Estimation and prediction of the real time information of the oceanographic parameters is of vital importance in India as more than 25% of the population resides along the coastlines. Information of the significant wave heights is necessary to deal with many oceanographic activities as almost all ocean engineering applications depends on it. Presently Indian National Centre for Ocean Information Services (INCOIS) provides wave forecasts on regional and local level ranging from 3hours to 7 days ahead using numerical models (www.incois.res.in). It is evident from real time observations that the predicted SWHs by a physics based model vary randomly and have non-linear relationship with observed values due to many reasons. Consequently predicted and actual values deviate significantly from each other with an ‘error’ which has to be removed to cater the needs of safe and secure lives residing along Indian coastline. Present work aims in reducing this error in numerical wave forecast made by INCOIS at Ratnagiri station on the south-west coast of India. For this ‘error’ between forecasted and observed waves at current and previous time steps were taken as input to predict the error at 24 to 48 hr ahead lead time in advance using a Geno-Wavelet Technique. Geno-Wavelet Technique is a combination of Genetic Programming (GP) and Discrete Wavelet Transform (DWT). This predicted error was then added or subtracted from numerical wave forecast to improve the prediction accuracy. It is observed that the numerical model forecast improved considerably when the predicted error was added or subtracted from it. It will add to the usefulness of the wave forecasts given by INCOIS to its stake holders.