GIS5935 Mod2: Data Quality Standards

 


In our recent GIS lab, we assessed the horizontal positional accuracy of two street datasets — the ABQ_Streets_Sample and the USA Streets dataset — using the National Standard for Spatial Data Accuracy (NSSDA). This standard provides a consistent framework for reporting the accuracy of spatial data by comparing dataset points to high-quality reference points.

For each dataset, we selected 20 well-defined check points and calculated the distance between the dataset points and their corresponding reference points. From these distances, we computed the Root Mean Square Error (RMSE), which summarizes the overall error, and then derived the 95% confidence accuracy, which indicates the expected maximum positional error 95% of the time.

Our results show that the ABQ_Streets_Sample dataset has an RMSE of 12.38 meters, with a 95% confidence accuracy of 21.44 meters, reflecting a high level of positional precision. The USA Streets dataset, by comparison, has an RMSE of 90.19 meters and a 95% confidence accuracy of 156.11 meters, indicating lower positional accuracy. These metrics are essential for understanding the reliability of spatial data and ensuring that analyses based on these datasets are appropriately interpreted.

Comments

Post a Comment

Popular posts from this blog

Module 6: Isarithmic Mapping

Image
  In module 6 we focused on Isarithmic mapping. Isarithmic mapping uses lines (isolines) to link points of equal value for continuous data like temperature or rainfall across a map. The pattern and spacing of these lines visually represent the distribution and rate of change of the phenomenon.  The lab for module 6 was quite interesting. We used data from PRISM to complete this lab. PRISM is a group out of Oregon State University that generates detailed maps of monthly and annual climate conditions like precipitation and temperature. It works by combining specific data points with a base grid, such as elevation data or a 30-year climate average. This approach is very effective in mountainous areas because it accounts for how terrain influences weather patterns. Monthly maps are created first, and then annual temperature maps are derived by averaging these monthly values, while annual precipitation maps are calculated by adding the monthly totals. As you can see when viewi...
Read more

GIS5935: Mod5 Surface Interpolation

Image
 In this lab, we explored different interpolation methods to create continuous surfaces from point data using ArcGIS Pro. Part A focused on generating a Digital Elevation Model (DEM) from elevation points using IDW and Spline interpolation. Comparing the two surfaces showed that Spline produces smoother results but can exaggerate extreme values, while IDW provides a more realistic representation of terrain. Raster calculations highlighted the areas where the two methods differ. Part B applied interpolation to water quality data in Tampa Bay. Thiessen polygons, IDW, and Spline methods were used to create continuous surfaces. Results showed that Spline can overestimate values in areas with sparse sampling, whereas IDW and Thiessen provide more conservative estimates. This lab emphasized the importance of choosing an appropriate interpolation method based on data distribution, density, and analysis goals.
Read more

Module 4: Data Classification

Image
This lab session was about understanding data classification. The lecture explained different types of data and how they're shown on maps. It then covered various methods for dividing data into categories for visual representation. In the Lab portion, we used census data for Miami-Dade County, Florida, specifically looking at the percentage of the population over 65. We created four different maps of this same information, with each map using a different data classification method. Those classifications were Equal Interval, Quantile, Standard Deviation, and Natural Breaks. We then arranged these four maps on a single layout, including standard map elements. The goal was to visually compare how each classification method displayed the data and what patterns became apparent or less obvious with each approach. Finally, we considered which classification method might be most useful for different purposes, such as identifying areas with a high senior population, and whether using percen...
Read more