Estimation of sea surface winds using reflected GPS signals

This dissertation presents algorithms and results for GPS-based remote sensing of ocean surface wind speed and direction. Results are given from aircraft experiments and simulations are shown for a spaceborne platform. The algorithm matches predictions of reflected signal power distribution in delay and Doppler with waveforms measured by a specialized receiver.; The wind retrieval algorithm for airborne receivers comprises two parts: a least-squares estimator of the surface slope variances and a solution for wind speed and direction. The state vector consists of the slope variances and covariance in the East, North, Up reference frame and a fixed path delay offset. The state estimate is that state which minimizes the sum-of-squares residual between the measured and predicted waveforms.; Model waveforms for aircraft altitudes are generated using an electromagnetic model developed by V. Zavorotny and A. Voronovich. Airborne experiments were conducted using a Delay Mapping Receiver (DMR) developed at the NASA Langley Research Center (LaRC). In situ measurements used as sources of truth include buoys and QuikSCAT scatterometer measurements. GPS-based estimates agree with in situ measurements to within 1 m/s and 5 degrees for data sets containing well-defined wind fields, long fetch, and high GPS satellite elevation angles.; The results of the visibility analysis are used in a sensitivity and wind retrieval resolution analysis. The sensitivity analysis identifies the regions of the glistening zone that have the greatest change in power for changes in wind speed and direction. The analysis shows that the waveform peaks and trailing edges should be used to estimate wind speed and regions of the waveform trailing edge should be used to estimate wind direction. The resolution analysis determines the accuracy with which wind vectors can be estimated over the glistening zone for varying wind conditions. Results show that wind can be retrieved to within 1 m/s and 20 deg over large portions of the glistening zone with as few as 3000 0.3 ms observations at low wind speeds. The resolution degrades for higher wind speeds, requiring far more observations to achieve comparable retrieval accuracy. (Abstract shortened by UMI.)...