A combination of multiresolution analysis (MRA) and a modified Schroeder integration method is tested as a means of identifying unknown oscillating signals in time series data, and estimating signal decay. Significant signals can be detected by applying a single-value criterion to each level of the MRA decomposition of the time series. Alternately, subject to certain constraints, significant signals can be detected during the modified Schroeder integration step by applying a reverse-integrated criterion vector to the reverse-integrated decibel transformation of each level of the MRA decomposition. Given that the modified Schroeder integration procedure failed to produce a better estimate of signal decay than other currently available procedures, the single-value criterion method of identifying significant signals is retained as the simpler and more flexible of the two methods of determining signal significance.