Validation of high-frequency radar ocean surface current observations in the NW of the Iberian Peninsula

• The accuracy of a CODAR HF radar network deployed along Galician Coast was assessed.
• Sites performance was monitored routinely in quarterly basis.
• Radial currents were compared with measurements from two moored current meters.
• Total currents were also compared, evaluating the impact of the averaging radius.
• Overall results proved that one of the sites should be revised and recalibrated.

An assessment of the accuracy of a four-site long-range (5 MHz) CODAR SeaSonde HF radar network deployed along the Galician Coast is attempted through a comparison with measurements from two moored current meters for the entire year 2012. Radial vectors comparison shows correlation coefficients and RMSE in the ranges 0.32–0.72 and 10–17 cm s−1, respectively. Observed discrepancies in current velocities could be partially attributable to the existence of bearing errors and the influence of the current profile, although other factors like the Stokes drift or intrinsic HF radar variance might also contribute. Evidence of bearing errors can be seen, with values lying between −15° and +10°. The study is also conducted on a quarterly basis in order to routinely monitor performance of the sites in terms of bearing offset evolution, with offset values remaining stable throughout 2012. Use of measured antenna pattern only improves performance at two radar sites. Comparison of total current vectors is carried out together with an evaluation of the impact of the selected averaging radius (AR). Lower AR values give rise to optimized metrics when radar sites involved in the comparison operate accurately, reporting correlations and RMSE in the ranges 0.56–0.74 and 8–13 cm s−1, respectively. By contrast, if at least one of the two involved sites operates erratically, the strategy based on reducing AR becomes counterproductive and the imposed standard value of 25 km seems to be appropriate.