Abstract
Attenuation problems arise when short wavelength radar is used for severe storm identification and structure analysis. These are illustrated by comparing observations from 5- and 10-cm Doppler radars. Reduced and fragmented storm representations were obtained with the 5-cm wavelength radar. Signal losses exceeding 30 dB so greatly distorted the reflectivity structure of one thunderstorm that the expected association between mesocyclone and reflectivity pattern was not evident. Attenuation of the received signal reduces the signal-to-noise ratio and increases the variance of the spectral moment estimates. However, velocity measurements remain unbiased as long as the received signal remains above the system noise level. Correcting for attenuation appears futile.
