Roland Martin is a GIS specialist in Arup’s Los Angeles infrastructure group. He recently volunteered online with Humanitarian OpenStreetMap Team (previously discussed in an earlier post) to assist in the mapping effort for Nepal. Below is a guest post by Roland on his experiences with OpenStreetMap and with the volunteering effort in particular.
A couple of years ago, I spent eighteen months living in South-East Asia. At weekends, keen to truly experience a part of the world I had dreamed of as a child, I would head out exploring, either by booking a cheap flight or just by heading out on my bike. Like all good explorers, I often got lost—in the cities, the quality of mapping data was often good, but as soon as you crossed into a rural area, the map was often completely blank. I needed a source of good mapping data which I could download onto my hand-held GPS, and OpenStreetMap was a natural fit, so I started editing it for myself. In the intervening years I’ve apparently made over a hundred thousand edits to the map.
The devastating Nepal earthquake that took place on April 25th this year has seen what might be the biggest mobilisation of OpenStreetMap editors in the site’s history. The Humanitarian OpenStreetMap Team (HOT) formed roughly five years ago to try to coordinate the site’s response to disasters and emergencies, and also to improve the quality of map data in vulnerable locations. In the aftermath of the disaster in Nepal, they have identified over twenty separate “tasks” requesting that particular features be added to the map by users. Initial passes saw major roads added and updated, while later tasks have included building footprints, residential areas, damaged areas, and other items. Within days of the event, the map went from being relatively sparse to extremely detailed, as illustrated by this animation.
You don’t even need much experience to be a “mapper”—in the last few weeks nearly 5,000 individuals have helped with the effort, and more than two thirds of them had never used OpenStreetMap before. I’ve found almost unlimited enthusiasm amongst the people I’ve shown the system—all you need is an eye for detail and a bit of time to spare.
Superficially, it might seem like a lazy way to help, as you sit in your armchair adding rivers, roads and buildings to a map, but the truth is that it has quickly grown into an enormously beneficial resource. Take this area in central Gorkha, for example:
In addition to the inclusion of buildings and roads in considerably higher detail than Google, users have also added locations of destroyed buildings (the brown areas) and a potential helicopter landing site. Furthermore, although not visible in the main map rendering, examining the raw data also reveals several spontaneous camps in the city which have been identified by comparing imagery from before and after the earthquake.
This is all extremely valuable information for the relief effort, and unlike some of the other crowdsourced mapping tools, which provide useful but short-term help, this is also a long-term resource—the data will remain available for Nepal as it recovers from these events and builds its future.
Anything crowdsourced is susceptible to various problems. Academics have suggested that crowdsourcing geographical data can actually be more accurate than commercial off-the-shelf products, but in practice this is very much dependent on the location—for every area that’s mapped to amazing accuracy, another slips off the radar and gets neglected. Secondly, there is a danger of vandalism, such as the recent high-profile case of a user’s anti-Apple statement which made it onto Google Maps, or even Google’s own contractors apparently vandalising OpenStreetMap.
The HOT’s system has a validation process in place—each tile edited by a user is cross-checked by other mappers, and the most significant omissions are usually picked up. But this is not a rigorous process, and substandard data does occasionally make it through. This leaves the team with difficult conundrum: crowdsourcing can only work if users feel they have as much right to edit and check as anyone else, but an academically rigorous checking process would require a much more hierarchical user-structure.
But for any failings there may be, the data that has been added to the map of Nepal in the last month has helped provide information for organizations such as Red Cross International, who are using the data, among other things, to map roads, estimate damage, and analyse population. Helping the mapping effort is a simple, effective way to add intelligence for the teams on the ground in Nepal and provides a way to make a big difference in your spare time.
If you want to help, all you need is an account with OpenStreetMap. Visit the HOT task manager, pick a task that looks interesting to you—there are tutorials to walk you through the editing process.
Curated topics from the April 25, 2015, Nepal Earthquake to help inform reconnaissance activities, identify impacted regions, and help document the timeline of earthquake response/recovery.
Information on Emergency Response, Social Impacts, and Community Resilience from the April 25, 2015, Nepal Earthquake.
