[HOT] Guinea Ebola Outbreak, Mapping the broader area
jgc at arkemie.com
Sun Apr 6 15:13:40 UTC 2014
The Landsat-8 mosaics have been extended again, to 35 images, towards
the South and East, and now cover all of Guinée, Sierra Leone, Liberia,
western part of Côte d'Ivoire, to the city of Bobo-Dioulasso in
Burkina-Faso (Fofana ;) - unrelated to Ebola outbreak.
For ease of use, the URL are still unchanged. You may need to refresh
your cache to view the new version.
Le 29/03/2014 00:56, Jean-Guilhem Cailton a écrit :
> Dear All,
> The Landsat-8 mosaics have been extended to 22 images, and now cover
> all of Guinea and of Sierra Leone (in addition to the same parts of
> Liberia as before, and border areas of other neighboring countries).
> The URL are unchanged. You may need to refresh your cache to view the
> new version.
> Best wishes,
> Le 28/03/2014 15:14, Andrew Buck a écrit :
>> Hello again everyone,
>> The task manager jobs for all of the main cities are coming along very
>> nicely. I have been amazed at how quickly the community responded to
>> these jobs.
>> In addition to the jobs in the cities, there is also some mapping that
>> needs to be done over a wider area for context. In order to help
>> understand, model, and predict the spread of the disease it is useful
>> to have population data for all of the villages and also road networks
>> to help understand where infected people who travel might have
>> traveled through. We can help with both of these issues using the
>> combination of what high resolution imagery we have (Bing, Pleiades,
>> Orbview, etc), as well as some very recent LandSat 8 imagery over the
>> entire affected area in 2 different color combinations. These Landsat
>> layers were put online by JGC and show natural color in the "432"
>> image and a traditional "false color" image in the 753 layer listed
>> below. Both of these image layers were taken at the exact same time,
>> however the 753 layer shows the area in 3 different infra-red colors.
>> By using these two layers together it is possible to tell apart
>> objects that are the same color (like a village with brown roofs, vs a
>> brown field). In these images villages will appear brown in the 432
>> layer and bright pink in the 753 layer.
>> The best way I have found to use these is to load them both into JOSM
>> with the 753 layer listed above the 432 layer. Then, turn down the
>> opactiy (transparency) on the 753 layer to about 50 to 60 percent and
>> you should see the villages look quite apparent. To start with just
>> try this in an area where a lot of villages are already mapped to get
>> a sense for how to set up the imagery layers and what to look for.
>> Basically, the villages will be a bright pinkish hue with a kind of
>> "shimmering" look from the 432 layer underneath.
>> Hopefully we can get maybe 5 to 10 people working on this over the
>> whole area we have Landsat for. I don't think it is worth creating
>> such a huge job on the task manager for this since it is such a broad
>> area. Just try to upload (and update) once every 1/2 hour or so.
>> Also, it would be a good idea to install the Geochat plugin for josm
>> and login to it as well as right clicking on the Geochat text box and
>> clicking 'show users on map'. This will let you see other Geochat
>> users in the same area. If all the people who work on this task use
>> that setup the risk of conflicts is basically 0. And finally, I am on
>> mumble if you have any questions about the setup of what to do. It
>> would be cool to have a room of people working on this together so you
>> can all share tips on there.
>> Anyway, thanks for all the help so far and let's keep going on this
>> new stuff as well. If you plan to work on this, please just post a
>> short message to the list here when you start so I can get a sense of
>> how many people end up doing this. It is a bit tricky to set up, but
>> is a really fun thing to work on once you have it all set up.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the HOT