2013 Pakistan Earthquake

The 2013 Pakistan Earthquake took place 24 September 2013, and the Humanitarian OpenStreetMap Team responded, making some improvements to the map data.

The epicenter was 60km north-east of Arawan around here on the map.

The description of the Earthquake by USGS Earthquake Hazards Program can be found here. A Wikipedia article has been created for the event; see wp:2013 Pakistan Earthquake, available in 18 languages as of early October 2013.

Find some pictures of the disaster here: bbc.co.uk

This strong earthquake struck the sparsely populated Balochistan province (map; en:Balochistan (Pakistan)). Large sections of Awaran town (map; en:Anwaran) are reported to be destroyed. Houses are often build with mud and bricks and did not resist to the earthquake.

In the mountainous region around the epicenter, the population is dispersed in agricultural valleys oriented north-south along the rivers.

Using our maps

We have maps of Pakistan! Browse this Overpass-Map that highlights landuse and buildings to get a feel for what we have. It shows how residential areas are distributed and can be used to find where to trace roads.

Geofabrik downloads for the whole of Pakistan (as Shapefile or OSM files)

Improving the maps

This area has reasonable bing imagery coverage, revealing a fair amount of scope for improving the maps of the area. Look for

• towns and villages - broad outline landuse=residential areas
• missing roads - They're pretty much all unpaved tracks, but can be mapped as roads. Try to find the main tracks leading up each of the valleys.
• fords - We should probably label a section of ford=yes where these tracks cross/follow the riverbeds although the rivers look like they must be dry most of the time
• rivers - Lots of clearly visible river beds although they look mostly dry. Could map the more major valley bottom rivers. There's also some basic OCHA river data in place which can improve using bing.
• Village buildings details and walls. - Not the priority. but see following task manager job.

Task manager job

If you're not sure what to work on, you could fill in details in these task squares:

http://tasks.hotosm.org/job/316

This is based on settlements we've already identified. However this isn't actually the priority, and if you're happy to hunt down missing data, then this isn't the best use of your time. It may be better to go looking for missing villages and other data in Balochistan outside of these task squares.

Those working on this task have been mapping every building and every track between buildings, and every wall within the villages (As discussed on the mailing list) It's all potentially useful map data but it is probably better to go looking for more missing villages, and missing tracks between villages across a wider area.

Advanced

GEOnet Names Server place names are presumably poor accuracy in the area, but probably good enough to manually match. Someone take the lead on that?

2010 07 Pakistan Floods/Imagery and data sources/DMA Maps Mosaic 1980s raster map source might be useful for this?

Mapping Zalzala Koh

An interesting curiosity of this earthquake was the formation of a new mud island on the southern coast named Zalzala Koh. wired reported new satellite imagery of the island, and yes, inevitably somebody has added the island to the map, initially as a node then later as a way:

http://www.openstreetmap.org/browse/way/239644298/history

Using the imagery from NASA

We're assuming that this new imagery can be used for deriving open licensed vectors. Possibly not technically correct since most "from NASA" imagery is actually licensed by commercial organisations. Anyway it would be good if somebody walked around the perimeter of this new island recording a GPS trace, and provides this to an OSMer (with open license) or uses it directly to add to the map (know anyone local who could do this?) Since the georectification of imagery is likely to be a bit off.

We'd also like to get a look at where villages are damaged by the earthquake. Progress has been made with this kind of thing, but clearer more rapidly actionable agreements with imagery providers could help inform HOT crisis mapping activities. More timely and accurate maps post-disaster, could save lives.

HOT email list communications

one link to each month's archive where email discussion takes place, then a listing of threads by title of first posting

• 2013 September
  • "Earthquake in western Pakistan kills hundreds and creates island off coast"
• 2013 October
  • "Walls in Pakistan villages"

there are likely other threads which contain information relevant to the mapping response for this event; add more information here if you know of these cases