An evaluation of water vapor radiometer data for calibration of the wet path delay in very long baseline interferometry experiments

The internal consistency of the baseline length measurements derived from analysis of several independent very long baseline interferometry (VLBI) experiments is an estimate of the measurement precision. In this paper we investigate whether the inclusion of water vapor radiometer (WVR) data as an absolute calibration of the propagation delay due to water vapor improves the precision of VLBI baseline length measurements. We analyzed 28 International Radio Interferometric Surveying runs between June 1988 and January 1989; WVR measurements were made at the G. R. Agassiz Station of the Harvard College Observatory in Ft. Davis, Texas, during each session. We found that the addition of WVR data decreased the scatter of the length measurements of Ft. Davis baselines by 5–10%. The observed reduction in the scatter of the baseline lengths is less than what we would expect from the behavior of the formal errors, which suggest that the baseline length measurement precision should improve 10–20% if WVR data at Ft. Davis are included in the analysis. The discrepancy between the formal errors and the baseline length results could be explained as the consequence of systematic errors in the dry-mapping function parameters, instrumental biases in the WVR and Ft. Davis barometer, or both.