These links speak to the question of ENSO sensor data subject to many sources of noise-
https://www.pmel.noaa.gov/gtmba/data-quality-control
https://www.climate.gov/news-features/blogs/enso/enso-signal-and-noise
In the context here, we can identify Lunisolar Tide and ENSO State "cross-talk" as a possibility, of spurious signal overlaid on desired signal.
https://en.wikipedia.org/wiki/Crosstalk
Wanqui Wang seems to be the NOAA guy for the sort of ENSO Model Ensemble we have been looking for. His Ensemble Member Set compares about 30 ENSO Member Models to average out into his ENSO Ensemble Mean Prediction.
https://www.cpc.ncep.noaa.gov/information/who_we_are/cpc_staff/Wanqiu_Wang.shtml