Noise properties of HF radar measurement of ocean surface currents

Abstract

AbstractHigh‐frequency (HF) radars are commonly used for coastal circulation monitoring. The objective of the study is to assess what is the minimum timescale of variability of the geophysical surface currents that are accessible to the radar measurement given the intrinsic noise of this measurement. Noise properties are derived from the power density spectra (PDSs) of radial current records, which are compared to a model of the PDS of idealized currents contaminated by an additive white noise. The data were collected by two radar systems operating in the Northwestern Mediterranean. Periods of 3 weeks to 7 months are considered. Most of measured currents are affected by a white noise effect. Noise properties vary in time and space and are not specific to a particular radar station or to the radar signal processing method used (beam forming or direction finding). An increase of the noise level reduces the effective temporal resolution of radar‐derived currents and then increases the minimum observable timescale of variability of geophysical currents. Our results are consistent with results of comparison found in literature between in situ sensors and radar measurements as well as between two radars operating along a same base line. The study suggests a self‐sufficient method, requiring no external data, to estimate the minimum sampling period to consider for getting data sets having a minimized contamination by instrumental noise. This period can also be taken for smoothing or filtering measured currents.