Data quality and sampling requirements for reliable wave measurement with HF radar

HF radar systems located on the coast can provide measurements of surface currents and waves. The maximum range of these measurements depends on the signal to noise in the radar backscatter spectrum. Previous work, comparing radar with buoy wave measurements, has shown that a second order peak signal to noise of 15dB can provide reliable measurements of the directional wave spectrum. However recent evidence suggests that this is not sufficient and some ideas for a more robust measure are discussed. For radar and buoy measurements averaging is required in order to reduce the variance in wave parameter estimates. The work of Sova (1995) on the impact of temporal sampling variability in radar Doppler spectrum estimation on wave measurement is reviewed. The implications of this for various sampling strategies that have been used for radar wave measurements is discussed and compared with the corresponding sampling variability of wave buoy measurements. The impact of increased averaging will be demonstrated using WERA data from the Eurorose Fedje experiment.