Temperature and dewpoint data from NWS radiosondes and first-generation Water Vapor Sensing Systems (WVSS-1) installed on United Parcel Service aircraft, were compared for a three month period during the spring and summer of 2001. The comparisons were made at locations where WVSS data were available within one hour either side of the 00 UTC or 12 UTC radiosonde data collection times, at airports located within fifty miles of the radiosonde launch site. The WVSS-1 system uses a thin-film capacitor similar to those employed on NWS radiosondes. The data were transmitted to the ground via the Aircraft Communications, Addressing and Reporting System (ACARS), and were displayed on an internet web page hosted by the Forecast Systems Laboratory. The first of thirty WVSS-1 systems were functional in the spring of 1997, though the majority have been installed since early 2000. Installation or hardware problems have resulted in some aircraft having known wet or dry sensor biases. These have been excluded from the comparison.

Nearly two thousand data comparisons were made at various mandatory sounding levels from 925 to 250 mb. The two data sources compare reasonably well, especially below 700mb. Air temperatures at 925, 850, and 700mb differ by an average of 1.2 degrees C at the three levels, with dewpoints differing by about 3.5 degrees C. Larger differences occurred at high levels, which can probably be explained by dynamic heating of aircraft at high Mach numbers, the normally larger spatial separation of the two systems at higher altitudes, and the known difficulties of measuring atmospheric moisture at very cold temperatures.

A discussion is also made of the WVSS-2, a second-generation WVSS that uses a laser diode to measure moisture. The WVSS-2 is expected to be more accurate than radiosondes at all levels, and require little maintenance.