Saturday, May 8, 2010

Historic Sites of Buncombe County

This project sprang from the inspiration of finding a listing of the National Historic Register sites in Buncombe County. This included a great deal of old photos of Asheville. With my love of history this was an impossible opportunity to pass up.

Data acquisition began with a KML file from the National Park Service. This file showed all Historic Register sites in Buncombe County. This file named a number of historic districts, but failed to fully define those districts. I was able to locate maps of three districts on the internet, which I loaded into ArcMap and georeferenced.  This still left a number of gaps in district definitions. In search of the boundaries of the remaining districts, I got in contact with both the Buncombe County and the state historic agencies. Both agencies were more than happy to supply their shapefiles of historic registry sites and districts. I completed my data search by contacting the D H Ramsey Library and filing a one time use agreement. My use agreement stated no publication or web postings so those pages will not appear here, but I have included a sampling of the pages that do not have the historic photos. 

One important lesson to remember, which I encountered in this project, is to know the nature of the data acquisition used and the degree of accuracy applied in the creation of the shapefile. In comparing the multiple shapefiles in ArcMap, it became obvious that there were several slight differences in the placement of points. Michael Southern, Senior Architectural Historian/GIS Coordinator at the N.C. State Historic Preservation Office was kind enough to explain the method used in creating the Site Survey used in the application for listing on the National Historic Register. He explained that the site survey is often prepared by historians who often have only a basic knowledge of GIS. They are asked to use a clear overlay, with UTM coordinates displayed on it, to determine the exact coordinates of a site by simply counting tick marks to make the determination. This leads to a result that is limited to the care taken in that process. Fortunately the NCSHPO has taken the time the time to confirm locations by using parcel records and satellite imagery. Where possible I used geocoding and knowledge of the sites to confirm the discrepancies between layers. One site yielded very interesting results. The NPS and NCSHPO showed different locations for the same site and geocoding resulted in a third location, all seperated by major geographic features. As expected, the NCSHPO shapefiles proved to be extremely accurate and most complete by defining all historic districts and became the basis of my maps.