Lately I've been working hard to improve my skills with Geographic Information Systems. As a Planner, GIS is a critical tool for researching, deconstructing, and analyzing human settlements. I've been using GIS for several years, yet was never confident in my ability to utilize the software packages or the datasets. I could do the work, but it was never intuitive. Fortunately that is beginning to change. Recently, GIS has taken on a new role in my life as I've been using it to determine and model advance indicators of insecurity. While there are plenty of competing organizations and individuals out there hoping to find ways to asess the probability and locations of conflict before it happens, the truth is, all these systems are bulky, expensive, slow, and not feasible for an individual user. Yet there is a demand among individual users and so my goal is to create a reliable statistical tool for common individuals with basic internet access, not to reinvent the wheel of security and defense.
|Image via Spatial Networks|
Surprisingly, the biggest obstacle hasn't been acquiring real-time data. Thanks to recent developments in social media, it has been remarkably simple to acquire and filter information on recent events as they happen. When a problem takes place in Kabul, I have full details on that situation within seconds and after only a matter of minutes am able to fully assess its scale and location. Knowing when and where things are happening is the easiest part.
|Google Maps - Map of Kabul|
Instead, the biggest challenge has been the acquisition of a useful base map. In short, maps of Kabul are terrible. Take for example this map acquired from Google. You will notice that the streets outlined in yellow do not remotely correlate to the actual roads in the satellite image. Someone should be fired for this.
|WikiMapia - Map of Kabul|
Other typical map options are equally limited. In the past I've been a solid user of Wikimapia, as it allows individuals to upload information and draw vector-boundaries around areas of interest, so its useful for studying remote geographies. It has been of great value when studying Somalia, and it is clear from the example that Wikimapia is densely loaded with relevant information in Kabul as well. Clearly this is better than Google, yet it has one majore flaw, it does not allow one to export the maps into any useable format.
Options do exist out there in the world to obtain high quality geographic data on Kabul, such as found through Spatial Networks, but if you are like me and must do the work with a limited budget, options are slim. Using ESRI's online ArcExplorer, I was able to pull up a collection of maps for comparison. Although they look suitable in the small examples to the right, once you actually begin to zoom inward, all feasibility of use at street-scale is lost. Bummer.
Today I made the breakthrough and found the winner to be OpenStreeMap.org. It functions basically like google earth, allows one to customize the map like wikimaps, but best all, allows the user to export the map as an XML file. The result is that I can integrate this map with my datasets and an actually useful product is in the making. I'm excited about the prospects of this new tool and look forward to sharing updates on its development in the near future. If any other GIS users out there have insight on ways to obtain useful data and maps for less-documented places like Kabul, feel free to send me an email or something - I'm always looking for new information.