Application of Wavelet De-noising for Wind Speed Retrieval from a 2.5 MHz Partial Reflection

Abstract

Atmospheric studies related to middle-atmospheric coupling, climate monitoring, steering profiles for expendable launch vehicles and re-entry missions require atmospheric wind profiles as an important input. Atmospheric wind profiling is conventionally carried out using rawinsondes and radiosondes flown in high altitude balloons, complemented by sounding rockets and ground-based radars. The partial reflection radar (PRR) can be effectively used for the same purpose to an altitude of about 100 km. But deriving wind profiles from radar samples is a difficult task due to contamination of radar signals by electromagnetic noise from various sources. In the present work, an attempt has been made to de-noise the signals from 2.5 MHz PRR located at Trivandrum (8.5º N, 76.9º E) using a novel wavelet de-noising technique. Comparison of signal to noise ratio (SNR) showed improvement of the order of ~ 2 dB on the average after applying the wavelet de-noising technique. This is a significant improvement given the high level of noise contamination in the signals received using the PRR. The wavelet de-noising technique was proved to be a useful tool for PRR signal de-noising compared to conventionally used threshold methods and can be operationalized for other such atmospheric Radars. 

Keywords: Wavelet de-noising, partial reflection radar, wind profiling