The environment changes. The pulse usually doesn't.
A sonar pulse is chosen for a set of water conditions. Those conditions move during a survey, not between them — so a pulse chosen on the deck is already the wrong pulse by the time it reaches the seabed.
What actually moves
Four properties set the acoustic channel, and all four drift continuously: temperature and salinity move the speed of sound, depth moves the pressure term, and suspended sediment changes absorption in a way that depends on frequency. The payload samples all of them.
What the model does about it
Sweeping turbidity with everything else held, at a required range of 220 m in the underwater configuration. The optimiser re-solves each tick; nothing below is a lookup.
| Turbidity NTU | Centre kHz | Bandwidth kHz | Pulse ms | Resolution mm | Usable range m |
|---|---|---|---|---|---|
| 5 | 351 | 299 | 0.5 | 2.6 | 365 |
| 180 | 351 | 299 | 0.5 | 2.6 | 275 |
| 550 | 275 | 247 | 7.4 | 3.1 | 252 |
| 1000 | 148 | 96 | 2.8 | 8 | 294 |
Read it downward. High frequency buys resolution while the water is clear; as sediment loads, the optimiser drops the centre frequency to keep the link budget closed, and lengthens the pulse to hold energy on target. Resolution degrades from 2.6 mm to 8.0 mm because that is the price of still seeing anything at 1000 NTU.
A fixed-frequency instrument has one row of this table. It is excellent in one column of water and blind in the next. The adaptive one has all four rows and picks between them without being told which water it is in.