Our colleagues at NiJeL are doing great work on community action through mapping. I highly recommend checking out there work.
Also, the UNDP's Threat and Risk Mapping Analysis (TRMA) project in the Sudan does participatory crisis mapping at the community level. Very much worth checking out as well.
I have heard from various community mapping initiatives, but wonder what are the different objectives of such efforts? I read for example that it is used to represent in maps different interests around territories for conflict negotiation, but I am sure there are many more objectives. Can you maybe explain when and why community mapping is helpful? I also found this interesting resource by IFAD: Good practices in participatory mapping (http://www.ifad.org/pub/map/PM_web.pdf) Thanks in advance!
Hi Christian, yes I agree with you there are many reasons for using community mapping, depending on the context. I used community mapping with a small group of villagers in Samoa with the aim of transforming an aerial map imbued with scientific data - into a map that was 'localised' with cultural, historical and environmental knowledge of the villagers. Too often maps - especially ones used by government authorities - are made using a dominant view and by experts, reinforcing the power relations between those who map and those who are mapped. So community mapping could be used to turn the tables around and enable 'vulnerable people' to map what they know. I think determining the objectives of community mapping is part of the wider mapping process, one which should involve all parties.
Thank you for sharing this great example from your experience in Samoa. Your comment - "Too often maps - especially ones used by government authorities - are
made using a dominant view and by experts, reinforcing the power
relations between those who map and those who are mapped." - reminded me of another community example where mapping provided the community with an opportunity to share their knowledge and come together as a community.
This map was done in a very basic way - on the floor of a church. The community recaptured their community that had been demolished by the apartheid regime in South Africa. The map is "housed" in a church that was left standing and has become the District Six Museum where they purposefully set out - Mapping personal histories to reclaim a place in history, recover lost land and promote social justice. In the same way as you pointed out, "turning the tables around and enable 'vulnerable people' to map what they know."
This makes me think that with today's technology, it would be a very interesting experiment to create the community map digitally and superimpose this map on the current map of the area. This would provide an opportunity to create a lost historical map and identify the forced changes that occured.
Nancy Pearson, New Tactics in Human Rights Program Manager
Hi Nancy, I just read your link about District Six - what a powerful story! I think your idea of superimposing a community made map with an existing one is fantastic...it would be interesting to select a few locations, perhaps those in conflict (it could be war/civil conflict, land dispute/reclamation or environmental conservation) and engage the local communities in mapping their own experiences and knowledge of the area. We could then compare with existing maps used by government authorities and see how different they are and where they overlap. This would be a great project to 'turn the tables around' and to empower vulnerable communities become cartographers in their own right! Any ideas which organisations would be interested in getting involved in such projects? And who would be willing to fund?
Note that community produced maps are a longstanding tool in various fields, I know much has been done with them in the conservation sector especially, to map protected areas, hunting grounds, sacred areas, etc. I think Patrick mentioned the UN Sudan project (http://www.sd.undp.org/projects/dg13.htm) which is a way to map conflict areas. So, there are tons of resources for how to do this, two great examples include:
The first is especially regarded as a standard. One interesting side note of such mapping is that it can also directly facilitate dispute resolution. While government maps, for example, are often an expression of oppression and authority, they are also then by definition not very accurate. Thus, the process of participatory engagement over a mapping table (or computer system, or cocktail napkin) can serve to "gently" correct government maps in an open context, minimizing the risk of violence. In brief, by getting adversaries to focus on a map rather than each other, conflict resolution is improved as arguments are detailed and based on data on a map. While not a perfect method, it is a tool. This is really an ancient idea, and many border disputes around the world have been peacefully resolved via diplomatic mapping conferences. Not all of course ;)
As for project ideas and proposals and funding, its possible I can help a little on this, depending on the project and cost of course. Or, I am sure mself and others can help develop proposals. If you have any ideas, please do feel free to list them here or email them...
Thanks for sharing this example, Nancy! I have another similar example to share in which communities are collecting a documenting information to preserve cultural heritage sites through geo-mapping. This example is called the Diarna Project, a project coordinated by Jason Guberman-Pfeffer. The following information on this project is taken from the Diarna website:
"Diarna, “Our Homes” in Judeo-Arabic, is a project dedicated to
virtually preserving Mizrahi (“Eastern”) Jewish history through the
lens of physical location. Satellite imagery, photographs, videos, oral
history, and even three-dimensional models offer a unique digital
window onto sites and communities disappearing before our eyes. As
structures decay and the last generation to live in these
locations passes on, we are in a race against time to preserve
priceless cultural treasures."
Challenge: Hundreds of Jewish sites from Morocco to Iran are rapidly disappearing, while the generation with first-hand knowledge of these locations is passing on...It is difficult to visit and impossible to preserve many of these sites, and while historians have written about these communities, few have provided geographic documentation.
Goals: Diarna seeks to memorialize Jewish heritage sites across the Middle East via three core products: (1) a map of Jewish sites across the Middle East stored in a digital database and plotted directly onto Google Earth satellite images; (2) a multimedia collection featuring archival and contemporary images and videos of these sites; and (3) dynamic education methods for sharing this information with the public, including virtual tours, interactive presentations, three-dimensional models, and curricula.
Audience: Diarna eschews politics by focusing on collecting factual information and allowing varied audiences to draw their own conclusions. The project’s data reveals that Jews lived in towns across the Middle East, built a vast array of communal structures, and left a mark on the landscape visible long after communities have completely disbanded. The collection’s factual material resonates with a range of audiences, including the general public, the Jewish community, as well as young Arabs, Persians, Berbers, and Kurds – who grow up (in most cases) surrounded by Jewish sites but without Jewish neighbors. Deepening appreciation of this often ‘hidden history’ can perhaps help promote tolerance.
Are any of you familiar with similar projects to preserve and map heritage?
Kristin Antin, New Tactics Online Community Builder
Mapping is collecting data and applying geographic coordinates to it. The people that have the most data about conditions and situations in specific communities are local stakeholders. Finding a way for them to easily transfer that specialized knowledge and creating incentives for them to do so is invaluable. The world is too big to have in house experts plotting every data point. It will have to be user-generated, regulated for quality, and in a standard format to make it easy to compare and present it. It won't be academic quality data, but it will be a good starting point and guide for development organizations and human rights organizations. That is one of our goals at CollaborAid
In my experience, I have seen community mapping be used as a way to empower local communities by validating their knowledge, ie, local knowledge. I witnessed this first hand while in the Sudan earlier this year when consulting on UNDP's Threat and Risk Mapping Analysis (TRMA) project. It was amazing to see just how powerful an effect maps have when used in focus groups as a participatory mapping. Local participants immediately took ownership of the process and being able to physically draw on maps did give them a sense that their own knowledge was being validated.
This thread of the conversation has been very interesting and has reinforced our New Tactics experience regarding the value and multiple purposes served by our "tactical map" tool.
Although our mapping tool is not a geo-map, it does create a visual map of relationships. The tactical map operates on the same principle of participatory mapping. It provides participants with an opportunity to take ownership of the process and draw out their own knowledge regarding the relationships and institutions within their local community and how these connect to national and international levels.
As you point out, a significant benefit to mapping processes is empowering the participants themselves to recognize and validate their own knowledge.
Nancy Pearson, New Tactics in Human Rights Program Manager
Though not as 'mobilizing' as we would like, the process of interacting with local groups to develop geospatial information on attacks on civilians and other human rights violations is a community building experience. For example, in eastern Burma and Thailand I interact with various groups like the Free Burma Rangers, Karen Human Rights Group, and the Thailand Burma Border Consortium to develop information on where attacks are taking place. Once we have such information, we review available satellite imagery and acquire new images as needed. The entire process really means a lot to local groups and people, they generally feel ignored and marginalized in world affairs so its important to them that outsiders are literally watching and finding evidence to support their reporting. This has also been my experience on Zimbabwe, Somalia, and elsewhere.
One other bit on this, the Open Street Map folks and other organizations are currently doing a participatory mapping excercise on Kibera slum in Nairobi. See http://mapkibera.org/. This is a perfect example of geospatial technologies used to engage and mobilize a community and it will be interesting to see how it develops.
Participatory mapping is very interesting indeed - something I'm looking at for disaster management as well, since many of the community based relief efforts involve participatory mapping.
Recently conducted a set of experiments in the U.S where I worked with folk from Google, OpenStreetMaps (OSM) and InSTEDD. We set up a OpenStreetMap server locally, which spit out WalkingPapers with a Google Enterprise overlay - users then walked around the block marking and annotating stuff on the paper maps, which was then fed back into the OSM server- Sahana could then fetch this data from the OSM server as tiles - what we had was updated content. What we achieved was a simple and effective participatory mapping platform which had the added advantage of findings its way back into the OSM server as well.
WalkingPapers is an interesting concept - it is an extension of OpenStreetMaps, which prints out a paper based map of the region selected. Users can then go around with the paper map, draw and mark and annotate on it, and then scan and send the map back to the OSM server - here it will be put into the queue where someone can come 'digitize' or translate the new map features to the OSM format, or the users who printed the map can make the changes themselves - this is like a poor man's GPS, I guess. Very interesting in a participatory mapping context.
Thank you for sharing this exciting endeavor with us - the mapping of the Kibera slum is taking place as this dialogue is happening. It will indeed be exciting to see how the project and process develops.
Nancy Pearson, New Tactics in Human Rights Program Manager
What are some of the over-arching challenges that geo-mappers face in their human rights work? What are the lessons-learned that can be shared?
What challenges have you faced regarding data collection? What lessons have you learned regarding data accuracy, integrity and hard analysis?
What challenges to national NGOs on the ground face in utilizing geo-mapping tactics for human rights work?
Are there opportunities for more collaboration and coordination among the tech-savvy, information activists, geo-mapping practitioners and human rights practitioners?
What challenges have you face regarding privacy and security for those whose data you are mapping? What lessons have you learned or words of advice?
I would like to learn more about how organizations have been able to use satellite imagery and/or geo-tagged information as reliable evidence. Is geo-tagged information target of particular scrutiny? What are the recommendations for the use of GPS logged data? Outside of Human Rights experts what other areas of expertise can support the analysis of satellite imagery?
Good questions, and ones that our favorite scientist Lars Bromley will no doubt answer very soon. In the meantime, I'd point you to the work that UNOSAT is doing since they do use satellite imagery analysis to provide evidence. A 5 minute video of UNOSAT's work will be posted on Crisis Mappers Net (CM*Net) in the next 24 hours, so do look out for it.
Also, the Joint Research Center (JRC) out of Ispra, Italy is doing some excellent work using automated change detections (algorithms from mathematical morphology) to count IDP populations. More here.
the Z-GIS group of the University of Salzburg is also involved in developing algorithms for counting refugee camps. I saw their presentation last year at the UNSPIDER (United Nations Platform for Space-based Information for Disaster Management and Emergency Response)
@Sandra Your example sounds interesting. I wonder whether you can explain a bit more why the development of algorithms is important in this project? How does it work?
Hi Christian, I'll try, I have not done any remote sensing for a while ;-)
well, the tricky thing with the automatic refugee dwelling extraction from a satellite image is that it is quite often made out of very different material. Only in arid areas and after an immediate emergency the refugees stay in tents which have a defined spectral reflectance and are easier to pick when classifying an image. In a protracted crisis and also in emergency areas where there is enough building material the houses/huts do not necessarily look so much different than the surrounding area, for instance if the roofs are grass thatched etc. Often there is also a mix of grass thatched houses, iron and plastic sheeting; which makes it more challenging as well.
It is possible to develop an object based extraction method which will look into the shape and size of an object instead. But it is not so easy to get the parameters right. I was talking to their team and also to the DLR (German Aerospace Centre) team who have been working on this for a while. They probably have some links to JRC; at least they had joint projects at a stage.I realised that the presentation that is online at the link I included in the last post is actually on vulnerability mapping in terms of floods.
Let me know if this answers your question, Christian.
I am by no means an expert on this, but there's a fair amount of literature on it and folks working on it. I think one detail of interest which we might not think of is that we like to assume that if we count structures we can estimate number of IDPs / refugees. However, in Somalia at least this was very difficult in that the number of IDPs per structure varied greatly. In many areas, the tents were basically only for food storage and preperation, and children / pregnant or nursing mothers, and people generally slept outside. So, the tents and structures are an indicator, but translating them into counts of people is extremely tricky.
Purpose of algorithms and collecting data on IDPs / refugees
Hi Sandra, Lars, Patrick and others,
This thread about algorithms and collecting data on the numbers of IDPs and refugees in camps is really interesting - but for those of us that do not work in geo-mapping, can you explain the use and purpose of this type of work? Is it to collect data on geographic areas that are otherwise difficult to access, for the purpose of planning logistical support for those IDPs and refugees? Or, is it being used for advocacy, somehow? I imagine it could be used in a number of different ways - can you share some examples?
Kristin Antin, New Tactics Online Community Builder
Camp mapping could be used for site planning and also advocatin
Kristin,
Camp mapping could be used for site planning and also advocating and the process described by Sandra is part of the exercice.
With this example in Padibe IDP camp in Northern Uganda (not planned) just by the facts of plotting the facilities we see that there is a potential risk case scenario to see military barracks close to the school. Again it has to be correlated with the figures of "criminality" reported but it is a situation at risk.
It is mentionned in another post from Sandra, colleagues dealing with data on the grounds are "extremly cold" to see the incidents within a camps, it is well know if you put latrines in the surrounding areas were there is no light is a risk factor for having more GBV.. however it is never mapped or tracking those incidents is always tricky. due to the risk that could be bring to the victim(s)
so site planning and geomapping will avoid those risky situation.
When Sandra was talking about aggregating data we borrow methodologies of the ecologist grid cells that could be 100m , 200, 500m just to represents concentration... how many incidents to you have in this grid cell (A1 like in an urban map) ? and then you see a kind of density... it will come! protection officers will do the steps one day, we just have to advocate a bit more those possibility of analysis the facts.
that could/should be correlated with the info in the ground.
mapping the data even if it is internal/ confidential
<pre>
About the so detailed information level (the camp mapping) I would like to share reports from colleagues in the field.
</pre>
<pre>
</pre>
<pre>
Maybe like that it is not clear, but after a map had been done to show
the distributions of ethnicities within a camp in chad; ethnic violence flared up stronger in the blocks where the the specific ethnie live.
And that it is not hard to assess whether this is coincidence, whether
the map which was supposed to be internal had actually been
circulated outside and had found its way to the militia groups active in
the area.
</pre>
<pre>
So as you all know, we have to take with "peck" and really be careffull with Geoinformation related to HRV...
</pre>
<pre>
<a href="http://www.cartong.org">CartONG,</a> Yann Rebois
</pre>
that is totally true, we encountered similar issues in Uganda as well which is why we only used the Sat image to come up with a sampling method for interviews and establishing an address system at the same time. And of course it gave us a rough idea on how many individuals could be residing in a particular area of the camp.
Thanks Sandra, so it means that certain pattern/objects in maps can be found through algorithms? A different approach would then be the North Korea Uncovered, where volunteers try to identify buildings?
Thanks for that link, Christian, that looks super interesting!
Well, for classifying satellite images it is always advisable to have ground truthed data on at least a sample plot to verify whatever you (or your algorithms) extracted. Groundtruthing is basically what the North Korea project does.
This is also necessary when digitising for instance roads on high resolutions satellite images; you might also need to know name of the road for your map/ your application and therefore have to go yourself or send someone to the ground to get it.
More Organisations involved in sat img extraction: Kibera
Hi Everyone,
I found some more resources/organisations involved in satellite image extraction with focus on humanitarian topics.
There is the RESPOND initiative (link points to Kibera Map on Population Estimate through Satellite Image Extraction) and Keyobs, one of the RESPOND Partners, who has worked on that particular example. Other RESPOND partners are UNOSAT and JRC (both have been mentioned by Patrick Meier in another post ).
The issue of evidence is a tough one. In its most formal sense, evidence means a court case, and in most legal settings the bar for evidence is extremely high. So, while satellite imagery very often serves as supporting evidence and corroborates other reporting, such as witness statements, it does not often serve as direct evidence itself. The reason for this is that we very rarely capture a crime "in progress" on imagery so to speak, or with enough detail. For example, while we can produce hundreds or thousands of images of villages destroyed in Darfur or Burma, and we can produce many images of adjacent janjawid or military encampments, we cannot to date produce any images of such militia/military units carrying out an attack directly. Therefore, the images are useful to prove that something drastic is happening, but not useful to incriminate anyone in particular.
One possible exception to this is work my group recently did with Physicians for Human Rights, where we located an image showing mass grave tampering in August, 2006, (see http://afghanmassgrave.org and http://shr.aaas.org/geotech/afghanistan/afghanistan.shtml). We literally saw a hydraulic excavator and dump truck atop what became an excavated mass grave, which afixed a date to the tampering, and filled a key gap in the evidence. Within about 72 hours of PHR releasing its findings, Obama announced an investigation, but we don't know where that will end up.
Another exception is work we did last year on the Russia / Georgia fight over South Ossetia. In this case, the issue of when exactly particular villages were destroyed was important. Specifically, the Russians claimed the villages were destroyed when Georgians controlled the area, while Georgia claimed the opposite. Once we were able to review imagery from mutliple dates, we could show conclusively that the villages were destroyed while Russia was in control, based on their own statements of when they took control and things like helicopters and other military assets visible in the imagery. This was cited by the EU fact finding commission on the conflict, see http://www.ceiig.ch/Report.html. Note that UNOSAT produced a couple images of burning houses in the area, which was one of the few examples of seeing an attack 'in progress' so to speak.
On the other hand, we worked hard on Sri Lanka earlier this year, and while imagery work was cited in the US State Department war crimes report, they specifically said that imagery did not identify culpability in this case and thus did not constitute effective evidence. I am still arguing some of that with them as we were able to link shell craters with specific mortar positions, for example. They did include a few images which they call satellite images, but I think some of those are air photos based on the angle. See http://www.state.gov/documents/organization/131025.pdf for the report, or the report of my project at: http://shr.aaas.org/geotech/srilanka/srilanka.shtml .
A final example, some work I participated in a couple years ago located "unknown" IDPs in Somalia, that then required the UN to mount a relief operation. Not evidence in the courtroom sense, but it definitely served to change policy in a very real sense.
So, in sum, its not necessarily a direct link between imagery and evidence. Like all such evidence, it depends on the situation and context, and whether the evidence needs to survive extensive cross examination. I can talk more about this if needed, just let me know.
I might be a bit iconoclast and will go in the sense that was
discussed the technology should not hide the content, (l'arbre cache la
foret) human rights violations imagery is a prof that human rights
violation occurs, more as a sensitization and awareness to the global
society and will provide advocacy at a higher level (political ?), (it is
always a scale level). So there is real need of classical,
and not
always e-information, of mapping to highlights the facts and act
locally (if possible due to the context ) so collecting information in
the ground as close as possible of the source of information , is vital
trough NGOs network. EG in DRC kivu, local NGOs are able to report
incidents and crime violation within a regular manner (trough paper
forms) to the UNHCR, big International NGOs are failing... incidents
are reported and actions are taken as quicly as possible for the
victims, (like providing PEP,and Pshychological support) but the data
are not really analysed, correlated with the other layers of
informations... for going further into the analysis. Info management
in this area is
like putting a plaster for short and immediate response, but geo
mapping will be able to reinforce the possibility to overlay the facts
of violation( SGBV, land use violation...) with the environment (presence of police, of military detachement..)
This is of course a good point, and one that resonates with me as I generally stay in Washington and observe human rights violations from satellites. Its important to point out that there is nothing inherently protective about getting a satellite image, it is all virtual, and may not influence the situation on the ground at all. I also worry that such visual tools like imagery simply provide more media content but don't necessarily cause people to care about issues. I cringe when people tell me how "cool" our work is, because it means they are focused on the technology and not what we are documenting. How do we call looking at Darfur "cool"?
On the other hand, it meant a lot to me when some Burmese sent me a note to tell me how great it was to know we were watching and trying to help them, though we had never met. They often feel isolated and alone, so the idea of satellites in space taking images of them was something they really liked, especially as it got the issue on CNN.
As with anything, mapping and imagery is part of the broader context, and needs to be deployed or used locally in many cases to have proper impact. Otherwise we just contribute to the Washington literature circus and media charade...
Privacy issues on the example of Google Street View
Back at the State of Map conference in August from the OpenStreetMap (OSM) initiative, I remember an emotional debate, when one participant proposed whether the OSM project shall also offer something like Google Street View. Some mappers said that they will abandon the project the minute somebody starts such an initiative. I observed that there is a fine line of what can or shall be mapped and what not. Suddenly the whole mapping issue became quite controversial and left the technological discussion. I think it is very important to discuss the case of Google Street View as it lead to various protests. Is it always right to publish all available geo-data, or is it better to close some data?
I am not sure whether this example is entirely relevant for human rights activism, but shall there be limits too? Publishing geo-data might help one cause, but easily threatens another cause. What do you think?
Re: Privacy issues on the example of Google Street View
I can't comment on the Street View issue, it seems it is being decided courtroom by courtroom. Regarding whether its right to publish all data, that is a tough issue and differs drastically from place to place. I remember when Google Earth started putting high-res imagery up of a certain area, the folks hiding from persecution in that area got quite scared that it would allow them to be found. From Google's perspective, they just were doing their normal stuff, but they were totally unaware (and remained so) of how publishing such imagery might impact people.
In the projects I am involved in we take extreme care to sanitize much of the data we get to remove anything that might cause persecution of specific people. We often degrade our published coordinates and redact things like witness names. We feel this is absolutely justified to avoid further harm and repersecution of certain people.
This issue of privacy is a critical issue for human rights advocates - for as you point out, the cost of such implications for those who are suffering persecution could mean people lives. Ethical questions and considerations must be considered with the advances of technology.
I'm interested to hear about how those of you involved in geo-mapping have identified what those ethical considerations are in your various uses of the technology (such as this issue of privacy) and what recommendations you can provide for others grappling with such questions.
Nancy Pearson, New Tactics in Human Rights Program Manager
<p>
Nancy, we normally work together with the<strong> protection officers </strong>(most of them have a human rights lawyers' background) and it is them who assess the risk, but it is tricky.<br />
<br />
Apart from us only dealing with the <strong>anonymized data</strong> (e.g. no names); we aggregate the data on a <strong>defined grid</strong> to not show the actual location where the perpetration happened but just a concentration of specific events. See also Yann's reply on that. <a href="/blog/new-tactics/geo-mapping-human-rights#comment-3087" class="active"></a>
</p>
<p>
<a href="/blog/new-tactics/geo-mapping-human-rights#comment-3087" class="active">Camp mapping could be used for site planning and also advocatin</a><br />
<br />
But I don't think it is ideal. So I would be very interested to hear how others are dealing with this.
</p>
<p>
Sandra
</p>
<p>
<a href="/blog/new-tactics/geo-mapping-human-rights#comment-3087" class="active"><br />
</a>
</p>
Visibility vs Vulnerability: ideas2 maximise 1st & minimize 2nd?
Most of the incidence mapping on human rights violations CartONG has been involved so far was on confidential data and the maps were for internal use only. Why? Because the mapped incidences were linked to refugees (or internally displaced persons - IDPs) or even subgroups of refugees (for instance women: Gender Based Violence - GBV) and there were valid concerns that public mapping would put the victim at risk as it can be traced back in such defined communities. One of the work arounds was to aggregate the data at a higher level; but even then the maps were considered too dangerous for public use.
There are really nice examples out there using web-mapping techniques for human rights mapping which would bring so much more visibility.But it is difficult to assess the risks involved for the individual. Has anyone looked into this before?
Has anyone had experiences with public human rights mapping where only certain groups of people are targeted? And yes, how was the risk minimized?
Visibility vs Vulnerability: ideas2 maximise 1st & minimize 2nd?
Thank you so much for your posting! I have always thought of geo-mapping as a way to portray the results of a situation (victims in Darfur, environmental consequences, etc.) instead of a situation (as in the case you mentioned on refugees) and risking a negative impact. It definitely provided a different perspective on geo-mapping. Geo-mapping can have a very positive influence, as mentioned throughout all the postings, but also have risks involved.
I was wondering if anyone has themselves recognized and identified risks within their own geo-mapping projects. It seems as if it would be hard to recognize the risks when your own project has only good intentions. Was it something you or your organization identified? Was it difficult to realize them? Or was it similar to Yann's experience, where a separate organization pointed out the risk?
One primary risk on our Burma project was pointed out by local project partners. In Burma, many villages at risk maintain a second, hidden set of shelters and supplies so that if the army attacks the villagers can flee to these hiding spots. When we first produced satellite imagery analysis for Burma, we made all of our imagery available via GoogleEarth and other methods. However, our local partners immediately let us know we were revealing some of these hidden villages, so we quickly modified our published imagery to only show those villages attacked, not the surrounding areas.
In other work, we've made sure to exhaustively review our reporting data before release to make sure we have no information in it that might get someone in trouble or hurt. Again, the local partners have the final decision on what we can release.
I wanted to point out the different risks associated with the different types of geo-mapping tools (this information comes from Tactical Tech's Maps for Advocacy Guide):
Hosting Maps on your Website (Map APIs)
These are digital maps that can be customized and embedded into your website. The data is hosted by other servers like Google Maps, Yahoo! Maps, Microsoft Virtual Earth or Open Layers.
This type of tool is easy and cheap so lots of people use it. A few things must be considered regarding privacy and ownership of data. The map's API provider reserves rights to use your data (for advertising, etc) so it is important to know the provider's terms of service. Also, to embed the map in your webpage, the user must register the root URL from where the map will be hosted to obtain the 'API key.' Again, read the terms of service to understand what association this creates between a user and a map provider.
Tactical Tech recommends OpenLayers b/c it is open source and free of some of the service terms that other map API providers present. Tactical Tech also mentioned that Mapstraction is a JavaScript map API library that lets users move between Google Maps, OpenLayers and other map API services.
Virtual Globe Software (like Google Earth)
This is software mapping applications that use satellite imagery at their core. Tactical Tech mentions that in addition to Google Earth, NASA WorldWind should also be considered.
The biggest privacy concern is around the availiability of high-resolution satellite data - in which objects can be seen on the ground (Christian, Lars, Nancy and Sandra have already commented on this issue in another thread and Lars posted on privacy issues that AAAS's partners faced in Burma).
Open Source GIS (Geographic INformation Systems)
This type of tool goes beyond displaying and sharing data and allows users to integrate, analyse and edit the data.
Tactical Tech suggested Quantum GIS (as did Lars and others in this dialogue)
Are there any risks associated with using this type of mapping tool?
What other types of geo-mapping tools am I missing - and what other risks regarding the use of these tools should be assessed?
Kristin Antin, New Tactics Online Community Builder
There is an amazing episode of the Brazilian series City of Men called "The Mail", where - long story short - some residents of a slum with the blessing of the drug dealers decide to put street names on the alleyways and small open spaces in the slum. Then the cops confiscate a map made with the new street names, and begin to use these street names to make their way through the favela in their cat-and-mouse combat with the dealers who run the place. (http://www.youtube.com/watch?v=cyu-DQZGEPQ)
In the episode of City
of Men, the street signs were simply changed around to confuse the
police. But in the age of all of the technology we are discussing here,
everything is instantly made available digitally, and not as easy to do
away with.
How does one "unmap" a place? Once data is published or placed in the public domain, it seems that it
would be extremely hard to take it back again.
Information is power. Maps, in the hands of those who would use them for other purposes, can be harmful. Not just in militarized or openly repressive situations. What about for example, if geomapped data meant to "protect" sacred places is actually used to better exploit these places? I am thinking of how in theory it would be "useful" to map sacred places in a country I work to prevent government and others from conceeding usufruct rights of these areas. But I can see that actually this could have the opposite effect, exposing these hidden places to more outside interference.
I realize some of these issues have already come up in your conversation... and I offer no technical insight, just questions...
Fascinating and insightful, many thanks for sharing. I'd say that the issue here is that there was just one map. One would perhaps simply need to make dozens of copies?
Very interesting question: "how does one 'unmap' a place?" Would love to hear what others think.
RE: Visibility vs Vulnerability: ideas2 maximise 1st & minimize
We, CartONG, have conducted all human rights mapping embedded in an UNHCR (United Nations High Commissioner for Refugees) project; so it was either the Protection officers there or within the UN Cluster System (with UNHCR chairing the Protection Cluster in many countries), assessing the risks.
I think a Rule-of-aThumb could be: if your target group is very defined and small in comparison to the rest of the population; reported violations can be traced back easier and repercussions/revenge could be more likely. Now if you are sure you can provide the necessary protection for your target group, you might opt still to make your findings public. But that is a tricky decision; especially if the perpetrators are the ones supposedly providing security (like military etc pp).
Yes, it would be good to get some more general ideas on what rules to apply on publishing data on vulnerables.
UNHCR does have a knowledge base and some protection related material for downloading, but of course their focus are the Displaced.
I would like to share a discussion I had early 2007 with a staff member of an organisation with a red cross in northern Uganda.
We were talking about mapping of the roads. Especially the opening of the new roads and well graded in order to access the returnee sites. Yes we are talking on the single tracks that come from your GPS device, the poly line that should be edited in order to join the rest of the road network, your 2 km road section from the main road to the new returnee sites with 150 dwellings.!
The staff was concern about the opening and mapping the roads. Why? The staff had a fear of easy movement of the LRA (Lord Resistance Army) . In a certain extent, the mapping of those tracks will facilitate the possibility of having again abductions and violence that we are all know. It is tricky to map information for some organisation that are interested by a “physical information” like for the NGOS taht are doing road rehabilitation,and in other hand it could brings some fear to other organisations.
We should understand of having or not mapped information that seems to be in the “public domain” and if the community or humanitarian are not interested of having presented on a maps the information related to there environment, we should respect them. How are we dealing with those dilemmas? especially in a context of the humanitarian clusters and information sharing "a tout va" where everything is shared with not so much control after all.We should really asses the impact or foot print and the potential consequences that our mappings could have. It is like the movement of the butterfly.
I’m playing the advocate of the devil, ;-)I might be ejected of the forum...
These are important issues of course, and I am wondering if they are of more concern in humanitarian mapping as opposed to human rights mapping. We are cautious with publicizing our data or satellite imagery, though this is usually temporary. After our investigation is complete we publicize pretty much everything we can. So, maybe it is all a matter of timing.
Again, we are generally not mapping roads that might enable LRA attacks or something like that. And, further, I'd never assume that my road mapping was better than any local knowledge...
What are the minimum requirements needed for geo-mapping?
Hi all,
It would be great for those of you doing this work on the ground, to share your experience regarding what the minimum requirements are that are needed. For example:
What kind of technology equipment is necessary?
What kinds of infrastructure is needed?
What kind of organizational support (local, national, international) must be in place?
What expertise is needed from people (internal to organizations, external as "experts")?
This would be very helpful to give a better understanding of how realistic it is for organizations to use geo-mapping to advance their efforts; and the steps organizations would need to take to make it effective.
Nancy Pearson, New Tactics in Human Rights Program Manager
Minimal requirments for geomapping: the end user should know
Some minimal requirements for geo-mapping the end user should know what he/she would like to see on the map.
Maybe some elements about your questions If we really want to act, a bit less for advocating we have to put in place a system, system in terms of data collection and processing for a better diffusion, system for the timeliness delivery and update of the facts, it could be based on the locals with more or less risks for themselves, or international, with less proximity of the source of the info, so the information my have some bias.
In terms of system data model or SDI (Spatial Data Infrastructure) should be defined ; as of today there is several initiatives on HRV, (with different topics, it might be interesting to develop a common dataset?
Then in terms of technology, sure the web server technologies, like the Google applications, or feeds (GeoRSS, WMS Web Map Service , WFS Web Feature Services) are the most appropriated to share and to be useed by GIS specialist and advocate and also to register the records like with the SMS ; however in the field it is not always possible or user friendly, or due to IT knowledge of certain local staff member of some organization it is limited (previous experiences).
About expertise and here you point out the biggest challenge, we, techies have to facilitate and to promote the power to the experts that knows how to act with HRV and knows the mechanisms or the systems for referral actions. They will be the one that wil say “I would like to have this information as it could be use on this perspective…” For instance, we get some request from OHCR, for mapping the catchment areas of the new police post created in an area (how many villages are covered In 5km, 10 Km…radius), they were interested by the coverage of the paralegals and the police to see if there is an overlap of the 2 entities… so the specialists will come with their own language and we will have to analyze through spatial analysis for instance, the facts.
With the Google tools, I will say we are more locating or plotting the facts than processing the info… but how could the data could be turned out into real information for decision? We should have a kind of HRV wish list of geo-products that could be used and become standard for HRV. Then we could think on the data that should be collected and how…
thank you for your comprehensive description on different approaches of mapping. Risking to be a bit superficial, I experience two main approaches so far to start a mapping project.
1) You start collecting data yourself to map your area with your topics and aims. For this you either rely on existing maps and mark your points of interest (see community mapping) or your walk around and map such points with a GPS advice for exact position. (OpenStreetMap)
2) You take existing data, which includes location relevant pieces and use it for maps. There are many available databases or own data with geo-information. There are tools such as Google Maps or Yahoo Map Maker (http://developer.yahoo.com/maps/), which offer easy tools to filter locations out of data. These tools allow different forms of easy maps through mashups.
Both above approaches focus on non-experts and still offer some quite impressive opportunities. The above described map of Iran protest during the elections is such an example.
some sample of data collection forms within the DRC system
Hi, I’m
maybe after alltoo “old school” ;-), ?!
It is true
that Martus is a standard. I should
reckon I had a look some years ago and the spatial component was not really
there… Since then as I did was not involved directly on HRV I didi not follow the evolution.
With Google
tools it is true that is simpler and simpler to share the information .
Do you have any doubts when we are talking on confidential data ? CartONG also used Google Application, but nothing was related up to know on HRV, due to confidentialiy
I would
like to share some data collections forms that are used in the fieldin DRC (mainly Kivu) by the protection
cluster/ UNHCR to report, act and advocate.
It seems to
be old fashion system, but it is working
on a weekly basis but if you had on top of it new geo-mapping spirit it could
be quite powerful, I think. And it will be great! And they are looking for
geommaping support. As of today, They prorduce or would like to produce some
maps showing concentration of incidents.
If you look
on the system (Flow chart) you really see the complexity on the stakeholders
that will need a specific kind of information from the raw data, to the already aggregated
information. What is interestingis to
see the aim of the data/geodata 1. program, 2 advocacy, 3, recommendation4, action and referral to the specialised
agency. About the standard, two
forms are attached as a sample. 1. 2
Our colleagues at NiJeL are doing great work on community action through mapping. I highly recommend checking out there work.
Also, the UNDP's Threat and Risk Mapping Analysis (TRMA) project in the Sudan does participatory crisis mapping at the community level. Very much worth checking out as well.
Patrick Meier, Ushahidi
I have heard from various community mapping initiatives, but wonder what are the different objectives of such efforts? I read for example that it is used to represent in maps different interests around territories for conflict negotiation, but I am sure there are many more objectives. Can you maybe explain when and why community mapping is helpful? I also found this interesting resource by IFAD: Good practices in participatory mapping (http://www.ifad.org/pub/map/PM_web.pdf) Thanks in advance!
Christian Kreutz – crisscrossed.net
Hi Christian, yes I agree with you there are many reasons for using community mapping, depending on the context. I used community mapping with a small group of villagers in Samoa with the aim of transforming an aerial map imbued with scientific data - into a map that was 'localised' with cultural, historical and environmental knowledge of the villagers. Too often maps - especially ones used by government authorities - are made using a dominant view and by experts, reinforcing the power relations between those who map and those who are mapped. So community mapping could be used to turn the tables around and enable 'vulnerable people' to map what they know. I think determining the objectives of community mapping is part of the wider mapping process, one which should involve all parties.
Mapper,
Thank you for sharing this great example from your experience in Samoa. Your comment - "Too often maps - especially ones used by government authorities - are
made using a dominant view and by experts, reinforcing the power
relations between those who map and those who are mapped." - reminded me of another community example where mapping provided the community with an opportunity to share their knowledge and come together as a community.
This map was done in a very basic way - on the floor of a church. The community recaptured their community that had been demolished by the apartheid regime in South Africa. The map is "housed" in a church that was left standing and has become the District Six Museum where they purposefully set out - Mapping personal histories to reclaim a place in history, recover lost land and promote social justice. In the same way as you pointed out, "turning the tables around and enable 'vulnerable people' to map what they know."
You can also read more indepth account of the community involvement in this great process in the tactical notebook, The Power of Place: How historic sites can engage citizens in human rights issues.
This makes me think that with today's technology, it would be a very interesting experiment to create the community map digitally and superimpose this map on the current map of the area. This would provide an opportunity to create a lost historical map and identify the forced changes that occured.
Nancy Pearson, New Tactics in Human Rights Program Manager
Hi Nancy, I just read your link about District Six - what a powerful story! I think your idea of superimposing a community made map with an existing one is fantastic...it would be interesting to select a few locations, perhaps those in conflict (it could be war/civil conflict, land dispute/reclamation or environmental conservation) and engage the local communities in mapping their own experiences and knowledge of the area. We could then compare with existing maps used by government authorities and see how different they are and where they overlap. This would be a great project to 'turn the tables around' and to empower vulnerable communities become cartographers in their own right! Any ideas which organisations would be interested in getting involved in such projects? And who would be willing to fund?
Note that community produced maps are a longstanding tool in various fields, I know much has been done with them in the conservation sector especially, to map protected areas, hunting grounds, sacred areas, etc. I think Patrick mentioned the UN Sudan project (http://www.sd.undp.org/projects/dg13.htm) which is a way to map conflict areas. So, there are tons of resources for how to do this, two great examples include:
The first is especially regarded as a standard. One interesting side note of such mapping is that it can also directly facilitate dispute resolution. While government maps, for example, are often an expression of oppression and authority, they are also then by definition not very accurate. Thus, the process of participatory engagement over a mapping table (or computer system, or cocktail napkin) can serve to "gently" correct government maps in an open context, minimizing the risk of violence. In brief, by getting adversaries to focus on a map rather than each other, conflict resolution is improved as arguments are detailed and based on data on a map. While not a perfect method, it is a tool. This is really an ancient idea, and many border disputes around the world have been peacefully resolved via diplomatic mapping conferences. Not all of course ;)
As for project ideas and proposals and funding, its possible I can help a little on this, depending on the project and cost of course. Or, I am sure mself and others can help develop proposals. If you have any ideas, please do feel free to list them here or email them...
Lars Bromley, AAAS-SHRP
Thanks for sharing this example, Nancy! I have another similar example to share in which communities are collecting a documenting information to preserve cultural heritage sites through geo-mapping. This example is called the Diarna Project, a project coordinated by Jason Guberman-Pfeffer. The following information on this project is taken from the Diarna website:
"Diarna, “Our Homes” in Judeo-Arabic, is a project dedicated to
virtually preserving Mizrahi (“Eastern”) Jewish history through the
lens of physical location. Satellite imagery, photographs, videos, oral
history, and even three-dimensional models offer a unique digital
window onto sites and communities disappearing before our eyes. As
structures decay and the last generation to live in these
locations passes on, we are in a race against time to preserve
priceless cultural treasures."
Challenge: Hundreds of Jewish sites from Morocco to Iran are rapidly disappearing, while the generation with first-hand knowledge of these locations is passing on...It is difficult to visit and impossible to preserve many of these sites, and while historians have written about these communities, few have provided geographic documentation.
Goals: Diarna seeks to memorialize Jewish heritage sites across the Middle East via three core products: (1) a map of Jewish sites across the Middle East stored in a digital database and plotted directly onto Google Earth satellite images; (2) a multimedia collection featuring archival and contemporary images and videos of these sites; and (3) dynamic education methods for sharing this information with the public, including virtual tours, interactive presentations, three-dimensional models, and curricula.
Audience: Diarna eschews politics by focusing on collecting factual information and allowing varied audiences to draw their own conclusions. The project’s data reveals that Jews lived in towns across the Middle East, built a vast array of communal structures, and left a mark on the landscape visible long after communities have completely disbanded. The collection’s factual material resonates with a range of audiences, including the general public, the Jewish community, as well as young Arabs, Persians, Berbers, and Kurds – who grow up (in most cases) surrounded by Jewish sites but without Jewish neighbors. Deepening appreciation of this often ‘hidden history’ can perhaps help promote tolerance.
Are any of you familiar with similar projects to preserve and map heritage?
Kristin Antin, New Tactics Online Community Builder
I highly recommend this book on Mapping Vulnerability which really goes to the heart of the excellent points you make.
Patrick Meier, Ushahidi
Mapping is collecting data and applying geographic coordinates to it. The people that have the most data about conditions and situations in specific communities are local stakeholders. Finding a way for them to easily transfer that specialized knowledge and creating incentives for them to do so is invaluable. The world is too big to have in house experts plotting every data point. It will have to be user-generated, regulated for quality, and in a standard format to make it easy to compare and present it. It won't be academic quality data, but it will be a good starting point and guide for development organizations and human rights organizations. That is one of our goals at CollaborAid
I thought this post on Social Mapping and Crisis Mapping might be of interest.
In my experience, I have seen community mapping be used as a way to empower local communities by validating their knowledge, ie, local knowledge. I witnessed this first hand while in the Sudan earlier this year when consulting on UNDP's Threat and Risk Mapping Analysis (TRMA) project. It was amazing to see just how powerful an effect maps have when used in focus groups as a participatory mapping. Local participants immediately took ownership of the process and being able to physically draw on maps did give them a sense that their own knowledge was being validated.
Patrick Meier, Ushahidi
This thread of the conversation has been very interesting and has reinforced our New Tactics experience regarding the value and multiple purposes served by our "tactical map" tool.
Although our mapping tool is not a geo-map, it does create a visual map of relationships. The tactical map operates on the same principle of participatory mapping. It provides participants with an opportunity to take ownership of the process and draw out their own knowledge regarding the relationships and institutions within their local community and how these connect to national and international levels.
As you point out, a significant benefit to mapping processes is empowering the participants themselves to recognize and validate their own knowledge.
Nancy Pearson, New Tactics in Human Rights Program Manager
Though not as 'mobilizing' as we would like, the process of interacting with local groups to develop geospatial information on attacks on civilians and other human rights violations is a community building experience. For example, in eastern Burma and Thailand I interact with various groups like the Free Burma Rangers, Karen Human Rights Group, and the Thailand Burma Border Consortium to develop information on where attacks are taking place. Once we have such information, we review available satellite imagery and acquire new images as needed. The entire process really means a lot to local groups and people, they generally feel ignored and marginalized in world affairs so its important to them that outsiders are literally watching and finding evidence to support their reporting. This has also been my experience on Zimbabwe, Somalia, and elsewhere.
One other bit on this, the Open Street Map folks and other organizations are currently doing a participatory mapping excercise on Kibera slum in Nairobi. See http://mapkibera.org/. This is a perfect example of geospatial technologies used to engage and mobilize a community and it will be interesting to see how it develops.
Participatory mapping is very interesting indeed - something I'm looking at for disaster management as well, since many of the community based relief efforts involve participatory mapping.
Recently conducted a set of experiments in the U.S where I worked with folk from Google, OpenStreetMaps (OSM) and InSTEDD. We set up a OpenStreetMap server locally, which spit out WalkingPapers with a Google Enterprise overlay - users then walked around the block marking and annotating stuff on the paper maps, which was then fed back into the OSM server- Sahana could then fetch this data from the OSM server as tiles - what we had was updated content. What we achieved was a simple and effective participatory mapping platform which had the added advantage of findings its way back into the OSM server as well.
WalkingPapers is an interesting concept - it is an extension of OpenStreetMaps, which prints out a paper based map of the region selected. Users can then go around with the paper map, draw and mark and annotate on it, and then scan and send the map back to the OSM server - here it will be put into the queue where someone can come 'digitize' or translate the new map features to the OSM format, or the users who printed the map can make the changes themselves - this is like a poor man's GPS, I guess. Very interesting in a participatory mapping context.
Mifan Careem
Thank you for sharing this exciting endeavor with us - the mapping of the Kibera slum is taking place as this dialogue is happening. It will indeed be exciting to see how the project and process develops.
Nancy Pearson, New Tactics in Human Rights Program Manager
What are some of the over-arching challenges that geo-mappers face in their human rights work? What are the lessons-learned that can be shared?
I would like to learn more about how organizations have been able to use satellite imagery and/or geo-tagged information as reliable evidence. Is geo-tagged information target of particular scrutiny? What are the recommendations for the use of GPS logged data? Outside of Human Rights experts what other areas of expertise can support the analysis of satellite imagery?
Hi Enrique,
Good questions, and ones that our favorite scientist Lars Bromley will no doubt answer very soon. In the meantime, I'd point you to the work that UNOSAT is doing since they do use satellite imagery analysis to provide evidence. A 5 minute video of UNOSAT's work will be posted on Crisis Mappers Net (CM*Net) in the next 24 hours, so do look out for it.
Also, the Joint Research Center (JRC) out of Ispra, Italy is doing some excellent work using automated change detections (algorithms from mathematical morphology) to count IDP populations. More here.
Patrick Meier, Ushahidi
Hi Enrique,
the Z-GIS group of the University of Salzburg is also involved in developing algorithms for counting refugee camps. I saw their presentation last year at the UNSPIDER (United Nations Platform for Space-based Information for Disaster Management and Emergency Response)
http://www.oosa.unvienna.org/oosa/unspider/Bonn2008.html
the url for their current projects is
http://www.uni-salzburg.at/portal/page?_pageid=142,125895&_dad=portal&_s...
Sandra Sudhoff, CartONG
@Sandra Your example sounds interesting. I wonder whether you can explain a bit more why the development of algorithms is important in this project? How does it work?
Hi Christian, I'll try, I have not done any remote sensing for a while ;-)
well, the tricky thing with the automatic refugee dwelling extraction from a satellite image is that it is quite often made out of very different material. Only in arid areas and after an immediate emergency the refugees stay in tents which have a defined spectral reflectance and are easier to pick when classifying an image. In a protracted crisis and also in emergency areas where there is enough building material the houses/huts do not necessarily look so much different than the surrounding area, for instance if the roofs are grass thatched etc. Often there is also a mix of grass thatched houses, iron and plastic sheeting; which makes it more challenging as well.
It is possible to develop an object based extraction method which will look into the shape and size of an object instead. But it is not so easy to get the parameters right. I was talking to their team and also to the DLR (German Aerospace Centre) team who have been working on this for a while. They probably have some links to JRC; at least they had joint projects at a stage.I realised that the presentation that is online at the link I included in the last post is actually on vulnerability mapping in terms of floods.
Let me know if this answers your question, Christian.
Sandra
I am by no means an expert on this, but there's a fair amount of literature on it and folks working on it. I think one detail of interest which we might not think of is that we like to assume that if we count structures we can estimate number of IDPs / refugees. However, in Somalia at least this was very difficult in that the number of IDPs per structure varied greatly. In many areas, the tents were basically only for food storage and preperation, and children / pregnant or nursing mothers, and people generally slept outside. So, the tents and structures are an indicator, but translating them into counts of people is extremely tricky.
Hi Sandra, Lars, Patrick and others,
This thread about algorithms and collecting data on the numbers of IDPs and refugees in camps is really interesting - but for those of us that do not work in geo-mapping, can you explain the use and purpose of this type of work? Is it to collect data on geographic areas that are otherwise difficult to access, for the purpose of planning logistical support for those IDPs and refugees? Or, is it being used for advocacy, somehow? I imagine it could be used in a number of different ways - can you share some examples?
Kristin Antin, New Tactics Online Community Builder
Kristin,
Camp mapping could be used for site planning and also advocating and the process described by Sandra is part of the exercice.
With this example in Padibe IDP camp in Northern Uganda (not planned) just by the facts of plotting the facilities we see that there is a potential risk case scenario to see military barracks close to the school. Again it has to be correlated with the figures of "criminality" reported but it is a situation at risk.
It is mentionned in another post from Sandra, colleagues dealing with data on the grounds are "extremly cold" to see the incidents within a camps, it is well know if you put latrines in the surrounding areas were there is no light is a risk factor for having more GBV.. however it is never mapped or tracking those incidents is always tricky. due to the risk that could be bring to the victim(s)
so site planning and geomapping will avoid those risky situation.
When Sandra was talking about aggregating data we borrow methodologies of the ecologist grid cells that could be 100m , 200, 500m just to represents concentration... how many incidents to you have in this grid cell (A1 like in an urban map) ? and then you see a kind of density... it will come! protection officers will do the steps one day, we just have to advocate a bit more those possibility of analysis the facts.
that could/should be correlated with the info in the ground.
<pre>
About the so detailed information level (the camp mapping) I would like to share reports from colleagues in the field.
</pre>
<pre>
</pre>
<pre>
Maybe like that it is not clear, but after a map had been done to show
the distributions of ethnicities within a camp in chad; ethnic violence flared up stronger in the blocks where the the specific ethnie live.
And that it is not hard to assess whether this is coincidence, whether
the map which was supposed to be internal had actually been
circulated outside and had found its way to the militia groups active in
the area.
</pre>
<pre>
So as you all know, we have to take with "peck" and really be careffull with Geoinformation related to HRV...
</pre>
<pre>
<a href="http://www.cartong.org">CartONG,</a> Yann Rebois
</pre>
Lars,
that is totally true, we encountered similar issues in Uganda as well which is why we only used the Sat image to come up with a sampling method for interviews and establishing an address system at the same time. And of course it gave us a rough idea on how many individuals could be residing in a particular area of the camp.
The sampling method we used was based on some previous hut sampling of IOM in Uganda (International Organisation of Migration, http://www.iom.int/jahia/jsp/index.jsp) and Eric Green's work; a scientist who was based in one of the camps, researching among others sampling. Here is his blog; I picked a blogpost on sampling: http://www.ericpgreen.com/2009/10/27/remote-control-sampling/
Sandra, CartONG
Thanks Sandra, so it means that certain pattern/objects in maps can be found through algorithms? A different approach would then be the North Korea Uncovered, where volunteers try to identify buildings?
Thanks for that link, Christian, that looks super interesting!
Well, for classifying satellite images it is always advisable to have ground truthed data on at least a sample plot to verify whatever you (or your algorithms) extracted. Groundtruthing is basically what the North Korea project does.
This is also necessary when digitising for instance roads on high resolutions satellite images; you might also need to know name of the road for your map/ your application and therefore have to go yourself or send someone to the ground to get it.
Hi Everyone,
I found some more resources/organisations involved in satellite image extraction with focus on humanitarian topics.
There is the RESPOND initiative (link points to Kibera Map on Population Estimate through Satellite Image Extraction) and Keyobs, one of the RESPOND Partners, who has worked on that particular example. Other RESPOND partners are UNOSAT and JRC (both have been mentioned by Patrick Meier in another post ).
Sandra Sudhoff, CartONG
The issue of evidence is a tough one. In its most formal sense, evidence means a court case, and in most legal settings the bar for evidence is extremely high. So, while satellite imagery very often serves as supporting evidence and corroborates other reporting, such as witness statements, it does not often serve as direct evidence itself. The reason for this is that we very rarely capture a crime "in progress" on imagery so to speak, or with enough detail. For example, while we can produce hundreds or thousands of images of villages destroyed in Darfur or Burma, and we can produce many images of adjacent janjawid or military encampments, we cannot to date produce any images of such militia/military units carrying out an attack directly. Therefore, the images are useful to prove that something drastic is happening, but not useful to incriminate anyone in particular.
One possible exception to this is work my group recently did with Physicians for Human Rights, where we located an image showing mass grave tampering in August, 2006, (see http://afghanmassgrave.org and http://shr.aaas.org/geotech/afghanistan/afghanistan.shtml). We literally saw a hydraulic excavator and dump truck atop what became an excavated mass grave, which afixed a date to the tampering, and filled a key gap in the evidence. Within about 72 hours of PHR releasing its findings, Obama announced an investigation, but we don't know where that will end up.
Another exception is work we did last year on the Russia / Georgia fight over South Ossetia. In this case, the issue of when exactly particular villages were destroyed was important. Specifically, the Russians claimed the villages were destroyed when Georgians controlled the area, while Georgia claimed the opposite. Once we were able to review imagery from mutliple dates, we could show conclusively that the villages were destroyed while Russia was in control, based on their own statements of when they took control and things like helicopters and other military assets visible in the imagery. This was cited by the EU fact finding commission on the conflict, see http://www.ceiig.ch/Report.html. Note that UNOSAT produced a couple images of burning houses in the area, which was one of the few examples of seeing an attack 'in progress' so to speak.
On the other hand, we worked hard on Sri Lanka earlier this year, and while imagery work was cited in the US State Department war crimes report, they specifically said that imagery did not identify culpability in this case and thus did not constitute effective evidence. I am still arguing some of that with them as we were able to link shell craters with specific mortar positions, for example. They did include a few images which they call satellite images, but I think some of those are air photos based on the angle. See http://www.state.gov/documents/organization/131025.pdf for the report, or the report of my project at: http://shr.aaas.org/geotech/srilanka/srilanka.shtml .
A final example, some work I participated in a couple years ago located "unknown" IDPs in Somalia, that then required the UN to mount a relief operation. Not evidence in the courtroom sense, but it definitely served to change policy in a very real sense.
So, in sum, its not necessarily a direct link between imagery and evidence. Like all such evidence, it depends on the situation and context, and whether the evidence needs to survive extensive cross examination. I can talk more about this if needed, just let me know.
I might be a bit iconoclast and will go in the sense that was
discussed the technology should not hide the content, (l'arbre cache la
foret) human rights violations imagery is a prof that human rights
violation occurs, more as a sensitization and awareness to the global
society and will provide advocacy at a higher level (political ?), (it is
always a scale level). So there is real need of classical,
and not
always e-information, of mapping to highlights the facts and act
locally (if possible due to the context ) so collecting information in
the ground as close as possible of the source of information , is vital
trough NGOs network. EG in DRC kivu, local NGOs are able to report
incidents and crime violation within a regular manner (trough paper
forms) to the UNHCR, big International NGOs are failing... incidents
are reported and actions are taken as quicly as possible for the
victims, (like providing PEP,and Pshychological support) but the data
are not really analysed, correlated with the other layers of
informations... for going further into the analysis. Info management
in this area is
like putting a plaster for short and immediate response, but geo
mapping will be able to reinforce the possibility to overlay the facts
of violation( SGBV, land use violation...) with the environment (presence of police, of military detachement..)
CartONG
This is of course a good point, and one that resonates with me as I generally stay in Washington and observe human rights violations from satellites. Its important to point out that there is nothing inherently protective about getting a satellite image, it is all virtual, and may not influence the situation on the ground at all. I also worry that such visual tools like imagery simply provide more media content but don't necessarily cause people to care about issues. I cringe when people tell me how "cool" our work is, because it means they are focused on the technology and not what we are documenting. How do we call looking at Darfur "cool"?
On the other hand, it meant a lot to me when some Burmese sent me a note to tell me how great it was to know we were watching and trying to help them, though we had never met. They often feel isolated and alone, so the idea of satellites in space taking images of them was something they really liked, especially as it got the issue on CNN.
As with anything, mapping and imagery is part of the broader context, and needs to be deployed or used locally in many cases to have proper impact. Otherwise we just contribute to the Washington literature circus and media charade...
Lars Bromley, AAAS-SHRP
Back at the State of Map conference in August from the OpenStreetMap (OSM) initiative, I remember an emotional debate, when one participant proposed whether the OSM project shall also offer something like Google Street View. Some mappers said that they will abandon the project the minute somebody starts such an initiative. I observed that there is a fine line of what can or shall be mapped and what not. Suddenly the whole mapping issue became quite controversial and left the technological discussion. I think it is very important to discuss the case of Google Street View as it lead to various protests. Is it always right to publish all available geo-data, or is it better to close some data?
I am not sure whether this example is entirely relevant for human rights activism, but shall there be limits too? Publishing geo-data might help one cause, but easily threatens another cause. What do you think?
I can't comment on the Street View issue, it seems it is being decided courtroom by courtroom. Regarding whether its right to publish all data, that is a tough issue and differs drastically from place to place. I remember when Google Earth started putting high-res imagery up of a certain area, the folks hiding from persecution in that area got quite scared that it would allow them to be found. From Google's perspective, they just were doing their normal stuff, but they were totally unaware (and remained so) of how publishing such imagery might impact people.
In the projects I am involved in we take extreme care to sanitize much of the data we get to remove anything that might cause persecution of specific people. We often degrade our published coordinates and redact things like witness names. We feel this is absolutely justified to avoid further harm and repersecution of certain people.
This issue of privacy is a critical issue for human rights advocates - for as you point out, the cost of such implications for those who are suffering persecution could mean people lives. Ethical questions and considerations must be considered with the advances of technology.
I'm interested to hear about how those of you involved in geo-mapping have identified what those ethical considerations are in your various uses of the technology (such as this issue of privacy) and what recommendations you can provide for others grappling with such questions.
Nancy Pearson, New Tactics in Human Rights Program Manager
<p>
Nancy, we normally work together with the<strong> protection officers </strong>(most of them have a human rights lawyers' background) and it is them who assess the risk, but it is tricky.<br />
<br />
Apart from us only dealing with the <strong>anonymized data</strong> (e.g. no names); we aggregate the data on a <strong>defined grid</strong> to not show the actual location where the perpetration happened but just a concentration of specific events. See also Yann's reply on that. <a href="/blog/new-tactics/geo-mapping-human-rights#comment-3087" class="active"></a>
</p>
<p>
<a href="/blog/new-tactics/geo-mapping-human-rights#comment-3087" class="active">Camp mapping could be used for site planning and also advocatin</a><br />
<br />
But I don't think it is ideal. So I would be very interested to hear how others are dealing with this.
</p>
<p>
Sandra
</p>
<p>
<a href="/blog/new-tactics/geo-mapping-human-rights#comment-3087" class="active"><br />
</a>
</p>
Most of the incidence mapping on human rights violations CartONG has been involved so far was on confidential data and the maps were for internal use only. Why? Because the mapped incidences were linked to refugees (or internally displaced persons - IDPs) or even subgroups of refugees (for instance women: Gender Based Violence - GBV) and there were valid concerns that public mapping would put the victim at risk as it can be traced back in such defined communities. One of the work arounds was to aggregate the data at a higher level; but even then the maps were considered too dangerous for public use.
There are really nice examples out there using web-mapping techniques for human rights mapping which would bring so much more visibility.But it is difficult to assess the risks involved for the individual. Has anyone looked into this before?
Has anyone had experiences with public human rights mapping where only certain groups of people are targeted? And yes, how was the risk minimized?
Sandra, CartONG
Thank you so much for your posting! I have always thought of geo-mapping as a way to portray the results of a situation (victims in Darfur, environmental consequences, etc.) instead of a situation (as in the case you mentioned on refugees) and risking a negative impact. It definitely provided a different perspective on geo-mapping. Geo-mapping can have a very positive influence, as mentioned throughout all the postings, but also have risks involved.
I was wondering if anyone has themselves recognized and identified risks within their own geo-mapping projects. It seems as if it would be hard to recognize the risks when your own project has only good intentions. Was it something you or your organization identified? Was it difficult to realize them? Or was it similar to Yann's experience, where a separate organization pointed out the risk?
One primary risk on our Burma project was pointed out by local project partners. In Burma, many villages at risk maintain a second, hidden set of shelters and supplies so that if the army attacks the villagers can flee to these hiding spots. When we first produced satellite imagery analysis for Burma, we made all of our imagery available via GoogleEarth and other methods. However, our local partners immediately let us know we were revealing some of these hidden villages, so we quickly modified our published imagery to only show those villages attacked, not the surrounding areas.
In other work, we've made sure to exhaustively review our reporting data before release to make sure we have no information in it that might get someone in trouble or hurt. Again, the local partners have the final decision on what we can release.
Lars Bromley, AAAS-SHRP
Thanks for the great thread of comments on recognizing and assessing risks when it comes to geo-mapping. (I hope that Sandra will receive a few replies on her question about how to maximize visibility and minimize vulnerability - this is an issue that many human rights advocates face)
I wanted to point out the different risks associated with the different types of geo-mapping tools (this information comes from Tactical Tech's Maps for Advocacy Guide):
What other types of geo-mapping tools am I missing - and what other risks regarding the use of these tools should be assessed?
Kristin Antin, New Tactics Online Community Builder
There is an amazing episode of the Brazilian series City of Men called "The Mail", where - long story short - some residents of a slum with the blessing of the drug dealers decide to put street names on the alleyways and small open spaces in the slum. Then the cops confiscate a map made with the new street names, and begin to use these street names to make their way through the favela in their cat-and-mouse combat with the dealers who run the place. (http://www.youtube.com/watch?v=cyu-DQZGEPQ)
In the episode of City
of Men, the street signs were simply changed around to confuse the
police. But in the age of all of the technology we are discussing here,
everything is instantly made available digitally, and not as easy to do
away with.
How does one "unmap" a place? Once data is published or placed in the public domain, it seems that it
would be extremely hard to take it back again.
Information is power. Maps, in the hands of those who would use them for other purposes, can be harmful. Not just in militarized or openly repressive situations. What about for example, if geomapped data meant to "protect" sacred places is actually used to better exploit these places? I am thinking of how in theory it would be "useful" to map sacred places in a country I work to prevent government and others from conceeding usufruct rights of these areas. But I can see that actually this could have the opposite effect, exposing these hidden places to more outside interference.
I realize some of these issues have already come up in your conversation... and I offer no technical insight, just questions...
Fascinating and insightful, many thanks for sharing. I'd say that the issue here is that there was just one map. One would perhaps simply need to make dozens of copies?
Very interesting question: "how does one 'unmap' a place?" Would love to hear what others think.
Patrick Meier, Ushahidi
We, CartONG, have conducted all human rights mapping embedded in an UNHCR (United Nations High Commissioner for Refugees) project; so it was either the Protection officers there or within the UN Cluster System (with UNHCR chairing the Protection Cluster in many countries), assessing the risks.
I think a Rule-of-aThumb could be: if your target group is very defined and small in comparison to the rest of the population; reported violations can be traced back easier and repercussions/revenge could be more likely. Now if you are sure you can provide the necessary protection for your target group, you might opt still to make your findings public. But that is a tricky decision; especially if the perpetrators are the ones supposedly providing security (like military etc pp).
Yes, it would be good to get some more general ideas on what rules to apply on publishing data on vulnerables.
UNHCR does have a knowledge base and some protection related material for downloading, but of course their focus are the Displaced.
Sandra Sudhoff, CartONG
I highly recommend this book on Mapping Vulnerability
Patrick Meier, Ushahidi
I would like to share a discussion I had early 2007 with a staff member of an organisation with a red cross in northern Uganda.
We were talking about mapping of the roads. Especially the opening of the new roads and well graded in order to access the returnee sites. Yes we are talking on the single tracks that come from your GPS device, the poly line that should be edited in order to join the rest of the road network, your 2 km road section from the main road to the new returnee sites with 150 dwellings.!
The staff was concern about the opening and mapping the roads. Why? The staff had a fear of easy movement of the LRA (Lord Resistance Army) . In a certain extent, the mapping of those tracks will facilitate the possibility of having again abductions and violence that we are all know. It is tricky to map information for some organisation that are interested by a “physical information” like for the NGOS taht are doing road rehabilitation,and in other hand it could brings some fear to other organisations.
We should understand of having or not mapped information that seems to be in the “public domain” and if the community or humanitarian are not interested of having presented on a maps the information related to there environment, we should respect them. How are we dealing with those dilemmas? especially in a context of the humanitarian clusters and information sharing "a tout va" where everything is shared with not so much control after all.We should really asses the impact or foot print and the potential consequences that our mappings could have. It is like the movement of the butterfly.
I’m playing the advocate of the devil, ;-)I might be ejected of the forum...
CartONG Yann Rebois
These are important issues of course, and I am wondering if they are of more concern in humanitarian mapping as opposed to human rights mapping. We are cautious with publicizing our data or satellite imagery, though this is usually temporary. After our investigation is complete we publicize pretty much everything we can. So, maybe it is all a matter of timing.
Again, we are generally not mapping roads that might enable LRA attacks or something like that. And, further, I'd never assume that my road mapping was better than any local knowledge...
Lars Bromley, AAAS-SHRP
Hi all,
It would be great for those of you doing this work on the ground, to share your experience regarding what the minimum requirements are that are needed. For example:
This would be very helpful to give a better understanding of how realistic it is for organizations to use geo-mapping to advance their efforts; and the steps organizations would need to take to make it effective.
Nancy Pearson, New Tactics in Human Rights Program Manager
Some minimal requirements for geo-mapping the end user should know what he/she would like to see on the map.
Maybe some elements about your questions If we really want to act, a bit less for advocating we have to put in place a system, system in terms of data collection and processing for a better diffusion, system for the timeliness delivery and update of the facts, it could be based on the locals with more or less risks for themselves, or international, with less proximity of the source of the info, so the information my have some bias.
In terms of system data model or SDI (Spatial Data Infrastructure) should be defined ; as of today there is several initiatives on HRV, (with different topics, it might be interesting to develop a common dataset?
Then in terms of technology, sure the web server technologies, like the Google applications, or feeds (GeoRSS, WMS Web Map Service , WFS Web Feature Services) are the most appropriated to share and to be useed by GIS specialist and advocate and also to register the records like with the SMS ; however in the field it is not always possible or user friendly, or due to IT knowledge of certain local staff member of some organization it is limited (previous experiences).
About expertise and here you point out the biggest challenge, we, techies have to facilitate and to promote the power to the experts that knows how to act with HRV and knows the mechanisms or the systems for referral actions. They will be the one that wil say “I would like to have this information as it could be use on this perspective…” For instance, we get some request from OHCR, for mapping the catchment areas of the new police post created in an area (how many villages are covered In 5km, 10 Km…radius), they were interested by the coverage of the paralegals and the police to see if there is an overlap of the 2 entities… so the specialists will come with their own language and we will have to analyze through spatial analysis for instance, the facts.
With the Google tools, I will say we are more locating or plotting the facts than processing the info… but how could the data could be turned out into real information for decision? We should have a kind of HRV wish list of geo-products that could be used and become standard for HRV. Then we could think on the data that should be collected and how…
CartONG, Yann Rebois
Dear Yann,
thank you for your comprehensive description on different approaches of mapping. Risking to be a bit superficial, I experience two main approaches so far to start a mapping project.
1) You start collecting data yourself to map your area with your topics and aims. For this you either rely on existing maps and mark your points of interest (see community mapping) or your walk around and map such points with a GPS advice for exact position. (OpenStreetMap)
2) You take existing data, which includes location relevant pieces and use it for maps. There are many available databases or own data with geo-information. There are tools such as Google Maps or Yahoo Map Maker (http://developer.yahoo.com/maps/), which offer easy tools to filter locations out of data. These tools allow different forms of easy maps through mashups.
Both above approaches focus on non-experts and still offer some quite impressive opportunities. The above described map of Iran protest during the elections is such an example.
Hi, I’m
maybe after all too “old school” ;-), ?!
It is true
that Martus is a standard. I should
reckon I had a look some years ago and the spatial component was not really
there… Since then as I did was not involved directly on HRV I didi not follow the evolution.
With Google
tools it is true that is simpler and
simpler to share the information .
Do you have any doubts when we are talking on confidential data ? CartONG also used Google Application, but nothing was related up to know on HRV, due to confidentialiy
I would
like to share some data collections forms that are used in the field in DRC (mainly Kivu) by the protection
cluster/ UNHCR to report, act and advocate.
It seems to
be old fashion system, but it is working
on a weekly basis but if you had on top of it new geo-mapping spirit it could
be quite powerful, I think. And it will be great! And they are looking for
geommaping support. As of today, They prorduce or would like to produce some
maps showing concentration of incidents.
If you look
on the system (Flow chart) you really see the complexity on the stakeholders
that will need a specific kind of information from the raw data, to the already aggregated
information. What is interesting is to
see the aim of the data/geodata 1. program, 2 advocacy, 3, recommendation 4, action and referral to the specialised
agency. About the standard, two
forms are attached as a sample. 1. 2
CartONG
Yann Rebois
Pages