WERA HF radar measurements of wind direction in the South Atlantic Bight

Wind direction measurements collected at 4 in-situ anemometers were compared with the WERA HF radar over a 20 month period to assess the accuracy of the radar system. The radar data was interpolated onto the less frequent in-situ time series. An unexpectedly large RMS error of 58.9° was calculated over all stations. A complex correlation coefficient of 0.65 was obtained. An improved correlation was exhibited at all stations with increasing wind speed, which is explained by an increase in radar accuracy. The radar operates at 8.35 mHz, and using this frequency the phase speed of the shortest wave sensed by the radar was calculated to be 5.3 ms -1 ; therefore in wind speeds under this threshold the radar estimates were discarded. Shallow water conditions were shown to affect all 4 locations when the dominant wave period exceeded a threshold limit for each station. Correlation improved when using only measurements made during deep water conditions. The results suggest that limited fetch affected all stations when the winds were blowing over the land. This is most evident at the Grays Reef station closest to the coast, where correlation actually declined in higher wind speeds that require longer fetch. Correlation of anemometers at different stations, and the radar between its grid cells, found that the in-situ measurements were more closely related to one another, suggesting the anemometers are in general more reliable. It was concluded that the WERA radar wind direction measurements are more accurate under a fully arisen sea state, specifically in high wind speeds for wind blowing over the open sea when deep water conditions apply. This was demonstrated when the radar data was corrected for these parameters; the grid cells at the 4 stations exhibited no significant difference in correlation with the insitu measurements.