AT 309 Lab 9: GCPs & Datums
Introduction
The remote sensing industry requires the use of geocoding and georeferencing. Geocoding is the process of giving an address or name of a place a coordinate referencing the location of it on the Earth's surface. The universal format is latitude/longitude, which is the exact format that ArcGIS Pro follows. It then uses the XYZ coordinate system to reference the data; "X" is longitude and "Y" is latitude. Georeferencing is also important; it is the use of a map utilizing an internal coordinate system that relates to the geographical coordinates of the real world. This helps to find the actual location of different subjects when looking at them on a GIS application.
For Lab 9, students were tasked with looking into geographic datums, identifying different types of datums, using the XYZ coordinate system, and solutions to issues with location data given for a mission.
Discussion
The dataset that was given shows ground control points (GCPs) near Bloom City, WI. Figure 1 shows the points in ArcGIS Pro.
From observation, the data is skewed to the edges of the mission with not many in the center, which can create skewness in the location data and can make it inaccurate. For the lab, metadata was not given for the data. This was a big issue because this prevented us from knowing what the unit of measurement was and left for the outside users to guess important aspects. We were given a text file to use for the XYZ internal coordinate system. Originally, we were given a file named "GCP_groundvalue_xyz.txt." Figure 2 shows what the file contained.
Figure 2: This is the original text file that was given for coordinates.
Now, this text file that was given was told that it was incorrect. Originally, I thought the reason would be for the misaligned titles that could potentially mess up the data. When an XYZ coordinate system was generated, the points all showed up inside Antarctica, which if we remember correctly, they should be in Bloom City, WI. Figure 3 shows the correct format for the text file.
Figure 3: This is the correct layout for the data.
As we can see, it was not the headers that were the issue. The creator of the document switched around the "Y" and "X" columns. The longitudes and latitudes were backward, which in turn completely moved the data outside of where it should be. With that out of the way, it was time to look at the "Z" values to see if they correspond with the terrain. Each point was selected individually to see if they had the same values from the text file.
ArcGIS Earth was then used to find the elevation values of the area to test the data. The same text file was added as a table for evaluation. Figure 4 shows the data being added.
Figure 4: This is the table being generated after it was imported.
When looking at the approximate elevation values, they showed to off around 30 meters. This is a significant amount of margin error that could potentially wreak havoc on data. This could be a simple calibration error or even a starting datum error. To fix the data and correct the offset, the National Geodetic Survey Data Explorer was used. Figure 5 shows Bloom City, WI on their explorer.
Figure 5: This is the area of Bloom City, WI on the National Geodetic Survey Data Explorer.
After finding the exact GPS location, a datasheet was pulled up to better understand the data. From this datasheet, we can find that the orthometric height was found using the geoid model GEOID12A. Using the GEOID12A model, we find that the geoid height is -33.852 meters. Since we have found the geoid height, we can subtract that absolute value from the original points. By doing this, it creates an offset of about 2 meters, which is a lot more tolerable than the error before.
Conclusion
From this lab, we can conclude that the use of geocoding and georeferencing is important for datasets. This allows for the use of coordinate systems to be used for data collected from UAS, which then can be used to geographically reference in the real world. This also brings attention to detail where the GCPs are located to reduce skewness in the data. This also helped to correct any data that can be inaccurate to enhance the accuracy of it and increase the usefulness of it.
Comments
Post a Comment