Introduction
This lab focused on why datums and coordinate formats matter when working with UAS and survey data. Even if a ground control point (GCP) file looks “normal,” mistakes like swapped coordinate order or using the wrong vertical reference can place points in the wrong location or make elevations incorrect. In this activity, I worked with a real GCP dataset that had multiple problems on purpose, and my job was to find the issues, fix them, and understand what caused the errors.
Overview
I imported the GCP file into ArcGIS Pro as XY data. The points showed up, but they were clearly not in the right place, which suggested that the coordinate fields were being interpreted incorrectly. After re-importing the data with the correct field setup, the points appeared in a more reasonable location.
Next, I checked whether the height (Z) values matched the terrain. By comparing the point heights with approximate ground elevations in ArcGIS Earth, I noticed a consistent difference across multiple points. A repeated offset like this usually indicates a vertical reference issue rather than random error. To better understand the cause, I used an authoritative reference source to confirm how ellipsoid-based heights and mean-sea-level-style heights can differ, then adjusted the dataset accordingly and documented the correct coordinate and vertical reference information in the final map layout.
Conclusion
This lab showed that small-looking mistakes in a GCP file can cause huge problems in UAS mapping. If the coordinate order is wrong, points can plot on the wrong side of the world. If the vertical reference is wrong, elevations can be consistently offset by tens of meters, which can damage products like orthomosaics, DSMs, DTMs, and any accuracy-focused deliverable. The key takeaway is that UAS data should never be trusted “just because it loads.” Coordinates and datums must be checked early, and external references like NGS can help confirm what the correct vertical relationship should be. After fixing both the coordinate order and the height reference, the dataset became usable and could be mapped with confidence.