The influence of spectrum functions and filters on crossing frequency of atmospheric turbulence.

Abstract

The effect on the crossing frequency was investigated for different spectrum functions and filters by systematically mixing the Von Karman and Kaimal spectrum functions with four filters: ideal band pass; ideal low pass; quadratic type and the sine type. Except for the quadratic type, when the difference is about 17% at high frequency and declines to zero at low frequency, the influence of spectrum functions is very small. The filters have a much greater effect. The predicted crossing frequency varies by as much as 50% among the low pass filters, and ideal one decade wide band pass filtering predicts crossing rates up to eight times higher than ideal low pass filtering for the frequency range studied. For a one decade wide ideal band pass filter and medium to high frequencies, the predicted crossing rate is found to be independent of the site elevation, roughness and wind speed and is given by 0.352n SUB 2 (per second), where n SUB 2 is the upper cut off frequency in Hz. (from authors' abstract)