Abstract
Wind tunnel tests have provided calibrations and intercomparisons of 14 Johnson-Williams (J–W) cloud liquid water content (LWC) measuring devices with 23 sensor heads from 10 research organizations. The absolute tunnel LWC was deduced using a rotating icing cylinder technique accurate to ∼5%.
A significant fraction of the systems arrived at the tunnel with nonfunctional shell or strut heaters, which can degrade measurements below 0°C. Several sensor heads exhibited airspeed dependencies. Switching heads sometimes produced calibration changes. At −15°C an instrument problem was discovered associated with icing of the compensating wire posts, which resulted in mild to severe measurement errors in 75% of the sensor heads at 103 m s−1.
Calibrations at −5°C revealed that J-W measurements usually varied linearly with tunnel LWC, but sometimes with a slope differing from unity, implying that the system dummy head did not always define the correct conversion from J-W output voltage to grams per cubic meter. No more than six of the 13 systems tested at −5°C agreed to within 20% of the tunnel LWC with each of their sensor heads, but at least 10 of 13 did so with one sensor head. At −15°C similar results were obtained, but most systems suffered from the aforementioned icing problem, resulting in unreliable small-scale measurements.
