2012 Mali Crisis

From OpenStreetMap Wiki
Jump to navigation Jump to search

Hot logo with text.svg

Following the 2012 Coup and de facto secession of the Northern part of the country with large displacement issues, Humanitarian OpenStreetMap Team is investigating activities to map Mali (Pre-Activation)

For more long term mapping and Mali OSM community news, see the WikiProject Mali page

Information about this activation

History of this intervention

  • April 2012 : Following the military coup in Bamako and the de-facto secession, the HOT contributors started to monitor the situation, to look for imagery and map various areas.
  • Dec 2012 : Following the resurgence of violence, contributors collaboration was re-activated.
  • Jan 2013 : At the request of UNOCHA-Mali and the humanitarian organizations, HOT formally activated for this humanitarian response, intensifying mapping and coordinating with UNOCHA-Mali and various humanitarian clusters providing aid in sectors such as water and sanitation, education, etc.


Pierre Béland and Frédéric Bonifas are coordinating this activation. For moer information, you can reach the coordinators by email at activation.mali, “ат”hotosm“ԁοт”org.

For discussions or help for mapping, you can reach the coordinators and collaborators on the irc://irc.oftc.net #hot.

Support Team

HOT activation in Mali covers a large territory and involve various activities in coordination with the humanitarian stakeholders. It requires the participation of many collaborators. Many contributors are supporting this activation through remote mapping with the OSM Tasking Manager to coordinate these efforts. Support team members are also doing various jobs such as find the necessary imagery, mount an imagery server, analyze / validate the database and various data files, geolocate Public services POI's, classify road network and support other mappers.

It is important to coordinate our efforts. You should not contact stakeholders in the name of HOT without the approval of the coordinators.

The following Support team contributors participate intensively to this activation, supporting the coordinators with various activities. They can be reached on the irc://irc.oftc.net #hot:

  • AndrewBuck
  • bremy
  • jgc
  • PovAddict
  • TomT5454

Discussions with humanitarians - Chat logs

There are ongoing weekly chats with the humanitarian workers in Mali regarding our collaboration with them in mapping for the crisis. Pre-meeting agendas and logs of these chats are available on the pages below:

Data download

OpenStreetMap offers many ways of downloading data as OSM XML or as other popular formats such as Shapefiles or for Garmin. These options tend to work worldwide (wherever there's data).

Set up especially for Mali, you can download up-to-date OSM data on Mali on the HOT Export Tool, including shapefile extracts. To get fresh data, please click on the "Start new run" button if needed.

An OSM file for Mali in Garmin format (imgsup.img) can be downloaded and transferred on a Garmin GPS. This GPS file will be updated regularly to facilitate the work of the humanitarian on the ground.

Documentation from UNOCHA and the humanitarian clusters

Experimentation of Crowdsourcing image recognition (2013-02)

It is often difficult to spot villages in the vast semi-desert landscape of Northern Mali. To progress more rapidly and ensure we spot all villages, user nicolas17 proposed and developed "Kuona", the Crowdsource Image Recognition application. Others have contributed to develop this application and incorporate the results into the Task Manager with a new approach. The task squares are not contiguous. They are determined by the result of the Crowdsourcing application. After one week of development, this seems very interesting for the future and we will surely continue to develop this concept:

Image collection

We use Bing High-res imagery when this is available. Extensive areas are covered. For others, we are discussing to obtain imagery. We also look for imagery to better identify the zones exposed to annual flooding.

How you can contribute to this activation

Our goal is to contribute to humanitarian relief by providing a base map of Mali including the road network, the villages and the public services. This facilitates delivery of goods and services by the humanitarian organizations and assure that remote villages are monitored properly.

Contributors can obtain support on the irc://irc.oftc.net #hot channel or by email at activation.mali@hotosm.org.

You have no GIS experience and want to contribute

  • Contribute to Crowdsourcing information from Satellite imagery. This task is simply to look at imagery and click to indicate if you see villages or dams. You can see some results of this work on this map.
  • People with knowledge of the country are using OpenStreetBugs, a simple application to add or correct on the map information about various public services (i.e. schools, health services, dams, water wells) or other significant infos about infrastructures. You comment the Reports already on the map or you add a Report to provide information. Experienced mappers will later treat this information and add to the map or correct the content.
  • The Bing Imagery layer from this map can be used to locate information on the map and report it either through the OpenStreetBugs application or by mail.

You are more experienced and want to contribute by editing the OSM database

Tagging Instructions for contributors that trace from Satellite Imagery

Tagging standards may be different that what you've traditionally tagged in OSM. Read the Mali Tagging Page to find help you identify and consistently tag features found in Mali.

Crowdmapping: you contribute by selecting a micro-task from our Task Manager

Priority areas to map
The Task Manager jobs for Mali coordinates mappers working similarly to map various priority areas.

For each job, a grid on a map divides the zone to map in micro-tasks that can be realized in an hour or two. The contributor selects a square area to map, edit in Potlatch2, JOSM or other OSM editors, and comes back to the Task manager to confirm that he completed the task. If he cannot complete, he simply unlock the task. This way, this micro-task can be completed later. Detailed instructions are given for each Job in the Workflow panel.

Verify the result of the Crowdsource application in JOSM

Experienced mappers can download the result file of the Crowdsource applicaton (GPX file) (source: Harry Wood).

In JOSM, the 'x' characters indicate where houses, villages and dams have been reported. This should not be imported in OSM but only used as a reference layer to detect rapidly the residential areas. Editors using this file should coordinate on the #hot irc to avoid duplication of efforts. If you right click on the loaded GPX layer you can change the color of the 'x' markers from the default grey to something more visible, like bright red.

Also, a handy way to quickly work these points is to zoom out so you can see quite a few, press z to select the zoom tool and drag a small box around one to trace, then press the 8 key on your keyboard to jump back to your previous zoom and repeat the process.

Import various databases, basic services utilities POIs

Various informations are necessary to complete the basic OSM map. We have various data files to geolocate such as populated places and public services (ie. health services, schools, water points, public and religious buildings). We generally discuss about these topics on the #hot irc and various contributors take responsability to geolocate these files.

We also coordinate with OCHA trying to improve this information.

Support from the local community and from humanitarians would also be very appreciated about this.

Road network: retrace major roads to assure that they are aligned with Bing Imagery

Humanitarian logistic to deliver goods and services requires to identify rapidly the major infrastructures. To respond to the two priorities identified below, experience OSM mappers are invited to look at major roads that may be used by the humanitarians NGO's traveling from neighborhood countries and across various parts of Mali.

To coordinate with other mappers, each contributor indicates below the road no. he is working on. Be careful to add bridges over the rivers and streams. When you have completed a road, confirm it in the comment column below. Indicate also portions of the road where no Bing imagery was available.

Indicate as changeset comment: #hotosm Mali [road] source=Bing

Road no From - To Contributor Comments
RN17 Gao to Niger Border pierzen 21:27, 25 January 2013 (UTC) Not covered by Bing High-Res from 15.54872, 0.5224588 to 15.5366699, 0.5589294
Task completed relation 2789200
RN15 Sévaré to Gao pierzen 15:16, 26 January 2013 (UTC) Task completed

relation 2789199

RN6 Bamako-Konobougou
RN6 Konobougou-Ségou FredB (talk) 23:16, 1 February 2013 (UTC) Bing and GPS
Task completed

relation 2789111

RN6 Ségou-Bia FredB 22:28, 27 January 2013 (UTC) Lot of work already done by meppen7.
Bing and GPS
Task completed
Rn6 Bia-San FredB 22:49, 27 January 2013 (UTC) Bing and GPS
Task completed
RN6 San-Ouan FredB 23:18, 27 January 2013 (UTC) Bing and GPS
Task completed
RN6 Ouan-Sévaré Skorasaurus (talk) 01:04, 6 February 2013 (UTC) task completed; GPS traces, bing imagery; no bing available from 13.9635, -4.1856 to 13.8838, -4.268
RN? Markala-Macina Skorasaurus (talk) 03:21, 5 February 2013 (UTC) task completed using bing; no gps traces available
RN? Markala - Diabali michael63 10:17, 4 February 2013 (UTC) Bing
Task completed
RN? Diabali - Nampala michael63 08:56, 4 February 2013 (UTC) Lot of work already done by AlkalIn, alexz and Divjo.
Task completed
Nampala - Lere pierzen (talk) 14:01, 8 February 2013 (UTC) Task completed
Lere - Niafunke pierzen (talk) 14:01, 8 February 2013 (UTC) Task completed
Niafunke - Tombouctou pierzen (talk) 14:01, 8 February 2013 (UTC) Task completed
RN? Tombouctou - Toya pierzen 01:56, 28 January 2013 (UTC) Task completed
RN? Toya - Oro-Mourou pierzen 01:56, 28 January 2013 (UTC) Work already done by AlkalIn.
GPS; a small zone not coverred by Bing High-Res
Task completed
RN? Oro-Mourou - Bambara Maoundé pierzen 01:56, 28 January 2013 (UTC) Work already done by AlkalIn.
Bing and GPS
Task completed
RN? Bambara Maoundé - Douentza pierzen 01:56, 28 January 2013 (UTC) Lot of work already done by AlkalIn.
Bing and GPS
Task completed
RN 18 or 19 Bourem - Tabakort - Algerian Border Mison (talk) 09:56, 6 February 2013 (UTC) Work already done by FredB, alexz
Bing and Orbview
Task completed
RN 18 or 19 Gao - Tabakort Mison (talk) 09:56, 6 February 2013 (UTC) Bing and Orbview
RN? Sévaré - Bandagara - Sangha
Anefis - Kidal
RN 20 Niger border - Ménaka Skorasaurus (talk) 22:33, 15 February 2013 (UTC) Work mostly done by numerous users; bing imagery
RN 20 Ménaka - Ansongo Skorasaurus (talk) 03:15, 21 February 2013 (UTC) complete, bing imagery covered by clouds at 15.8874, 1.6507; no bing imagery from 15.78, 1.15 to 15.82, 1.274; 15.68, 0.72 to 15.68, 0.767; 15.68, 0.52 TO 15.68, 0.56

relation 2789262

RN 8 Gao - Bourem relation 2790249
RN 14 Somadougou - Nandoli Xapitoun (talk) 17:22, 2 March 2013 (UTC) relation 2795781

Monitoring and mapping done in 2012

April 2012

December 2012

  • Tasks manager jobs with Bing imagery for various areas.

April 2012 : cities

These priorities were identified in 2012. In january 2013, various Task Manager jobs are defined

City Importance Imagery to use Imagery offset Comments Source
Bamako High Bing source=Bing
Tombouctou High Bing none Many streets already mapped source=Bing
Mopti High Bing Orbview-3 is available but much older (Offset -20.00; -0.00) source=Bing
source=OrbView-3, USGS, 2005-03-31
Ségou High No imagery for now
Gao High Bing None Bing Imagery shows hundred of flooded islands in the Upper Niger Delta during rainy season. These should be tagged as islands. Relations are to be created with riverbanks (role=outer) and islands (role=inner). Residential areas are also to be traced. source=Bing
Kidal High Bing source=Bing
Diabaly High Bing source=Bing
Konna Medium Bing source=Bing
Djenné Medium OrbView-3
source=OrbView-3, USGS, 2005-03-31
Kayes Medium OrbView-3 (PAN and MS, see below)
Léré Medium Bing source=Bing
Douentza Medium Bing source=Bing
Niafunké Medium Bing source=Bing
Goundam Medium Bing source=Bing
Diré Medium Bing source=Bing
Bourem Medium Bing source=Bing
Ansongo Medium Bing source=Bing
Anefis Medium Orbview-3 Bing aerial imagery available in the surroundings of the city source=OrbView-3, USGS, 2006-08-02
Ménaka Medium Bing source=Bing
Andéramboukane Medium Bing OrbView-3x source=Bing
source=OrbView-3, USGS, 2005-03-31
Aguelhok Medium Orbview-3 source=OrbView-3, USGS, 2006-05-02
Tessalit Medium Bing source=Bing
Tinzaouaten Medium No imagery for now



Please indicate here any imagery that can be used to map Mali.

Bing (for rivers and flooded area)
High resolution Bing imagery was first limited to the cities of Timbuktu and Bamako. In june 2012, many other areas start to be added. In many areas (see east of San as example), we see large portions of the territory that correspond to water at low levels of zoom and flooded areas at high level of zooms.
These flooded areas during rainy season should be remapped. The riverbank and islands should be traced correspondingly to the new Bing highres imagery. Observe farmlands and residential areas showed at high zoom levels. Large proportions of the territory in the Inner Niger Delta around Mopti and along the Niger River are under water during this season.
Other imageries like Orbview show farmlands in these areas.
Orbview 3 (for farmlands)
1464 OrbView-3 images (mostly 1 m resolution, panchromatic ("pan"), and a few 4 m resolution, multispectral ("ms") of Mali are available, as TMS.
  • For JOSM: tms[19]:http://osm.arkemie.org/cgi-bin/tiles/1.0.0/mali/{zoom}/{x}/{y}
  • For Potlach: http://osm.arkemie.org/cgi-bin/tiles/1.0.0/mali/$z/$x/$y
They were processed at level L1Gst by USGS, and have less than 10% cloud cover.
Like for Bing, the usual warning about adjusting the offset with GPS tracks before tracing from imagery applies.
Each scene covers about 200 square kilometers. When there is an overlap, the most recent image goes on top.
In case of overlap between "pan" and "ms" images, the "pan" images, which have better resolution, go on top. If you wish to access "ms" images anyway, they are also available in their own layer: replace "mali" with "mali_ms" in the TMS URLs:
  • For JOSM: tms[19]:http://osm.arkemie.org/cgi-bin/tiles/1.0.0/mali_ms/{zoom}/{x}/{y}
  • For Potlach: http://osm.arkemie.org/cgi-bin/tiles/1.0.0/mali_ms/$z/$x/$y
Use source=OrbView-3, USGS, 2003-2007.
Public domain maps

Context info

Cultural info


For help using OpenStreetMap, or other requests/information:

  • hotosm, “ат”gmail“ԁοт”com – Official contact address for (private) communication with HOT.
  • HOT mailing list – For public discussion.
  • IRC chatrooms include a irc://irc.oftc.net #hot channel – Though #osm is always busier.
  • See Contact for other OpenStreetMap channels.