Many measurements of hurricane waves have been made from deep water production facilities in the Gulf of Mexico. Measurements made on different sides of the platforms differ from one another and the incident wave field because the platforms diffract and radiate waves. For many purposes, we would like to know the incident wave field. Forristall and Aubault (OMAE2013-10860) used WAMIT diffraction calculations to successfully invert wave spectra measured under a TLP model in the Marin offshore basin. We have now used similar techniques to invert spectra measured at offshore platforms during Hurricanes Gustav and Ike. We do not have any measurements of the undisturbed wave spectra for testing the results. The tests were made by checking whether inverse calculations on all the gauges deployed at different locations on the platforms could produce the same undisturbed wave field.
Wave directions are needed for the diffraction calculations. Information from the wave gauges can be used to find the directions by optimizing the agreement among the inverted power spectra. To perform the optimization, we varied both the mean direction and spreading at each spectral frequency. The rms difference between the inverted probe spectral densities was minimized at each spectral frequency.
When spectra from four gauges on a platform are inverted, they agree reasonably well with each other. The average of the inverted significant wave heights is slightly lower than the average of the measured significant wave heights. But when spectra from pairs of the four probes are inverted, the results differ depending on which pair is used. This result implies that our inversion method cannot be used on data from platforms with two probes, and casts doubt on the accuracy of four probe inversions.