Here is a comparison of ground based magnetometer and UAS based magnetometer surveys over the same site.
For the ground based surveys I utilized a QUSPIN, INC. total field magnetometer (QTFMv1) as both a base and a rover. I had an Emlid Reach M2 GNSS receiver on the rover and an RS2 GNSS receiver as the base. I post processed the GNSS rover data vs the base data. I corrected the magnetometer rover with the base, then did a bit a data cleaning and filtering. I then merged the magnetometer data with the GNSS data and performed lag corrections to fix the offset between the two.
For the unmanned aerial system (UAS) based surveys, I utilized a Geometrics Inc. MagArrow magnetometer towed below a DJI Matrice 300 RTK (M300) quadcopter UAS. The flight height was approximately 60 ft and the MagArrow was at a height of approximately 55 ft. I did similar processing to the ground based surveys, but the GNSS receiver on the MagArrow does not have real-time kinematic (RTK) or post-processed kinematic (PPK) capabilities so I had to use the positions available from the integrated GNSS receiver. I have merged the corrected GNSS positions from the UAS with sensor payloads in the past, but I have found that the offset of the magnetometer has from the center of the UAS platform is not constant so it usually isn’t worth the effort.