The flux data represent the response of a filtration system to interactive progression of multiple fouling mechanisms. Analyses using the classical fouling models of filtration (Hermia's model and its modified forms) are often used by linearization of the flux data (dJ/dt) to identify the effects different fouling mechanisms. However, digitally acquired flux data with very short time intervals (high resolution) exhibit oscillations caused by particle-particle and particle-membrane interactions. As a result, dJ/dt can have both positive and negative values and not suitable for linearization. In this study, high resolution flux data were analyzed and compared in terms of magnitude and frequency of flux oscillations over time using Stockwell transform (S-transform) which is a modified form of the Short Time Fourier Transform (STFT). Flux data sets from filtration runs using membranes with different molecular weight cutoff (MWCO) (5, 10 and 30 kDa) were analyzed to characterize the progression of flux loss under submerged and cross flow filtration conditions. The magnitude and frequency of flux oscillations, and flux loss patterns for each filtration run were unique and allowed identification of the critical times when flux oscillation characteristics changes as a result of changing fouling dynamics under different filtration modes. For the submerged filtration conditions, all membranes exhibited significant flux loss immediately after the start of the filtration run due to cake formation followed by cake compression and penetration of particles into the membrane which resulted in flux oscillations at long filtration times, especially with 10 and 30 kDa membranes. Under cross flow conditions, membranes with smaller MWCO (5 and 10 kDa) did not exhibit significant flux loss over time; however, there were significant flux oscillations in the high resolution flux data indicating temporary blockages which did not have long lasting effects. The spectrograms produced by S-transform of high resolution flux data provided unique fouling signatures of the filtration systems and allowed comparison of the fouling progression for each filtration system. (C) 2016 Elsevier B.V. All rights reserved.