In this study a one dimensional stochastic forcing model is used to estimate latent and sensible heat flux feedback factors from sea surface temperature anomalies at three-hourly, quarter degree resolution. A new dataset (SeaFlux version 1) compiled from satellite observations of diurnally-varying latent and sensible heat fluxes, accurate to within 14 W m-2 and 6 W m-2, respectively, is utilized. Western boundary currents and regions of strong horizontal motion in the ocean are neglected, as well as tropical latitudes, due to the assumptions of the model. Previous studies have used 30-day running means to calculate latent and sensible heat flux feedback factors. The latent and sensible heat flux feedback factors are estimated on sub-monthly and monthly time scales by utilizing discrete wavelet transforms to filter time series in a manner that retains the time localization (i.e. lag information), in addition to providing frequency decomposition and ensuring energy conservation, of the input signal. The estimation of sub-monthly feedback factors in this study not only quantify the magnitude and direction of oceanic mixed layer heat flux responses to forcing on short time scales, but also illuminate dominant atmosphere-ocean processes on different time scales.