Wind-noise reduction for infrasonic measurements using adaptive line enhancement

Abstract

Measurement quality and analysis capability of infrasonic signals are both affected by background wind-noise. Physical filters, i.e., barriers and pipe arrays, are traditionally employed to reduce such noise. However, limited efficacy, site dependence, cost, requirement of space, non-portability, and frequent maintenance are some of their major drawbacks. This work proposes an adaptive filtering-based adaptive line enhancer (ALE) noise cancellation scheme as an alternative. Two infrasonic sensors (Chaparral Physics 50A), are adjacently deployed. One sensor is fitted with a conventional four-armed non-porous hose array (physical filter), while the ALE scheme is applied to the second sensor, sans physical filter. In high wind-noise conditions, the ALE scheme seems to behave as a lowpass filter (cutoff at 0.2 Hz), with a maximum attenuation of 26 dB at 8 Hz, while the physical filter provides significant attenuation only above 4 Hz with a maximum attenuation of 17 dB at 8 Hz. Generally, at other frequencies, the ALE scheme provides up to 20 dB superior noise attenuation as compared to the physical filter. The ALE also provides up to 6 dB gain in the signal-to-noise ratio as compared to the physical filter, due to non-attenuation of the infrasonic signal.