A few examples of projects undertaken inside and outside of university
Open Street Map
Changes I have made to the OSM network in Kisoro as of 01/18
I have been involved in mapping Kisoro in south-west Uganda since I was introduced to OSM whilst working with GeoData. Kisoro lies 10km to the east of DR Congo, and 7km to the north of Rwanda. It assists with UN humanitarian operations in DR Congo, as well as hosting a large refugee camp which fills when violence flares in Congo's North Kivu province.
My interest in Kisoro stems from a 7 month stint I spent working there prior to studying at university. I was involved in a water and sanitation programme that sought to extend safe drinking water supplies to the more inaccessible areas of the district through provision of rain water harvesting tanks. Kisoro has highly seasonal rainfall, storing water in the wet season is a cheaper and more effective solution than wells due to the geology around Kisoro not being conducive to borehole construction.
I observed that there was poor centralised knowledge of who lived and worked in some of the remote areas. This is a niche that OSM effectively fills, using a combination of local knowledge, GPS tracks collected on Strava and satellite imagery I have been involved in extending the OSM network away from the main Kisoro-Kabale road and out to the villages
I hope to continue my contribution. There have been a number of new contributors in the area since I began. Whilst this is welcomed the edits made are of varying quality, so my focus will be on amending and improving these.
Spatial distribution of poverty
I came across WorldPop data during my time working with GeoData and was fascinated by the work that they do in creating raster datasets that show population density at 100m2 resolution across much of Africa, South America and Asia. Their methodology encompasses regular population, such as that derived from censuses and combines it with nightlights and other sources to locate where people within census zones reside more accurately - visit the WorldPop website for more detail on their methodology.
WorldPop also produce rasters at 1km2 resolution showing the proportion of a population within any given cell living in poverty. Two levels of poverty are available - one at an income of $2 a day and one at $1.25. Converting these rasters to points, and then spatially joining these points with a shapefile of each sub-county within Uganda using the mean of all the pixels within a particular polygon. Using the road network I calculated the average time it would take to travel to the nearest town. I also calculated the time it would take to travel to Kampala, Uganda's economic hub.
My results showed that there was a significant correlation between increased travel time to nearby towns and rate of poverty. Possible reasons for this include, but are not limited to; reduced access to capital for entrepreneurs, perishing of agricultural products on the way to markets, increased transportation costs. The link between level of poverty and distance to Kampala was also striking, some figures (there is some disagreement) quote the capital cities contribution to national GDP at around 65%. A lot of companies are reliant on Kampala for their supply chains, and the increased cost of transportation as well as the increased difficulty in communications associated with distance make it harder for far-flung sub-counties to develop at a rate equivalent to ones is close proximity.