I would like to understand what would cause a constant offset in the RTK GPS measurement. For example, we estimate the GPS location by averaging the GPS measurements at the same physical location for a constant sample size. Let’s assume the GPS measurements follow a Gaussian curve. What I have observed is on different time of the day, when we conduct such measurements. It is not the variance of the Gaussian curve is changing but the mean of the Gaussian curve is shifting (drift). For all the experiments, same base station is providing the RTK correction. And the base station is not being moved physically.
Is it something that can happen to the RTK GPS? What would be the cause? What is the best practice to minimize such phenomenon to happen.