Wavelet analysis of velocity-time data obtained from a bubble column using laser Doppler velocimeter (LDV) was carried out. The principal time scale/s, which contributes more to the coherent pail of turbulence than the other time scales, was identified based on the analysis of scalewise wavelet coefficients and its variance. Three methods namely, derivative, global and local (Donoho's method) were applied and compared with each other. The results of the derivative method were validated by studying flatness factor, a quantitative measure of local intermittency in the wavelet coefficients at various scales. Further, a new methodology has been developed for the estimation of bubble chord-length distribution in the column using the information in the principal time scale.