Warning: All L2 Waveform products are calibrated in amplitude at 1kHz only. There is no phase calibration applied at this stage.
Before using these waveforms to process wave parameters, please follow the calibration process described below or the results will be wrong.
Level-2 (L2) Waveform data is amplitude-calibrated at 1 kHz. But there are amplitude deviations at other frequencies, and there are phase shifts which are not reflected in the L2 data at all. The phase shift is a frequency-dependent shift in the phase of the observed wave relative to the input wave, tantamount to a time delay at that frequency.
The L2 data can be adjusted over frequency by applying dimensionless complex factors over frequency immediately after Fourier transforming the L2 data. The file, called L2_fsw_tables_full_res_adjustment.txt, consists of a table for the B sensors and a table for the E sensors. Each table has 5600 complex entries, extending from 2.13623 to 11962.89 Hz, in steps of 2.13623 Hz. Above the highest frequency the WFR filters roll off; no calibration measurements exist, but one could apply the last value to any frequencies above that.
The table is constructed assuming 16384 data points are to be FFT'ed. If fewer than that are to be transformed, then the table can be decimated to accommodate a shorter data set. The procedure is; FFT the L2 data at the desired resolution and then perform a complex multiplication of the FFT'ed dataset and the E or B adjustment table.
If comparing results of the frequency-adjusted L2 data with the onboard survey spectra, note that the L2 data has units of volts/meter and nanoTesla for E and B respectively, whereas the onboard survey spectra have units of RMS volts/meter and RMS nanoTesla.