Pretest activities included wind sensor checkout and calibration in a wind tunnel at the Technical Center. The test bed was established on an existing equipment tower on a well-exposed mountain overlooking JNU. This site is one of three existing ridge-top weather stations associated with JHIS, and has a suitable tower, equipment, and communications infrastructure to support installation and continuous operations of the test equipment. The site is subject to extreme meteorological and climatic conditions where snow and the buildup of rime ice on exposed surfaces can be substantial. With assistance from NCAR, nine anemometers, all with heater capabilities, were installed along with other instrumentation including an ice detector, temperature/relative humidity probe, and Internet-capable video cameras. Data was acquired from the mountain via a high-speed wireless network in an unattended mode during the 6-month period. Continuous remote monitoring of sensor data and video was accomplished via a server with Web and FTP capabilities.
Approximately 3500 hours of test bed data were collected and analyzed over the 184-day period. Video motion clips were constructed and used to visually assess icing. In-depth data analysis was performed for the ~37 hours where winds exceeded the current limits for airport operations. With a few exceptions, the existing JWHIS mechanical anemometers exhibited minimal icing. Mechanical sensors from a second manufacturer experienced significant icing, and were found operationally unsuitable due to extended ice-related outages. Favorable performance was displayed by the ultrasonic anemometer. The sensor was determined through internal status as well as video data to be degraded by icing 0.8% of the 6-month period. Further analysis of the periods of degraded performance (with respect to the JAWS 1-minute wind reporting requirement), revealed that in 84% of affected cases, there were at least 75% raw data samples available to derive a valid wind report. During the entire field experiment, only 81 sporadic 1-minute wind reports were derived from 25% or less raw sensor data. The pressure-type sensor that was evaluated exhibited encouraging performance; however the amount of data collected was limited due to delayed sensor installation. This sensor is under further evaluation as part of a follow-on 2001–02 wintertime assessment currently being conducted in Juneau.