While it is true that the characterization of the peak wind speed is the largest change from the conventional (CONV) F420/gust recorder to ASOS, there is also an interesting calibration difference. Comparative data from 16 stations show a systematic difference in wind speed as a function of wind speed, the signature for a calibration difference.
Wind tunnel calibrations of cup anemometers usually are generalized with a straight line. The constant from this linear regression defines the wind speed output with zero rotation of the cup wheel. This fact can be implemented in two ways. The F420 simply set the threshold of the output range at two knots. There was no speed indicated below two knots. The digital ASOS uses an algorithm to convert average frequency to wind speed.
Both the CONV F420 cups and the ASOS cups were calibrated in the same wind tunnel at Sterling, Virginia. This paper will show the results of an analysis of 35 wind tunnel calibrations of ASOS or Belfort model 2000 cups. In general, 28 cups fell well within the specification of the larger of ±2 knots or ±5% over the range of seven to 117 knots. The test measured the difference between the wind tunnel speed and the ASOS or Belfort output with a 0.1 knot resolution. There was no use of the algorithm in this test since frequency was not measured. The preliminary conclusion is that the ASOS data are confirmed by wind tunnel tests. Any difference between ASOS and CONV suggests the error rests with CONV