Talk:2010 Chile earthquake/Imagery and data sources

From OpenStreetMap Wiki
Jump to: navigation, search

Contributions of Willett and Nicholas

information originally appearing on the main page, relating the origin and context of B. Willett's contributions --Ceyockey 11:27, 28 February 2010 (UTC)

(2010-02-28 - 0:38 UTC)

Thanks to Bruce Willet in Chile for this first snippet via slow modem;
he will be downloading more overnight..

http://maps.geography.uc.edu/~cgn/maps/Chile/vectors/

Chris Nicholas

UN-SPIDER

(2010-02-28 - 2:15 UTC)

Thank you Bruce ! We have streets for Santiago, Talca, and Puentes as
personal geodatabase .rar file in the same area, and more uploading
overnight.

The *precise* lineage of this data is unknown, but I understand there
is no problem sharing it....

Chris Nicholas

UN-SPIDER

(2010-02-27 - 22:39 UTC)

I have been thinking about it for a while!!  I think my Punta Arenas data is legit, as I have worked for local / state gov for a bit

Bruce D. Willett

GIS Specialist              Punta Arenas, Chile

Description of contents

(added to main page - prior content here removed upon appearance on main page) --Ceyockey 13:05, 28 February 2010 (UTC)

Use

I might just not be using my tools correctly ... I opened the .shp file in ArcExplorer 9.3 (Java viewer) and I couldn't get information to appear as maptips, such as names / descriptions. I had all of the related files in the same folder. Should a newer version of ArcExplorer be used, or is there something preparative which needs to be done to view file contents in ArcExplorer. (this could have gone onto the email lists - and probably should have). --Ceyockey 13:05, 28 February 2010 (UTC)

I used QGis to open the shapefiles and have a look at their content. --Jgc 23:21, 1 March 2010 (UTC)

Is JAXA/ALOS imagery good enough for some things?

Chile jax imagery shot 188mRes c-32 92712by-70 68092.png

This is a shot from inside JOSM showing 188meter resolution around -32.92712x-70.68092. Perhaps good enough to note that there is a roadway bridge present and to trace some roadways and waterways? I agree, though, that the highest priority tracing and annotation isn't possible with this imagery. The roadway is visible under the drawn element and gps. --Ceyockey 18:02, 28 February 2010 (UTC)

Misleading information about pre-quake JAXA/ALOS imagery

I just noticed that there are now two sections with information about JAXA/ALOS imagery. One of those is post-quake, another one is pre-quake. It seems as if much information about the post-quake images have been copied from the section about the pre-quake data. This would be ok, but there are errors:

  • The text about the pre-Quake images says: a satellite image of Chile after the earthquake from JAXA.
  • Also, it reads: Note that the resolution is nearly as bad as the landsat imagery, so you cannot identify actual buildings... I was the one who originally said that about the post-quake-data. Has somebody confirmed that the pre-quake-data has the same low resolution? (I understand that it is reasonable to assume this, but anyway, I think it should be checked.)

--Brian Schimmel 11:40, 1 March 2010 (UTC)

Hello Brian,

(I did not do the copy&paste job .. just commenting) I just took a look at pre and post quake imagery in the same resolution and view. They both do seem to have the same low resolution. Also, they are out of register by a small amount (~10 meters or so, bad guesstimate) which doesn't make much of a difference, but if you want to do toggle-scanning (flipping between the two rapidly), you'll want to fine-tune the registration in JOSM. Also, the pre-quake imagery seems to be darker than the post-quake. Wondering if we can feed this back to JAXA by way of request to better match the two datasets? I'm not sure what is possible in this area.

--Ceyockey 11:55, 1 March 2010 (UTC)

International Charter Imagery

There seems to be a whole lot of useful information on the linked page, but to me this is very cluttered and I can't find my way through it. Can somebody expand the information in this wiki and point to the really usefull stuff? I think useful would mean:

  • High resolution satellite imagery taken after the quake, acessible via WMS
  • High resolution satellite imagery taken before the quake, acessible via WMS (still usefull for areas which are not covered by high-res Yahoo material, because they are poorly mapped)
  • Geo-referenced information about incidents like colappsed buildings, impassable roads, etc.

But right now all I can find are a dozen maps with circles to indicate the epicentres, and web-clients for satellite imagery that is as worse as landsat, and pdf files which are smaller than 400kB and thus will not contain that much information.

Also, it states For additional resources and imagery requests, please contact spaceaid@unoosa.org I'd like to know if someone from OSM has already contacted them, and told them about our needs and possibilities. I will come back to the computer in two hours, if there still is no more information, I think will contact them and document the communication right here. --Brian Schimmel 19:08, 1 March 2010 (UTC)

Oh, I just see it says Clarification on re-usability of Charter imagery within the OSM project has been requested. Target area requested from the Charter are the following: 1. Concepción - Lat 36° 44’ 37” S; Long 73° 06‘ 33” W, 2. nearby Coronel - Lat 36° 58’ 39” S; Long 73° 10‘ 23” W. So someone asked about the re-usability, but did someone ask if there is even more information? I'm not sure what audience the page is targeting, but if they had for example a high-res WMS, would they put it on the page? --Brian Schimmel 19:26, 1 March 2010 (UTC)
You may have a look at http://groups.google.com/group/crisismappers-chile for updates. --Jgc 23:25, 1 March 2010 (UTC)
Thanks, Jgc, there's some nice information there, although until now, nothing really striking, since no high-res-map is available. --Brian Schimmel 01:07, 2 March 2010 (UTC)
It's me, once again. Above, I said I would contact spaceaid@unoosa.org if nobody indicated that they were already contacted. I read the article once again and this time it seems to me as if somebody had already contacted them, though I am not sure. I already drafted an email, but be aware that I did not send it and probably won't do because now I believe somebody already asked. But just in case someone else likes to contact them or contact anybody else, I'm sharing that draft here:

Dear ladies and gentlemen,

as one if the many volunteers working on OpenStreetMap, I want to ask about available satellite imagery from the areas in Chile which are affected by the recent earth quake. We are aware of the information at http://www.un-spider.org/page/3287/spaceaid-available-space-based-information-earthquake-and-tsunami-chile but since it reads "For additional resources and imagery requests, please contact spaceaid@unoosa.org." we are asking if there is anything more than that.

We aim to manually trace such imagery to create up to date maps, which are then provided to local rescue teams, etc. This has proven useful for the recent disasters in Haiti, but for Chile, we currently lack useful sources. We are currently aware of the sources mentioned at http://wiki.openstreetmap.org/wiki/2010_Chile_earthquake/Imagery_and_data_sources which include many low resolution images ( > 6.5m ground resolution) but include high resolution material only for Santiago which is believed to be about 10 years old.

As far as I can tell we currently could use the following kind of satellite imagery:

  • Mid resolution from the affected areas in Chile, even if it is outdated (to have a source for the road network, etc., independent from current damages). By "mid resolution" I mean high enough to recognize the position of residential roads, etc. Such material for Santiago is of limited use, since we already have that.
  • Current high resolution images from the affected areas, including, but not limited Santiago. By "high resolution" I mean high enough to recognize damages on buildings, roads, bridges and other infrastructure, as well as emergency camp grounds, etc.

Most of our volunteers are able to use satellite images delivered via a WMS-Server, but there are always some professionals around who can use and convert any other common data format.

We would be glad to be updated on any new data sources.

With best regards,

Once again to clarify: This email has NOT been sent. --Brian Schimmel 01:07, 2 March 2010 (UTC)
Yes Brian, I think they are aware of the needs. It is my understanding that the present limitations come from the image suppliers. --Jgc 21:37, 2 March 2010 (UTC)

Geoeye

Any news about high-res Geoeye imagery for Chile? --ALE! 09:17, 4 March 2010 (UTC)

The WMS-server mentioned now works in JOSM (have not tested Potlatch). I've added a way [1] to show the area covered by hires imagery around San Javier. Comparing with existing GPS tracks, the images seem to be off by a few meters, so try to adjust them first. Do we have a statement about legal usability for OSM tracing yet? --Lyx 08:15, 12 March 2010 (UTC)

Eros-B

Did someone use the Eros-B WMS with success? I get "Server returned HTTP response code: 500 for URL: http://irs.gis-lab.info/?layers=eros-b-chile&bbox=-72.9775883,-36.7408756,-72.9718592,-36.7351465&srs=EPSG:4326&width=500&height=499". --Lyx 05:14, 5 March 2010 (UTC)

I have used it successfully, ending a couple of minutes ago. --Ceyockey 13:12, 5 March 2010 (UTC)
Works indeed; turns out that I originally tried it in the area of Concepcion instead of Constitution.--Lyx 19:29, 6 March 2010 (UTC)
For me both areas do NOT work. I just use the direct launch link for Potlatch but I do not see any sat imagery. Anybody else has the same problem? --ALE! 17:58, 7 March 2010 (UTC)
I was referring to working in JOSM; sorry for not considering Potlatch. --Ceyockey 00:17, 10 March 2010 (UTC)

resolution

I think it would be useful to add an additional column resolution in order to identify the quality levels of each imagery source (could also be something like < than source X or an approximation if no exact value is available). At least we should include the information in the column image set. Comments? --Marc 10:42, 5 March 2010 (UTC)

Maybe some functional values like "building tracing", "major road tracing" and "variable", the latter to indicated that there are some areas suitable for fine work and others not. Also, it would be useful to indicate where cloud cover makes imagery useless ... maybe this should be part of the boundary relation? --Ceyockey 13:13, 5 March 2010 (UTC)
Yes please. I haven't really had time to look, but if we now have imagery which is good enough to do some serious tracing then there's lots of pages which need updating to make that clear. Wiki tip : DELETE the old sentences which are no longer true. :-) -- Harry Wood 14:49, 5 March 2010 (UTC)
I've started to include more information about imagery utility in OSM Trace Matrix; for instance, see http://projekte.eiops.de/osm-matrix/?zoom=16&lat=-36.62875&lon=-73.10852&layers=B0FFFFFFFFFFFT (square 15/9729/19971) and nearby. I've only done a few squares, but in principle, this could be done for all squares in a region. For practicality's sake, it would be useful to be able to select a group of squares and update information for all of them at the same time ... this is not in the current functionality of the tool, though. --Ceyockey 14:12, 6 March 2010 (UTC)

Budget for disaster images

For Haiti and Chile it took about two days until the first high res images were available for OSM tracing. When the high res images finaly was available it took about one day to trace the images.

Technicaly the pre quake images could be online a few minutes after the quake. The satelite image producers want to earn money and it's very logical, that they can't give away all images for free.

To improve speed there should be money available to buy satelite images. There must also be some people that deside what to buy and they should be able to response very fast in case of emergency.

--Robotnic 09:38, 6 March 2010 (UTC)

My understanding was that so far we haven't paid any money for these sets of imagery. Some of them may have only been released due to direct prompting by people from OSM, but I think many were just released by the imagery companies for free, for anyone to use (but with attribution requirements). I think with some of the Haiti ones there was a delay introduced while Mikel and others checked that it was OK for us to do our tracing from them. This needs to be checked since we're releasing with CC-BY-SA2, so no commercial restriction, also our type of attribution (wiki page mentions + loosely enforced 'source' tags) may not meet their requirements.
Maybe we should be looking into having pre-arranged agreements (covering future crisis situations) for contact channels, and agreement that we are allowed to do vector tracing with our license.
I imagine we should also try to give a mention of these imagery companies when we're doing blogs/write-ups/presentations about the crisis mapping. Also on map outputs even (see my static images for an example of this) though we must always be clear that this is not required attribution. That kind of promotion will encourage them/others to be forthcoming with imagery.
Could also look at spending money. The only time we've bought imagery so far was for WikiProject Palestine Gaza I believe.
More information in Mikel's Humanitarian OSM Team/Haiti Strategy And Proposal#Imagery. He makes the very good point that funding may be beneficial in helping restarting OpenAerialMap.
-- Harry Wood 19:41, 6 March 2010 (UTC)