Case Study: Closing a Half-Mapped Siphon After Three Seasons

The Siphon That Stayed Open for Two Years

Season one ended with the lead diver tying off at a constriction 412 meters into Sump 4, gas reserves spent, silt-out building behind him. Season two stalled in the same restriction when a flood pulse redeposited sediment and erased the previous line's tie-offs. The expedition journal — three years, eleven push dives, two broken stages — reads like the Ressel logbook, where the Sump 1 through Sump 5 traverse consumed staged exploration from 1981 until 1999 before anyone signed the end-of-sump survey. Curt Bowen's Beyond the Sump writeups document the same pattern across multiple French and Mexican siphons: multi-season tactics are the norm, not the exception.

Siphons (water-filled passages separating dry cave sections, per the standard sump taxonomy) resist single-season closure because the mapping budget per dive is tiny. A CCR push at 80m might yield 90 minutes of bottom time and 200m of new line — then six hours of deco. By the time conditions repeat (same viz, same flow, same tidal phase), a year has passed. Push-dive subsets drive a measurable share of American cave-diving fatalities, and the temptation to cut corners to close a multi-year lead is the common thread. Siphon closures need a method that treats each season as a patch, not a complete map.

The Lot basin shows the pattern clearly. Sumps in the Padirac, Ressel, and Émergence du Ressel systems take French sump-push teams multiple seasons to close because the access logistics — porter relays of CCRs across a kilometer of dry cave to reach the sump pool — limit each visit to a small window. A team that loses momentum after two seasons of unclosed lead often abandons the project entirely; the institutional memory walks away with the lead diver. Hölloch in Switzerland, with its kilometer-long flooded sections, has the same dynamic at multi-decadal scale. Wakulla and the WKPP push program in Florida fight a different version of the same problem: the bottom phase is so deep and the deco obligation so long that any single dive can only extend the survey by a few hundred meters of new passage.

Across all three regions, the binding constraint is not skill or gas but continuity — the ability to carry partial knowledge forward between visits.

Stitching Three Seasons Into One Quilt

EchoQuilt was built for exactly this failure mode. Each dive's ambient sound, DPV telemetry, and compass heading become a patch in a larger quilt — georeferenced to the dive's anchor points, timestamped, and queryable across seasons. The Lot-basin team's workflow on their three-year closure: season one's silt-out zone became a patch tagged "unclosed_west_432m", season two's flood-redeposit section became "sediment_rebuild_Apr24", and season three's push extended the stitching into clean passage all the way to an air bell at 618m.

The quilt representation matters because siphon push-diving is cumulative. Every ambient-sound sample from season one — the flow whine at 7.4 Hz behind the constriction, the regulator-exhaust echo off the far wall — remains available when season three's CCR diver replays it on approach. The dive leader doesn't re-enter "unknown" water; he re-enters water that's 40% stitched, with the missing 60% framed by patches that narrow the search cone. Voronya Cave's Samokhin expeditions pushed the terminal sump from 17m to 50.5m across multiple seasons — the same incremental patchwork approach EchoQuilt formalizes.

Three concrete patterns the Lot team relied on:

1. Silt-out reconstruction. When season two's flood pulse erased visual tie-offs, EchoQuilt's archived ambient signatures let the diver relocate the season-one line by matching flow tone, not sight. The quilt reconstructed passage geometry even when the water was unreadable.

2. Gas-budget forecasting from prior passes. Each prior patch came with swim time, CCR work-of-breathing, and scrubber temperature logs. The season-three push used those curves to predict bottom-time remaining at the 550m mark within 4% — enough margin to push past the 560m choke and still turn the dive safely.

3. Cross-team patch handoff. When the original lead diver couldn't make the spring window, a teammate dove the season-three push using the predecessor's stitched quilt. The handoff leaned on the same logic that underpins scaling phreatic work across multi-kilometer systems: patches are portable, individual memory is not.

EchoQuilt kept every pass in a single living map. The final closure dive wasn't a push into unknown water — it was a confirmation run that turned three seasons of partial quilt into one publishable survey.

Advanced Tactics for Multi-Season Siphon Closures

Teams running push programs on French siphons, Ressel-style multi-sump traverses, or Yucatán cenote extensions benefit most from three tactical layers on top of the basic quilt:

Season-boundary drift correction. Sumps silt up, flood pulses rearrange cobble, and bedding-plane passages migrate centimeters between seasons. EchoQuilt's cross-season delta view highlights where ambient-flow patches no longer match between years — a drift of more than 1.2m at a given control point tells the dive leader to treat that section as new water on approach. Borrowing from the Hadley Rille campaigns playbook, the team checkpoints drift before every expedition block rather than relying on last year's map.

Gap-reel choreography for partial leads. The half-mapped zone between season two's silt-out and season three's clean push ran 180m. The team pre-positioned a gap reel exploration kit at a known junction, so the closure dive didn't burn push time on reel-swaps. EchoQuilt flagged the junction automatically as the highest-connectivity point from the prior two seasons' patches.

Acoustic signature staging. Before each season's push dive, the team played back archived scrubber clicks and flow whine in a briefing. Divers matched their CCR's current signature against the quilt's stored baseline and recalibrated inhalation rhythm to keep solenoid firing outside the 6-10 Hz band where the siphon's flow signal lived. Without that calibration, their own rebreathers would have masked the very signal they needed to close the sump.

Scrubber chemistry as a temporal marker. Sofnolime canisters from different production lots produce subtly different acoustic signatures as they exhaust — the granule fracture pattern shifts as the absorbent breaks down. Tagging each push dive with the canister lot number and the runtime hours on the scrubber lets the EchoQuilt engine treat scrubber-noise drift as a known quantity rather than as ambient mystery. The Lot team kept a binder of canister lot stickers stapled to dive logs across all three seasons, and the cross-reference made the season-three quilt 12-15 percent quieter in the diver-self-noise band than the season-one quilt — a real, measurable improvement that came from logging discipline rather than equipment changes.

Air-bell handoff protocols. When the season-three closure dive surfaced in an air bell at 618m, the diver had two minutes of standing-up time before the deco clock forced descent. Those two minutes were the most valuable of the entire three-year project. The team had pre-rehearsed exactly which observations to make: photo of the air bell ceiling at three angles, voice note of any faint surface drip indicating dry-cave continuation upstream, single fingertip touch on the floor sediment to test whether it was bedrock or loose silt. Every observation went into the EchoQuilt voice channel and was tagged with the dive's master timestamp. The air-bell minute is where multi-season siphon projects either succeed in finding the next lead or close out the system entirely, and it has to be choreographed in advance.

The three-season siphon closed on push dive thirty-seven. The survey went to the French Federation in a single file.

Ready to Close Your Own Half-Mapped Siphon?

French sump-push teams, NSS-CDS expedition leads, and WKPP-style project directors running multi-season closures have the hardest job in survey cave diving — you can't rehearse what you can't see, and every season you lose to silt-out erodes institutional memory. EchoQuilt keeps each push as a patch in a persistent quilt, so season three builds on season one's hard-won ambient data instead of starting over. Join the Waitlist for Cave Diving Survey Teams and help us shape the multi-season replay tooling before your next push window opens. We prioritize teams with unclosed siphons going into their third consecutive expedition year.

Share your siphon's current penetration depth, your team's CCR rotation (JJ-CCR, rEvo, AP Inspiration), your scrubber-lot logging discipline, and your federation affiliation (FFESSM, NSS-CDS, WKPP, QRSS, GUE Tech), and we will scope a per-rig calibration set for your scrubber-canister inventory, the federation-shared patch schema for cross-season handoff, the air-bell observation choreography template French Lot-basin and Ressel teams use, and the multi-season replay tools you can use during your next push window. The third-season quilt should know more than the first-season quilt, not less.