Model a vessel
102,629 admitted AIS records have source spectra and a Fast TL estimate. Juan’s 674 records also have all eleven one-to-one RAM frequencies.
Map, SL, TL and RL →Can AIS explain the measured shipping noise?
Some selected high-exposure windows are louder than their lower-exposure comparators. The evidence does not yet establish ship-specific attribution or out-of-sample prediction.
Positive contrasts—and a reversal
Three of four preselected H01W pairs have a positive measured energy contrast at both 63 and 80 Hz. One reverses at 80 Hz. None establishes repeated directional agreement; the close-cargo case is directionally disfavoured.
Inspect the four pairs →Traffic →Which ships can we model?Propagation →How does sound reach the hydrophones?Measured sound →What did the hydrophones record?Comparisons →Do the observations agree with the predictions?
Inspect the measured sound
30 of 113 merged Juan arrival windows contain some calibrated SPL. Partial support is not a complete six-hour window.
SPL through time →Compare prediction and measurement
The Highlighted 21 retain measured spectra, fixed 6 m predictions and admitted draught sensitivity. The measured curve includes the whole soundscape.
Highlighted comparisons →Inspect the wider study
Propagation sensitivity, spectral structure and hydrophone-array consistency remain distinct tests, with their own scopes and limitations.
Measured products → · Propagation →Scope and evidence cutoff
This is a presentation refresh over saved products. The broader report receipts are dated 2026-09-04 15:13:25 UTC; the Juan RAM and arrival products retain their own source receipts. This is not a live worker-status display.