This weekend I’ve been doing some “humanitarian” aerial mapping in response to the floods in North India. My first time trying to map a country I’ve never visited, where the map might be used for immediate important purposes. I must say it’s a bit nerve-wracking, for those exact reasons:

• Firstly I’ve never been there and I’m not that familiar with the mountainous terrain. Sometimes it’s easy to be tracing a river and then you realise you’re actually tracing a road after all. When you’re not familiar with it, it’s much easier to get confused when tracing aerial imagery. We’re lucky that the bing aerial coverage for this area seems to be good high-resolution stuff.
• Secondly the mapping might turn out to be important. So, what if my amateurish edits lack something that would have been filled in by another mapper, if only I hadn’t filled the map up with simple-minded route-tracing (eg with no idea of which roads are major/minor)? What if I’m tracing a road from an old aerial photo, when locals would know that the road isn’t there any more? What if I’m not tracing the things that would be important for humanitarian relief purposes? What if I’m accidentally joining up roads that don’t join? etc etc etc. Harry Wood’s 2011 discussion of “blooms and weeds” in mapping summarises some ideas about whether importing low-quality data into OSM has a retarding effect on the map’s subsequent growth. Maybe rushed naive armchair mapping has a similar effect…

… See full entry

Another iteration of my visualisation of OpenStreetMap edits - here’s an animation showing, for each year 2005-2012, the density of edits according to their geographic location:

… See full entry

Here’s a data-driven way to decide where the next OpenStreetMap recruitment drive should be. Take the geographical spread of edits in 2012, and divide it by the geographical spread of the population, to see where there are most/least edits “per person”:

More details on my blog article about it.

OK, if you want to know where in the country has good pubs, how do you do it? Well, here’s what I do: download a data extract of all the pubs in the UK/Eire from OpenStreetMap, and use density estimation to look at the distribution of pub attributes such as whether it serves real ale, or food, or has wifi. That’s the normal way, right?

In my blog article about pubs in UK & Eire I’ve posted more plots, as well as the source code and some extra info.