The coherent arraying of antenna elements by widely distributed ground-based antenna systems has proven to be a valuable technological approach for high precision astrometric measurements and imaging via Very Long Baseline Interferometry (VLBI) and has been performed with considerable success by radio astronomers for several decades. The fundamental factor limiting the precision in which these measurements can be conducted, however, is due to the turbulence-induced refractivity changes of the atmospheric medium (troposphere) through which the propagating waves must traverse. For radio science applications, this problem can be significantly reduced via three well-demonstrated means: (1) proper choice of ground site location (i.e., dry, high altitude climates), (2) conducting observations during non-turbulent times (i.e., nights vs. days, winter vs. summer), and (3) employing relatively long integration time (on the order of minutes) compensation through the use of water vapor radiometers in data post-processing. For communications applications, however, this may not necessarily be the case, and a means to accurately estimate the water vapor variability of the troposphere at short time scales will be required to efficiently combine signals from ground-based antenna elements in an array environment, particularly for transmit arraying.
It is thus the goal of this research effort to identify and validate a means in which phase fluctuations induced by the atmosphere can be accurately measured which could be employed to ultimately improve the coherent combining of several spatially separated signals transmitted from ground to space without the use of an active source (i.e., receive
signal). The method in which this will be accomplished is through the use of a passive radiometric technique capable of accurately determining phase fluctuations on the necessary time scales to provide real-time phase compensation to realize transmit arraying at Ka-band frequencies and higher. To improve the accuracy over the state of the art in radiometric water vapor retrieval techniques, a novel blind source separation technique has been developed and demonstrated. Utilizing experimental data using a water vapor radiometer and a two-element interferometer, it is statistically shown that the approach described herein improves water vapor retrieval accuracy, particularly during cloudy conditions, over the state of the art.