Global Forest Watch: Using Open Data to Save the World’s Forests

Video thumbnail (Frame 0) Video thumbnail (Frame 716) Video thumbnail (Frame 1737) Video thumbnail (Frame 2683) Video thumbnail (Frame 3637) Video thumbnail (Frame 4845) Video thumbnail (Frame 5608) Video thumbnail (Frame 6312) Video thumbnail (Frame 9252) Video thumbnail (Frame 9931) Video thumbnail (Frame 10535) Video thumbnail (Frame 11063) Video thumbnail (Frame 11805) Video thumbnail (Frame 12342) Video thumbnail (Frame 12894) Video thumbnail (Frame 13679) Video thumbnail (Frame 14163) Video thumbnail (Frame 14967) Video thumbnail (Frame 15728) Video thumbnail (Frame 16627) Video thumbnail (Frame 17486) Video thumbnail (Frame 18021) Video thumbnail (Frame 19121) Video thumbnail (Frame 19782) Video thumbnail (Frame 20607) Video thumbnail (Frame 21331) Video thumbnail (Frame 24827) Video thumbnail (Frame 37991)
Video in TIB AV-Portal: Global Forest Watch: Using Open Data to Save the World’s Forests

Formal Metadata

Global Forest Watch: Using Open Data to Save the World’s Forests
Title of Series
Part Number
Number of Parts
CC Attribution 3.0 Germany:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Release Date

Content Metadata

Subject Area
Global Forest Watch (GFW) is an interactive online forest monitoring platform designed to empower people everywhere with the information they need to better manage and conserve forest landscapes. Thanks to open data, GFW is able to do the following: Monitor when and where forests are changing. NASA’s freely available Landsat and MODIS data has allowed hundreds of scientists and researchers to develop innovative solutions to monitor landscape changes. Algorithms are now used to process and analyze this remotely-sensed data to show when and where forests are changing with surprising precision and speed. Understand why forests are changing. Open data showing boundaries of land allocated for specific purposes, such as commodity production and conservation, as well as land management, allows us to understand why forests are changing. Are trees being cleared for palm oil? Are certain swaths of forests still standing because they are managed by indigenous groups? Gauge the significance of deforestation. Additional open data provided by research institutions, governments, and others is used to understand the implications of deforestation on biodiversity, climate change, and provision of ecosystem services. For example, was a recently clear-cut area of forest home to endangered species? Was it a carbon rich primary forest? Spark further innovation. GFW’s open-source code and APIs allow others to leverage GFW’s analysis tools and open data to create additional forest monitoring and management tools.
Web 2.0 Forest Mobile app Computer animation Sine Associative property Open set Resultant
Area Implementation Divisor Software developer Digitizing Multiplication sign 3 (number) Water vapor Database Open set Equivalence relation Product (business) Forest CAN bus Computer animation Integrated development environment Forest Website Quicksort Computing platform
Mathematics Computer animation Bridging (networking) Multiplication sign Forest 1 (number) Group action Traffic reporting Resultant Computing platform Product (business)
Satellite Arithmetic mean Mathematics Computer animation File format Online help Cycle (graph theory) Group action Theory
Group action Process (computing) Computer animation Forest Bit Group action
Ocean current Freeware Open source Divisor Code Decision theory Electronic program guide Maxima and minima Internet forum Forest Information Associative property Exception handling Information Military base Open source Electronic mailing list Knowledge engineering Group action Euler angles Open set Forest Particle system Data management Computer animation Internet service provider Blog Figurate number Quicksort Physical system
Tablet computer Forest Computer animation Different (Kate Ryan album) Internetworking Time evolution Forest Multiplication sign Surface of revolution Connected space Neuroinformatik
Area Group action Email 1 (number) Mathematical analysis Shared memory 3 (number) Group action Web 2.0 Process (computing) Computer animation Forest Automation Computing platform
Email Metropolitan area network Context awareness MUD Information Mapping Line (geometry) Digitizing Source code Maxima and minima Special unitary group Emulation Forest Sign (mathematics) Wave Casting (performing arts) Computer animation Forest Personal area network Right angle Summierbarkeit Wide area network
Forest Manufacturing execution system Computer animation Information Multi-agent system Mathematical analysis Combinational logic Mathematical analysis Special unitary group Summierbarkeit Hypothesis
Area Metropolitan area network Workstation <Musikinstrument> Maxima and minima Coma Berenices Disk read-and-write head Demoscene Value-added network 2 (number) Forest Process (computing) Multi-agent system Physical law Summierbarkeit
Area Forest Pay television Computer animation Multi-agent system Summierbarkeit Area Product (business) Product (business)
Area Metropolitan area network Pairwise comparison Cone penetration test Mapping Weight Menu (computing) Coma Berenices Arm Area Value-added network CAN bus Pointer (computer programming) Explosion Computer animation Personal digital assistant Term (mathematics) Electronic meeting system Pairwise comparison Summierbarkeit Traffic reporting
Metropolitan area network Computer animation Weight Workstation <Musikinstrument> Information Grass (card game) Open set Pairwise comparison Summierbarkeit Open set
Forest Mathematics Mathematics Computer animation Information Forest Emulation Product (business)
Area Mathematics Forest Context awareness Computer animation Strategy game Personal digital assistant Workstation <Musikinstrument> Right angle
Area Mathematics Computer animation Lecture/Conference Personal digital assistant Workstation <Musikinstrument> Mereology Summierbarkeit System call Field (computer science)
Forest Metropolitan area network MUD Computer animation Personal digital assistant Forest Execution unit Reduction of order Line (geometry) Reduction of order Wide area network
Area Forest Lecture/Conference Multiplication sign Electronic meeting system 3 (number) Normal (geometry) Menu (computing) Error message
Level of measurement Service (economics) Open source Code Civil engineering Multiplication sign Workstation <Musikinstrument> Source code Device driver Solid geometry Product (business) Formal language Mathematics Conservation law Endliche Modelltheorie Descriptive statistics Predictability Area Information management Standard deviation Validity (statistics) Building Forcing (mathematics) Software developer Binary file Cartesian coordinate system Computer animation Query language Network topology Pattern language Block (periodic table)
Axiom of choice Satellite Presentation of a group Existential quantification Multiplication sign Decision theory Source code Workstation <Musikinstrument> Mereology Data quality Dimensional analysis Derivation (linguistics) Mathematics Bit rate Meeting/Interview Different (Kate Ryan album) Forest Endliche Modelltheorie Error message Vulnerability (computing) Physical system Area Mapping File format Software developer Moment (mathematics) Sampling (statistics) Bit Flow separation Data mining Process (computing) Chain Website Cycle (graph theory) Asynchronous Transfer Mode Web page Point (geometry) Metre Service (economics) Open source Image resolution Graph (mathematics) Product (business) Prototype Punched card Dependent and independent variables Forcing (mathematics) Projective plane Polygon Set (mathematics) Vector potential Computer animation Query language Personal digital assistant Statement (computer science) Pressure
after well done as you the good
afternoon and evidence Alassane CTO and founder of the solid the result is that the agency we designed and developed we design and developed a web and mobile applications that of associations baby eyes interactive maps and all
sorts of the joke digital products have focused on environment social development open that open that up and transparency every so it is a founding partner of Global Forest Watch which is a product I'm going to 2 % to today and we were responsible for the design and implementation of the main website and most of the tools in the in the platform also today implementation of the again this platform has been helping preserve forests for 2 years now and the truth any of you know it before but I'm going to try to give you a high-level overview not technical but I'm happy to answer any technical questions afterward so 1st of all white
white white Global Forest Watch uh I think this phrase says is 0 50 hectares of forests are lost every minute of every hour of every day of every year during the a database of this conference and area quot equivalent to 40 times the city of Bonn at will have been lost in forests forests are fundamental for millions of people's lives they are fundamental for climate change they affect biodiversity water and air quality and they pressured by economic and financial factors before a
Global for Watch when we when we started thinking about creating this platform there was a huge that the gap and that that was not timely reports about forests came out every 5 or 10 years they were not comparable because they were made with different methodologies they were disparity they were inaccessible sometimes not accurate many times they were unreliable because governments don't want to look bad and and kind of change the results they were expensive and they were too technical unity GIS expertise this product is aimed the bridge that
gap would give every week we aim to bring together science policy and citizenship using technology and design what is
that about I'm talking about I'd like to to summarize it in these 3 phrases that must be available and that that we've made up a big the big progress in that we have terrabytes of that that every they knew that they generated from mobile phones satellites from everywhere but that that must be usable in that it might be must be in a format in a way that can be accessed and used most important of all that there must be used is that is not used to to to create the impact change then this a that probability doesn't mean anything we try to to our
theory of change around around how to make that that help to deliver that that drives the change is is summarized in this in this cycle were were available that can be combined the great
insights those synsets can move to actually that action creates dialog and the dialog generates the need for new data or that that the drive to to to create new that to believe 1 of the things that
needs to happen we think is we need to do the job from matters of fact to matters of concern because people actually use that that when that that that is around a concern facts and truths are not as important as comfort as concerns going to show you how Global Forest Watch aims to go from matters of fact to matters of concern but 1st I have to tell you a bit of a Global Forest Watch what it is set of
tools that includes an API that the association probability of is relevant information delivered by public API it costs over 165 datasets uh this figure gross every month a lot of them are global and near real-time and they come from uh right now 150 different other providers the vast majority of them are open that up except for a few notable exceptions is action-oriented is not another that the particle is not that that that we try to empower people everywhere and all kinds of all sorts of people with the information they need to actually create impact the information and the tools to to actually manage forests landscapes of forest resources it brings together science and that knowledge it uses technology to create there to be provide precise information that can guide policy and decision making is not that a person is not in a world that is that that can be analyzed that can be combined and that can be used all it's what tools are open including the open source code bases for all the tools and create the coalescent that the because free and simple to use is collaborative is open to contributions data that date code is open to pull requests and is open to received data from all the other actors and users can also contribute by sharing information by creating stories by uh validating forest from the ground a this discussing the blogger discussing forums and most of all
is a partnership this is a list of the the current list of golfers watch partners swallow started we were 10 partners we 19 o 2 years after so I think this is 1 of the big factors of success of these of these of this tool that all of us watch was coming and it has been removed for forward by the World Resources Institute and they are the the the the the big promoters behind we say this is that
forests transparency revolution why because for the
1st time we have a 2 where anyone in the world with an internet connection can when and where the 1st patient is happening from the computer the tablet or the mobile and the candidates with days of difference between 1 the deforestation happen and when they are still looking at in the in the fall automated we can go
from detections from acetylate processing analysis and sharing on the web in a matter of hours for some tools like fires for some the Pacific fires this is less than 3 or 4 hours for some more complex ones like a forest alerts it can be 1 is action oriented
it takes that there were people can see we don't aim for people to come to a platform from the can create alerts that are rooted to the e-mails might as a or even a creative with that's going to trigger action whenever I knew deforestation in the palate is detected in the area of interest they chose however we believe how do we do we do we go for action but
you're making that up and make it accessible it curious and helps create all sorts of that the from the forestation and reforestation protect various primary forest forest use concessions and digital imagery like the the low-cost cast step these peaks and use source information
here you can see there what's available in the map this is roughly how did the amount of that that you have on the as he sees not just 1st is that there is a lot about context so that that relates to forests including a people rights resource rights etc. it
enables combinations sharing and analysis it makes it easy to be surprised to download to combine or to share information 0 here what you
are seeing is the animation of the latter Laertes thesis
global that it is not that big of blood in their head they're not salads but this is the 1st station based on land that the and this is the process user would use to describe whether it's in
in our in our protected area south of the Berlinale as you see the whole process may be like
30 seconds this person
receive their and alert whenever there's news deforestation detected in a protected area there within hours of of within days of that happening we also create new
that inside we use the existing that to combine it and to create Francesmary products and a is an and and tools that people can use to understand that the status of
forest in in in a broader way in
this case what you're seeing is a dick climate
leg off will force watched and this is how a user would customize comparison between to areas in Brazil in terms of forest-related climate data this reports come because
some eyes and they can they
interactive and they can be shared or downloaded over as much
that is founded on on on open data it wouldn't exist without open data and then we use Open Data heavily to know when the 1st spatial happens but also to know why the 1st spatial happens and furthermore to know what that means not all the 1st station is is created equal if the bears the
first one monitor when and where the 1st patient is happening by detecting forest change globally in near real time sorry
this information you can see that the product you and the product which had a shows the forestation alerts in the last roughly 14 years this is something that was created by the University of Maryland and this is based on a 100 per cent on 3 available that from 2nd
wife why is this for spatial happening we can compare that that that was the context was players
for example in this case you will see an area in the center of Brazil heavily deforested but for some reason here there's no the 1st station if you open the land rights air later you will see that it matches exactly an area that we managed by local people you can understand why the 1st Asian happens or what its strategies or policies that help a deforestation from happening In this
example we are in a more nail you're seeing a heavily forested area and when you open the all Parliament concessions if it becomes clear what's causing that the part that the 1st station in this case is the 1st station for all of call for all of their fields the
3rd is knowing what the significance of that 1st is but would you see here the
darker green is what we call an intact forest landscape in the year 2000 dark and on and light green was in fact no human activity had never occurred there as you see we detected these lines which are obviously logging roads and wherever you see a logging road there is there that the the light green is i reduction is then is degraded to the forest that's not in that anymore here you have
and at the same example somewhere else but this area and in this case we're not looking at the other units we're looking at the other it's the sellers may have happened when when this was taken it was that it was made but this may have happened a week ago you see degraded forest is a a a heavily forested and new roads are a showing up here if we let this stay like that it will be
degraded forest in in in very little time so the important thing here is that we can detect the stationed at this level errors we can avoid the the the for spatial from happening we can act
here you can see a how the slogan roles are showing up there just weeks ago when you open the protected areas they here you can see that this is actually a protected area normal general should be should be happening here His indicating that could be a National Park Peru another example
what you see here is tiger conservation corridor and you have you can see that the corridor is about to be talking by the 1st station yeah this is how we will look at the importance of the station is not just the trees are being cut is where they are being cut how they affect people how they affect a biodiversity how they affect climate finally very important
well for as much as they said if this open source the code is source the API stop source they are based in our micro-services approach which means new services can be created in any language as long as they and the year to certain standards so anyone is invited to create new models for the API that respond to different users we welcome developers to use the API to create derived products uh that they can use developers can extend any of the tools I take a uses and query and of the nominal data programmatically uh you can create suspicious descriptions of the war with your application when when you know deforestation is affected and of course we are happy to hear any any ideas of how these problems could be its extended and what's what's there for the future what we're thinking about of course we have seen a lot more that have more timely that they're going to real-time that a more about that than more local that that this is something that it's going to happen more growth crowdsourcing tools for that the validation foreground validation we're exploring Artificial Intelligence in solutions the flag relevant changes for example when you so those roads showing up there you would you would be hard to detect a logging roads by a trying to trying to find a linear features in in those solids and a 1 of the big steps we want to go from earlier detection to prediction and that can be done with deep learning knowing a knowing what patterns have happened in the past what driver what drivers affect the 1st what areas are more at risk we can go from early detection to prediction it we can take that step we will avoid even though slogan rose from showing up I have a few examples of how global force let's have the jury has been used by governments by the NGOs and by civil society I don't have time to talk about them but I'm happy to to tell you is about the review want thank you very much thank you very much don't for
this great presentation we have questions or comments from the audience Miller thank you are there any WMS services or something that can be consumed and so relative from inside on it I mean actually we consume they're all the that the source is not an and we create our own but is there is there is a rising to indigenous territories in Brazil for example is WMS and consume it again and we have a bunch of different things layers error in especially in the commodities and the fires of sites of gophers what they are and there are several of them the name any more questions yes will you be adding central data was already there it's coming it's coming we're not as was the case we're working they want we have plenty of time now because we only gonna have to presentation so don't keep the thank you for example is impatient so you based mainly on 1 side door you said 15 sources sources use more this so it depends them as I showed you there's several a forest change so so 1 of them for example is that you in the 12 years 30 by 30 resolution meter that's Landsat but there is another layer of format which was based on Maltese and there's layers that are derivatives of those and and that that have been programmatically a process to to give new products and the civil layers which are a done manually and in Brazil there's this what as the the system other this month there nevertheless although this fundamental in in Brazil is done by looking at imagery that doesn't have to realize that the roads can be sent and tagging it manually so there's there's all different kinds of sources Sources fires for example are there is the problem which I believe is Landsat you with a bit of models as well and I'm should so there's there's uh and now we experimented with experimenting with wheather sentinel that so in general you don't have it your database but you're taking the data from yes it's usually like that we a take that take open-source products or talk to scientists who are doing that and they those products and put them on the page or weak something's that what what we do is harmonizing different datasets for example you have protected areas what the person that is not the case because there's a uh harmonize opposite but you have mining concessions for several different countries with different Attributesand different formats different projections so they are harmonized and and and and merge that's that's that's part of the thank you just the bad things for the presentation of I would like to know if it would be possible to use the data for a Mapping in OpenStreetMap if 2 OK the in the new and absolutely absolutely every piece of they think that force which has a license and and some of them are not owned by by gophers what in fact most of them are the only golfers what's so it it we should look at that places this but and you might be unsafe saying that 90 % of the power of that that there our our creative commons they can be used anywhere they can be derived and they can be used and actually there's a project called loading roads Base also convened by WI but more resources to the that's precisely about tagging roads air from set in a satellite imagery tagging them into into OpenStreetMaps located 2nd question would be if it did it done essentially downloadable yeah they present downloadable in most of the cases that it can even be cured queries and download for sample just for geospatial the for a polygon of your choice or for idea of your of your choice the API of models for that and then again the vast majority of their of their that the sets can be downloaded only encourages for people to the use of of course so there was only the 20 minutes alarm so we still have some time punch questions higher this was as nation under 1 of the question you quite often see such as borders and its size and what happens after some years of some months they disappear so I would like to understand is that we have a funding model in place that that we can ensure that this those who continue or you have that a new way question model to convey this thing so I think 1 of the things that have distinguished this project from the start is that this the founding partners made a big effort for these not to be a a product of a three-year project the intention was never to have a three-year produce we know that if a digital product is not taken care of it just becomes obsolete very very area so this is a very important part of the dimension so in the funding mode is not about how you get the money but how you use it so we knew that 1st of all it was important to really is very early it was important to show the potential of these to attract new partners apparently that that work was started with them partners we are 19 and all those partners to efforts in expertise in that in the air in in infrastructure and also in funding so that's that's the important of getting people together and how do you get people to to to to come by by evolving the tool everyday the school has had like maybe 20 pressure reduces since he was born it has been redesigned is that is a perpetual prototype we know we can get a perfect we but if we get it early is going to make a difference and that's that's what we try to do we try to reduce very even if I did that means that sometimes you have certain problems not working again perfectly but you have to show the potential and you and you engage people and I think that's the the most important part thinking about we're not thinking to build up that's going to be equal to that which development is going to be closed after after the 1st stage of development we are i ter rating on a prototype proprietary perpetually OK I have a final question maybe looking into the why of the deforestation process quite a lot the if if you may get some statements about that that's quite a responsibility in regard of data quality so you say you start to look into the data quality issues regarding Graham crowdsourcing how do you view was that at the moment so yeah there's there's sigh and I'm not going to know that there's a lot of issues around that data quality my point is that we didn't have that the data before these data may not be perfect and it may need to be validated by someone but is data and it's an eye in the sky is not they're asking a government to tell you how much the for spatial happened is that this is this their data that hasn't been manipulated by anyway so of course they may not be perfect but of course the it means something that is there because then someone can go and validated for example and that that logging road that they were shown there if you didn't exist OK fair enough it died as x is that most of the time they do we made a we may go wrong about this exact date maybe or the exact peaks so should falling and other things this is not intended to point fingers is not intended to go to companies and governments and say hey you're doing wrong actually the recessive cycle commodities and there's going to be any 1 called financed a . golfers what's the parts that are intended for companies to actually use these tools to air to reduce the 1st station in their supply chains to understand the whole there that's a IKEA should understand where at there they would there they're using comes from and and if that involves any of deforestation risks and when they understand it and they have these tools they probably are going to to to be more conscious and make decisions regarding the 1st station the thank you very much