We demonstrate the effectiveness of semi-empirical Brooks, Pope and Marcolini model and hybrid large eddy simulations in calculating the blunt trailing edge wind turbine noise at higher Reynolds number conditions. The 4 million element meshes of sharp and blunt trailing edge airfoils were tested at a Reynolds number of 3.2 million and an angle of attack of 4 degrees. The predicted airfoil self-noise by the modified semi-empirical formula with a low frequency directivity function and an additional term for large thickness ratio was compared to the experiments. The sound pressure level spectra from the hybrid large eddy simulation show that the predictions agree well with experimental measurements at the same observer location in the peak frequencies of the blunt trailing edge noise and sound pressure level rates of change at lower frequencies are also similar to experiments. The modified semi-empirical formula and the hybrid large eddy simulation can be considered as promising tools for high vorticity flow problems, especially for flatback airfoils for use on large wind turbines.