Ionosphere effects at the nanosecond level observed in common-view time transfer

The single-channel "Common-View" GPS time-transfer technique is frequently used to minimize the inaccuracy caused by satellite clock errors, including Selective Availability. Even when carefully conducted, these measurements are subject to receiver noise, satellite ephemeris errors, multipath effects, and variable propagation delays. Many of these errors can be reduced if continuous measurements are made on all of the satellites that are in common view. We report results using this method with multichannel, modular, L1, time receivers over a 4,000 km Fast-West baseline. The effect of averaging on the sources of error is discussed. The experiments show measurement noise at the level of about 1 nanosecond for averaging times between 100 seconds and 2 days without post-processing. In addition, significant diurnal effects af the level of up to 10 nanoseconds peak-to-peak are often observed. These appear to be due to the expected differences between the actual ionospheric delays and the predictions of the built-in model. Techniques for mitigating these errors are discussed.