<p>The work is concerned with the study of the breaking surface wave effect on the intensity and spectral characteristics of a scattered radar signal in laboratory conditions.</p><p><span>The experiments were carried out on the reconstructed TSWiWaT wind wave flume of the IAP RAS. The channel is 12 m long, the channel cross-section varies from 0.7 x 0.7 m at the entrance to 0.7 x 0.9 m in the working section at a distance of 9 m. The airflow speed on the axis is 3-35 m/s, which corresponds to the values of the wind speed U</span><sub>10</sub><span> of 11-50 m/s.</span></p><p>The wave characteristics in the flume were measured by an array of three wave gauges positioned in the corners of an equal-side triangle with 2.5 cm side, the data sampling rate was 200 Hz. Such a system gives the opportunity to retrieve 3D frequency-wave number spectra of surface waves.</p><p>The airflow parameters were measured using the profiling method. The velocity profiles were measured in the working section using an S-shaped Pitot tube. Microwave measurements were carried out using an X-band coherent Doppler scatterometer with a wavelength of 3.2 cm with sequential reception of linear polarizations.&#160; The absolute value of the radar cross-section (RCS) on the wavy water surface was determined by comparing the scattered signal with the signal reflected from the calibrator with a known value of the RCS - a metal ball with a diameter of 6 cm. The dimensions of the observation cross-section were 40 cm x 40 cm, the incidence angles were 30&#176;, 40&#176;, 50&#176; for the upwind direction, the distance to the target was 3.15 m.</p><p>Two series of experiments were carried out. In the first case, wind waves on the surface of pure deep water, developing under the action of a fan generated wind, were studied. In the second case, a train of three waves was generated at the beginning of the channel, with the fan turned on, in order to simulate shallow water an inclined plate was placed under water in front of the measurement area. As a result, the breaking waves occurred at a fixed point and at weaker winds compared to the first case.</p><p>As a result, an increase in the scattered signal intensity during artificial wave breaking in the case of weak winds was noted. For strong winds, the effect turned out to be insignificant, despite the increased amplitude of the waves under study. The Doppler spectra analysis is also presented.</p><p>This work was supported by the RFBR projects No. 19-05-00249, 19-05-00366. &#160;</p>
Implementation of Machine Learning Algorithms in Spectral Analysis of Surface Waves (SASW) Inversion