Greetings,

I am assisting a PhD student who is working on a GIS project quantifying the amount of built structures within certain high risk zones in several small municipalities in Chile. Neither he nor I have found building footprint data (OSM has some, but very incomplete coverage in these areas), and now he’s looking at heads up digitizing –which seems feasible because most of his zones are rather small and not densely developed… a test of a zone ended up having <1% human-made structures… Others could have up to 15% maybe.

At any rate, the snag is finding imagery at high enough resolution to accurately digitize. I have searched USGS offerings on Earth Explorer, but these are not current enough nor at high enough resolution. SPOT has 1.5 m resolution imagery of the study area from this month, but it is $440 per image. My institution has some sort of partnership with DigitalGlobe that I’m trying to dig up more information on, but I think access is limited to certain people.

So, would anyone happen to know of alternatives for high-res imagery in that part of the world? Is there a source I'm forgetting? Perhaps there is a source of aerial photographs in Chile that I’m not aware of?

Another idea is to just use Google Earth Pro, which could be used for the heads up digitizing, and it has SPOT imagery for the study area dated within the past 6 months –it looks pretty good. Google says people can use Google Earth for non-commercial purposes. Does anyone have experience using imagery from Google Earth for research purposes? On the other hand, drawing polygons in Google Earth is not optimal: when you add a new individual polygon it is a separate layer. The end result would be exporting a bunch of single building footprints as individual KML files and merging them all into one layer in Arc or QGIS –doable but suboptimal. 

Along those lines, the imagery available as a basemap layer in ArcMap is dated (DigitalGlobe from 2010). We found new structures in the study areas that had been built since then.

Or maybe heads up digitizing is not the best option… I’m open to suggestions.

Thanks,

Phil White

_____________________

Philip B. White
Earth Sciences and Environment Librarian, Assistant Professor
Jerry Crail Johnson Earth Sciences & Map Library
University Libraries

University of Colorado Boulder
184 UCB
Boulder, CO 80309

303-735-8278