is haze that you can hear right now as people are tweeting about it in various ways it would be the only source of that information other than what we might gather with remote sensing Hi Nigel, I'm Grace Wong with the livelihoods program my question relates to kind of the data that you have I mean this is a tremendous effort and a great resource but a lot of the data that you have are data that's designed to be mapped so you have forests the concessions, the protected areas but there is a whole slew of data that's really critical to forests that are not easy so things like tenure things like the adat thing forests and that's vague and uncertain and as this resource becomes used for decision making and you have a suitability mapper there on your commodities website I wonder how you see those risks and what might be some ways to kind of manage these types of risks that's a brilliant question there is a tab on Global Forest Works called the people tab and we are in the process of adding data to that tab right now so the first edition was made a couple of months ago when we added the first local people's tenure data to that which is community managed forests for Cameroon which we happen to have because we've been working on mapping in Cameroon for 15 years the government and NGOs and others there so if you go into Global Forest Works and you click on the people tab and you pull up that and go to Cameroon you'll see the first data set on Global Forest Watch for that on July the 24th we and Rights and Resources International Andy Whites Group will be publishing a report on global mapping of traditional, tenure indigenous peoples lands mapping and so on and to coincide with that report we'll be adding we'll be adding a bunch more tenure and traditional peoples maps and data layers to this forest Australia, Canada Australia, Canada Brazil, Guyana, DRC Liberia and a couple of others I can't read my handwriting so that's coming up on July 24th we're in discussion with Aman about adding their data set here in Indonesia which is a very interesting data set and I think that will happen eventually but as you say this is extremely complex and sensitive first of all those groups may not want that information to be published because it's useful in negotiations so how you actually visualize that the boundaries of those areas is a challenge if it's there are indigenous territories in Brazil with very clear and well defined boundaries by the government so that's easy to put up it's already out there and so we're in the process of adding that for countries like Colombia and so on you've got a Hido maps for Mexico and stuff like that but when you get to a place like Indonesia where there's tremendous conflict over where these lands begin and where they end and which particular rights are associated with which pieces of those landscapes the visualization of that is a huge challenge so we're basically looking for very creative ways to address those problems such as showing the boundaries as rather indistinct regions on the map that don't disclose exactly where that area is but make it very clear for example to a foreign investor or a company like Unilever that might have palm oil mills near there that are supplying it as deforestation is moving into that area make it very clear to them that there are people here who broadly claim this area as theirs and are in conflict with these companies that are your suppliers so in the commodities space this gets very very interesting and the companies are very keen to see that type of information and be able to work with it because the risks associated with that for them as investors are huge in terms of brand and reputation so this is a very high priority for us and we're moving through it as fast as we can but there are additional complexities and then of course there's stuff that simply there's not data available for what's very inspiring is the extraordinary I think expansion of participatory mapping efforts going on around the world so those databases are growing rapidly and so we're committed to having that up here in a format that's useful that respects the the opinions and demands of those communities completely and makes that as transparent as possible for everyone to see it and doing that as fast as we possibly can that's where we are right now on that if you have any suggestions or input we would be very happy to hear them sorry just a quick my name's Annie postdoc here at Seafall we'll hear a bit more about the bottom up approach that you have do you filter the stories that come through and can you distinguish between the different sources and do you set scientific articles I mean is there a way to to get to the sources and distinguish between that so the crowd sourcing piece of this is very rudimentary right now what we did what we decided to do was simply create a crowd sourcing function in the system and then see what happened we had no idea what would happen and the reality is that actually there's been very little use of that very few people compared with hundreds of thousands of people who have gone in and looked at what's on the system almost nobody has uploaded their own information you can go in and upload anything you want so if you've done a research paper at this protected area in Sulawesi you could go in and drop a pin touch that and everyone going in and looking at Sulawesi would see that research paper no one's doing that if you are APP and you want to push back or push back on accusations being made against your company you could go in and put that information in the system but they're not doing that so we're very intrigued by why there hasn't been more uptake of this and we've got several colleagues who are working hard to understand that learn from the best out there on crowd sourcing to see what the next generation of this should look like for Global Forest Watch in terms of the curation of what comes in from the crowd sourcing right now if you upload something it's instantly accessible on the site we'll remove the standard languages we'll remove anything that's inappropriate or abusive but we haven't had to do that yet so basically anything we go up it's very clearly identified as not endorsed by us or our partners it's there as crowd sourced materials so take it or leave it if you're looking at that on the system so that's kind of where we are right now I think perhaps as we get into more specialized applications of Global Forest Watch with much more intense engagement with specific user groups for example around palm oil and slave labor then you might get into a much much more interesting and much more active dynamic with these kinds of tools looking at where you've got potentially some real incentives for people to want to share that information I've been using a lot the state of the forest for my work and usually the public figure in that report is kind of different from the official government figures and especially for example the impact of the fires in 1998 and it's always an analysis in that report that claims that the government like covering up the actual incidents of area that affected by fire or any other public figures and you mentioned earlier that you will be closely working now with the government so do you think that would change? I mean now your figure, public figure would be more likely similar to the government official figures I need your opinion on this and the second one would be possible to analyze to analyze the forest cover resulted from rehabilitations or any other in our restorations activities because the government Indonesia for example has spent a lot of money in this and to analyze effectiveness has been so difficult because we have a lack of this actual forest cover that actually increase from that effort so would be able to analyze from that and my last questions do you think that Indonesia has been successfully reducing its emissions because we have a lot of government 26 or 41% Great questions and I'll start with the restoration one because we haven't talked about that yet so we have a sister initiative to global forest watch at WRI called the global restoration initiative we're very active on that it's a whole new area of work for us that we're investing a great deal in and many others are moving into that area as well we've been in some discussions with the Indonesian countries on that and global forest watch is the tool that we aim to use to measure success of restoration the challenge is the resolution of the data sets and so on the resolution of the data the satellite data for doing that so picking up early stage restoration early stage growth regrowth is difficult because it's not, it's very sparse it's hard to distinguish from pasture crop and picking up fine scale like riparian restoration or agroforestry system is difficult as well sometimes difficult to distinguish from other land uses so those are technical questions that we are very actively working on and the higher resolution data and combinations of high and medium resolution data could help us greatly to achieve advances in that area what's on the system right now is from the from the Matt Hansen Google data set I showed the pink of tree cover loss you can click on another button and you get a blue which is tree cover gain what you see very clearly across Sumatra there is expanding plantations you see the blue of the plantations big square blocks of blue coming up very dramatically across that landscape very nicely we also detect restoration of larger areas as they mature but won't as I said won't pick up the early stages of restoration so on so there is a technical remote sensing challenge there related to the confusion with other land uses and the resolution of the data that you require and the cost of handling those higher resolution data sets I expect to see there are advances in the technology over the next two to three five years that should solve that problem I would say somewhat boldly but based on what we know is going on in the industry basically we expect to have near daily very high resolution data available and you wouldn't need to look at that across the whole world every day of course once every year once every couple of years you could do a systematic analysis or in certainly the landscape level with that kind of data it's very very useful we're already talking with some of the companies here who are as you know well one company in particular that's committed to large-scale restoration efforts around their plantations and how can they monitor and report on success with that in a credible way is a question that they're asking us in terms of the fire data and discrepancies in what our analysis and your analysis and the government's analysis might say clearly there are methodological differences there and there are political interests as well one thing we did last year that was very helpful we thought at least for us in understanding and validating the medium resolution alerts the alerts that come up on Global Forest Watch are actually at one kilometer resolution from the MODIS system derived from MODIS by NASA many people said they're not accurate you're saying we've got all these that's not true so we tested that with rapid eye data we spent nearly $100,000 buying rapid eye 5 meter data post facto and did a fire scars analysis and C4 has that data we gave it to C4 as well to David here and that showed 97% accuracy of the MODIS data so we were very happy to see that we published that also in those maps along on our blog and now we've moved up to the 50 centimeter resolution data even finer scale so for specific areas where you've got really serious issues you can do you'll be able to do an incredibly precise fire scar analysis to the tree I mean this tree burned that one didn't this was the precise boundary and with the archive depending what's there from Digital Globe you'll be able to see a lot of the history of fire in that place as well potentially so that will help basically our approach is everything that we're doing on this is made public it's published if others want to challenge the results they need to publish their material and their methods and then we can have a scientific discussion about that if they want to challenge it without doing that it's not a very useful discussion and on the emissions I don't think I've got much to say about that I think there are others here who know much more about that than I do it will be the work we're now we're now doing with the carbon flex and carbon stock at high resolution of course will help inform that and we look forward to working very very closely with the MRV team here at BP Red the team that's just been set up under Ibu Hade to see how what we're doing globally on that could be taken to even higher resolutions here and inform policymaking and public opinion and media work on this here in Indonesia okay thank you unfortunately we've run out of time so I think we're going to have to end there but thank you for your interest and thank you Nigel and WRO