GIS5935 M1.2: Standards

Figure 1. Sampling Locations with reference points on top colored red (click to enlarge)

    For this project, we were tasked with assessing horizontal accuracy of various road datasets/shapefiles following NSSDA guidelines. To assess horizontal positional accuracy, I selected 20 well-distributed intersection points visible in both the ABQ Streets and StreetMapUSA datasets. Each point was digitized using provided orthophoto imagery as the reference. I extracted X and Y coordinates for both datasets using the Add XY Coordinates tool in ArcPro and calculated the horizontal error for each point. Using these errors, I computed the Root Mean Square Error (RMSE) using our given worksheet and applied the NSSDA formula to determine 95% confidence accuracy for each dataset

    Using the National Standard for Spatial Data Accuracy (NSSDA), horizontal positional accuracy was assessed for two street datasets using 20 well-distributed test points. The ABQ Streets dataset tested at ±13.48 meters horizontal accuracy at the 95% confidence level, with a calculated RMSE of 7.79 meters. The StreetMapUSA dataset tested at ±283.04 meters horizontal accuracy at the 95% confidence level, with a calculated RMSE of 163.53 meters. These results indicate that the ABQ Streets dataset provides significantly higher positional accuracy than the StreetMapUSA dataset for the sampled intersections.