We don't do much with the RTK portion, but instead are using a PPK process. This involves doing some conversions with the PPK Timestamp text file, and we are seeing a repeatable one foot error (or blunder) in the elevation. Every other aspect of the process is well within accuracy expectations, the relative accuracy in all three dimensions is as expected, the absolute accuracy of the XY is within expectation, the absolute accuracy of the Z (elevation) misses our GCP by one foot, which is the same as the 30 cm noted in this discussion. We will be doing what was suggested with taking some photographs with the UAV on the GCP. One photograph with the UAV facing due north and sitting plumb, one with the UAV facing due north and leaning as far northerly as possible, and one with the UAV facing due north and leaning due east. We are trying to better understand the values represent in the timestamp file.