Guest post by Mike Grasso, Environmental Studies ’13
Amanda Wooden is in preparation for the publication of a book entitled Another Way of Saying Enough: Environmental Protest & Conflict in Kyrgyzstan. The book explores the environmental disputes taking place in Kyrgyzstan. Professor Wooden wanted to utilize a number of maps of Kyrgyzstan found in a 2006 atlas to analyze the spatial relationship between between localized issues of public concern, the distribution of collective action addressing these concerns, and the proximity to potential hazards. These maps of Kyrgyzstan, however, are not as accurate as current 2011 maps.
The G.I.S. team’s job was to georeference and vectorize these maps. The first step was to scan the maps from the atlas so that they could be accessed online. The electronic copies were then downloaded into ArcMap as raster data. A base map downloaded from the GADM database of Global Administrative Areas was already added into ArcMap, so the atlas maps had to be scaled down to be the same size as the base map. Then the maps were aligned as best as possible so that the georeferencing could begin. W e looked for outstanding, unique geographical features (ie. large lakes, peninsulas, rivers, etc.) and then added control points. Control points come in pairs and are the georeferencing tools that do the stretching and adjusting. The first point is placed on the distinguishable feature on the incorrect map and the second point is placed on the same feature on the correct map. The incorrect map will automatically shift after selecting the second point. Each section usually takes three to six pairs of points to correct them.
After georeferencing, the map legend points were vectorized. Vectorization is the process of taking raster data and converting it to vector data. In order to do this, we zoomed in on the map and panned through the entire map clicking on every legend point on the map, assigning a different shape to each different environmental hazard. Some of the map legend points that were vectorized are flooding-prone areas, rock-fall prone areas, avalanche ar eas near roads, and mudflow and floods hazard. The map legend points were vectorized because vector data is able to be edited and used to run spatial analysis queries. For example, with the m
ap legend points as vector data, we could select a point on the map that could represent anything (a city, town, ski resort, etc.) and run a location query to find out how many potential avalanche risk site there are within a 5 mile radius of that point.