Integrating Multi-frequency HF Doppler Radar Estimates of Wind Direction with Other Meteorological Observations to Obtain Surface Wind Patterns over Coastal Ocean Areas

The University of Michigan, Stanford University and ERIM International have operated two four-wavelength, high-frequency (4-25 MHz) phased-array radars in coastal regions (Monterey or Chesapeake Bay) for extended periods since June, 1997 to determine current profiles to a depth of about 1.5 m. The radial components of the "Bragg resonant" ocean gravity waves (wavelengths of about 7 to 31 m) are used to estimate the currents. This paper deals with another characteristic of these radar returns which is related to the winds in the lower atmospheric boundary layer (ABL). The magnitude of echoes from the Doppler peaks is proportional to the height variance of the resonant ocean waves; the relative strength of the two peaks (for waves moving toward and away from the radar) provides a measure of wind direction relative to the direction in which the radar points. We reported at the second AMS Conference on Coastal Atmospheric and Oceanic Prediction that a preliminary study demonstrated that it is possible to estimate wind directions representative of areas a few kilometers on a side within about 20 degrees.

In this paper, we will describe how these wind direction measurements can be combined with other data to obtain more complete wind fields over coastal ocean areas. The sparsity of measurements in these regions makes additional information particularly welcome. We will describe how the HF radar estimates of wind direction can be combined with Doppler weather radar measurements of radial wind components. Very briefly, it is possible to determine overall wind speed if we know the wind direction (from the current measuring radars) and the component of the wind in a direction that is not orthogonal to that wind direction. Obviously, assumptions are required concerning the wind profile below the level of the weather radar observation. Another method that will be discussed involves speed interpolation from conventional ground level anemometer observations, combined with iterative adjustment to achieve mass conservation.