Why the Cap-Ferret sand spit is collapsing

benjamin gasque gasque 

The Cap-Ferret sand spit (SW France) exhibits a pattern of coastal instability that
combines chronic shoreline retreat (8.7 m·yr⁻¹ at the tip, Robinet et al. 2025), sudden
vertical collapses of emplaced structures (WW2 blockhaus 2024, 2026; oyster-farm
sector 1936, 1977; Hortense promenade 1999, 2000, 2014, 2019), and progressive
deepening of submarine pits (Hortense-Pointe depression volume nearly doubled
between 2005 and 2015; BRGM 2019). We show that these three signatures cannot be
simultaneously explained by the conventional sediment-budget framework, which is
energetically under-determined by approximately 30 % when benchmarked against the
observed channel scour rate under conservation of mass.
We formalize four candidate mechanisms, independently testable: (H1) submarine
groundwater discharge focused through the Sables-des-Landes aquifer exit, (H2) postdredging
cascade driven by the 2003-2005 bypass operations (3.1 × 10⁶ m³), (H3)
cyclic granular compaction under multi-frequency coupling during storms, and (H4)
localized aquifer depressurization — decomposed into H4a (municipal AEP pumping,
active, testable) and H4b (deep Lavergne reservoir, falsified as dominant source by
mass balance). Three independent lines of evidence converge on the H4a attribution:
(i) H/V Nakamura analysis at the RESIF station FR.LRVF identifies a stable
fundamental resonance f₀ = 0.323 Hz with A₀ ≈ 5, stable across tidal extremes (Δf₀ =
0.0000 Hz), consistent with a saturated 80-150 m thick confined aquifer rather than
free-gas; (ii) an EGMS spatial partition against two off-peninsula control sites
(Andernos, Biganos) shows a 6-7 mm differential subsidence of the peninsula relative to
its Bassin surroundings over 2019-2023, rejecting a regional signal; (iii) replication on
the 2019-10-21 event shows a spatial cascade (Bartherotte first, Mimbeau second,
Horizon delayed) consistent with the inverted aquifer-roof geometry. Three closed-form
closure tests further constrain the framework: (iv) cumulative Coulomb stress transfer
falsifies a systematic Parentis fault reactivation (2/6 events exceed ΔCFS ≥ 10 kPa —
secondary contributor only); (v) Taylor infinite-slope analysis on the Hortense pit
bathymetric record 2003-2024 shows a 26 % progressive loss of safety margin under
storm-peak pore pressure, supporting H2; (vi) a Palmgren-Miner cumulative damage calculation cleanly partitions the six instrumented events into storm-driven direct
compaction (3/3, D ≫ 1) and calm-sea events requiring pre-accumulated multi-winter
fatigue (3/3, D ≪ 1), validating H3 as a bimodal discriminator. Each hypothesis
nonetheless retains testable quantitative signatures that discriminate it from
conventional explanations.
We demonstrate that an integration of pre-industrial bio-engineering — palisade
techniques introduced by Brémontier (1786) and fluvial fascine-work documented since
the 16th century — addresses four physical mechanisms simultaneously (grain
trapping, energy dissipation, creep blocking, piping interruption) that are recognized
separately but never combined in the peer-reviewed coastal literature. Applied
differentially to the three distinct geomorphological settings of Cap-Ferret (exposed
Gasque · Cap-Ferret integrated framework · v1.3
4 / 47
dune / tidal-channel talus / low-emergence spit), this bio-engineering toolbox delivers
equivalent protective performance at 5 % to 3 % of the cost of conventional hard
engineering, while preserving ecosystem services.
The paper does not propose a new theoretical mechanism. It formalizes the crossdisciplinary
integration of physical mechanisms that remain compartmentalized in the
current literature, and demonstrates that site-specific combinations based on 240 years
of empirical validation on the French Atlantic coast offer a robust, reversible, and lowcost
alternative to the hard-engineering paradigm.