Geography 5223 - Elements of GIS: Part II
Final Project: Pennsylvania Archeological Sites Database

Group 10
John Taber (project coordinator), Jim McLay, and Chris Yeakle

Introduction: The Pennsylvania Historic and Museum Commission (HMC) maintains 30,500 archaeological and cultural sites in the state. These sites' locations were drawn on approximately nine hundred mylar overlays, hand-drawn using USGS 7.5-minute quadrangles as a base. They have not been converted to digital form.
For this project, we were assigned three quadrangles (Avonmore, Slickville, and Vandergrift) in the western part of the state. The process we have incorporated began with warping the JPEG scans of the mylar overlays to the UTM coordinates used by the Digital Raster Graphs (DRGs) of the USGS quadrangles.

Procedures: These examples show the process involved with the Slickville quadrangle.

Figure 1: This is a scan of the original mylar for the Slickville quadrangle.

Figure 2: This is the same image, adjusted through the ImageWarp extension of ArcView, to line up with the DRG for Strickville.

Figure 3: This screen capture shows the digitized archaeological sites, extracted from the warped JPEG image, overlaying this image.

Figure 4: This screen capture shows the same sites overlaying the DRG.

The sites digitized from the Strickville quadrangle were combined with those from Avonmore and Vandergrift.

Figure 5: The sites from all three quadrangles. (Strickville is in the lower left.) Once the sites were consolidated into one set of shapefiles (one each for points, lines, and polygons) these shapefiles were projected from the UTM coordinates of the DRG to geographic coordinates.

Figure 6: The final task for this project was to join with an existing table to determine which sites were on a slope less than one degree. This screen capture shows those sites.


These three sheets took the three of us, from start to finish, each about two hours. To complete the entire 900-sheet project would require about 1,800 man-hours.

The RMS error was never more than twenty meters on one axis. Because sheets were warped by different people at different times, there might be a slight scale error between JPEGs - this would not be considered a warp error. Other error is associated with the standards of the input maps, and the accuracy and precision of the digitizing.


Hosted by