Premium
Lorentzian Filter Correction of Turbulence Measurements on Oscillating Floating Platforms: Impact on Wind Spectra and Eddy‐Covariance Fluxes
Author(s) -
Ezraty R.,
Mor Z.,
Bodzin R.,
Assouline S.,
Tanny J.,
Fratini G.,
Griessbaum F.,
Lensky N. G.
Publication year - 2021
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/2020wr027583
Subject(s) - anemometer , wind speed , eddy covariance , turbulence , filter (signal processing) , environmental science , mechanics , meteorology , physics , acoustics , engineering , electrical engineering , ecology , ecosystem , biology
Turbulence and eddy‐covariance measurements on a floating platform over water surfaces can be contaminated by platform oscillations, which may affect the calculated air‐water exchange. The conventional method for decontamination of the platform oscillations from the wind velocity measurements requires the installation of an additional sensitive, and often costly, motion sensor. This paper examines a new mathematical decontamination method, termed Lorentzian filter, which avoids the need for such an instrument. The method, based on the Lorentzian function, capitalizes on the pseudo‐harmonic behavior of the platform oscillations and reduces the amplitude of turbulent wind velocity data detected as artifacts at the specific natural frequencies of the platform. The Lorentzian filter was applied to wind velocity data measured by sonic anemometer and eddy‐covariance system over the Dead Sea, Israel, for 30 days. We examined two approaches of dealing with motion contamination: Lorentzian filter decontamination and motion sensor decontamination. Both approaches were compared to nonfiltered raw wind velocity as well. Using the 3D wind velocity series, we examined the wind spectra, the cospectra of water vapor concentration and horizontal wind speed with vertical wind speed, and H 2 O and momentum fluxes. The Lorentzian filter performed very well in decontaminating the wind spectrum, meaning that it efficiently identified the contamination in the natural oscillation frequency and returned a decontaminated wind velocity time series. The cospectra and fluxes were less prone to the contamination of platform oscillations, presumably due to low correlations between the spurious wind velocity components and other measured scalars, such as water vapor.