[HOT] Valparaíso, Chile Fires Response / Respuesta a Fuegos en Valparaíso, Chile

Banick, Robert Robert.Banick at redcross.org
Wed Apr 16 12:47:01 UTC 2014

Fantastic stuff Andrew. Julio, may I ask whether anyone from OSM Chile is in touch with ONEMI or any other response agencies in Chile? Surely Andrew's analysis would be of value to them.

Robert Banick | Field GIS Coordinator | International Services |  American Red Cross
2025 E Street NW, Washington, DC 20006

-----Original Message-----
From: Andrew Buck [mailto:andrew.r.buck at gmail.com] 
Sent: Wednesday, April 16, 2014 1:45 AM
To: Banick, Robert; HOT at OSM (Humanitarian OpenStreetMap Team); Julio Costa Zambelli
Subject: Re: [HOT] Valparaíso, Chile Fires Response / Respuesta a Fuegos en Valparaíso, Chile

Hash: SHA1

Hey everyone,

I spoke for a couple hours on skype screenshare with Álvaro Monares who is local to Chile and we made some progress on situational awareness for this response.  Most importantly, there is a satelite image posted online taken the 13th of April (right while the fire was burning but before it got to the populated areas).  Although the license of the image is not suitable for tracing in OSM I did load it up in QGIS and traced 3 very rough polygons around where the fire area burned as of the 13th.

I then uploaded these polygons to a umap layer that we can use to plan out where to focus our mapping efforts.  The map below shows the definite burned areas in red and the areas we are pretty sure burned on the 14th in yellow (click on the areas for more details).  You can also see the blue dotted lines showing the two task manager jobs already created for this job (click on the blue dotted lines for links to these jobs).

Here is the link to the umap:


We also downloaded several Landsat 8 images to try to map landuse and to see what could be seen.  There is a mostly cloud free image (at least for the area we are interested in) from about a week before the fire but that still just shows the forest and whatnot the same as we see in the Bing imagery.  More interesting is a LandSat 8 image taken the 15th, however there is a cloud mass covering exactly the area we are interested in for that image.  The next Landsat pass will be on April 22nd so we will have to wait a few days before that imagery becomes available.  We should monitor this situation and process this image as soon as it becomes available.

We also checked through the MODIS daily fire map data and landuse data, however this is too low resolution to be of use (500 meters per pixel).

Finally, we downloaded the ASTER elevation data for the area.  Since the areas that burned are very hilly terrain with lots of steep canyons and valleys, there will now be a very high risk of flash flooding, soil erosion, and landslides in the area where the fire burned.  I used the ASTER elevation file to produce a slope map and then downloaded the OSM building data and for each building I had QGIS calculate the slope of the ground that the building sits on.  This will allow us to see where there are buildings in areas at risk for landslides and we can do a similar analysis to see where buildings might be at risk of flash flooding.  These high risk areas should receive special attention in the mapping process to help recovery teams work safely in the affected areas.

Good work to everyone so far on the tracing.  The first job is almost completely done and the second was at about 60% I think.  I am going to bed for the night, but I will take another look at the QGIS analysis tomorrow to see what I can put together for a risk assessment.

- -AndrewBuck
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/


More information about the HOT mailing list