Estimating error

I have reports based on long GPS occupations of the same site, which returns X Y and Z coordinates as well as the SD for each.

I want the most accurate coordinates for the point as well as an estimate of the error/uncertainty.

Attached are the XYZ and SD for 8 different reports of the same point.

I had been averaging all of the coordinates and ignoring the SD, but that can't be right.

So 8 different populations with a mean and SD for each and I want to combine all 8 and get 1 mean and SD for the whole group. Thanks,