Long-Term Spectral Average (LTSA)
       LTSA (Long-Term Spectral Average) plots allow large time series data sets to be viewed and analyzed including searching for and noting acoustically significant events, like whale calls. In 2006, each HARP deployment was capable of producing about 2000 XWAV files (~ 2 TB total), but viewing and analyzing each one of these XWAV files is not practical. LTSA plots provide a means of presenting an overview of these large data sets in a compressed format and allow quick linking to noteworthy events in the finer time scale XWAV data, which originally were used to generate the LTSA plots.
       An LTSA of time-series data is essentially a spectrogram (three dimensional time-frequency-energy plot) where each frequency spectrum plotted along time is averaged over a much longer period than one windowed frame of a Fast-Fourier Transform (FFT), as for typical spectrograms. The quantization time of an LTSA is defined as the duration over which consecutive single-window spectrum are averaged.  The averaged spectra are then plotted sequentially with energy shown as color.
Above is an example of a 24 hour LTSA from 35 GB of data sampled at 200 kHz
off the shore of Southern California. Notice how well ship echosounders and dolphin clicks
and whistles can be identified, and periods when the region is relatively quiet.

The LTSA plot above is from the first two hours of the previous figure, and the three
plots A, B, and C are spectrograms from the XWAVs used to generate the LTSA at the corresponding times A, B, and C in the LTSA plot, notice the differences.

Credit: Sean Wiggins