[en] According to the conclusion of the simulation experiments in paper I, the Tikhonov regularization method is applied to cyclone wind retrieval with a rain-effect-considering geophysical model function (called GMF+Rain). The GMF+Rain model which is based on the NASA scatterometer-2 (NSCAT2) GMF is presented to compensate for the effects of rain on cyclone wind retrieval. With the multiple solution scheme (MSS), the noise of wind retrieval is effectively suppressed, but the influence of the background increases. It will cause a large wind direction error in ambiguity removal when the background error is large. However, this can be mitigated by the new ambiguity removal method of Tikhonov regularization as proved in the simulation experiments. A case study on an extratropical cyclone of hurricane observed with SeaWinds at 25-km resolution shows that the retrieved wind speed for areas with rain is in better agreement with that derived from the best track analysis for the GMF+Rain model, but the wind direction obtained with the two-dimensional variational (2DVAR) ambiguity removal is incorrect. The new method of Tikhonov regularization effectively improves the performance of wind direction ambiguity removal through choosing appropriate regularization parameters and the retrieved wind speed is almost the same as that obtained from the 2DVAR. (electromagnetism, optics, acoustics, heat transfer, classical mechanics, and fluid dynamics)