Vertical difference between Calibrated target location and surveyed position

Hi, I have been exploring Pix4Dmatic as an alternative to Pix4Dmapper when it comes to getting corrections from a Real Time Network (RTN) on our RTK enabled drone. We connected the drone to the RTN and collected the images. Once the mission was completed we then connected our Trimble R10 rover to the same RTN to tie a handful of targets. Within Pix4Dmatic I specified the datum information and completed the initial calibration of the collected images. When I bring in the tied targets and mark them within the images as check points the horizontal error from the calibrated position to the surveyed position has showed great result being approximately 0.1’. My problem is when it comes to the vertical. Vertically I am seeing approximately 2.5’ from the computed position to the surveyed position. I can confirm it is not a rod bust. I am not sure where this vertical shift is coming from. My goal is to obtain the best results possible without constraining to the collected targets for the vertical difference. Hopefully someone has an explanation and solution?

Thank you.

Hi jlitts,
Can you confirm what coordinate system your RTN is correcting on? Is it NAD83 (2011)? In addition, can you confirm what vertical coordinate system your RTK drone is using. Is it the GRS80 ellipsoid, EGM96, etc. And then are you entering that information as the image coordinate system?