The channel carries a fixed tidal/surge volume flow rate Q (m³/s) — water that must get through the barrier one way or another while the tide fills or drains the basin behind it. As gate leaves lower into the flow, the total open cross-section A(t) shrinks. For an (approximately) incompressible fluid, conservation of volume flow rate — the continuity equation — links flow rate, area and velocity:
Q = A · v → v(t) = Q / A(t)
Since Q stays roughly constant while the tide keeps pushing, shrinking A(t) forces v(t) upward — the same water has to squeeze through a smaller gap in the same time. This is exactly the scour hazard that real storm surge barriers (the Oosterscheldekering, the Maeslantkering, the Thames Barrier) are engineered around: a fast, concentrated jet through the last open gap can undercut the foundation.
- Sequential (edge-first) closes one gate fully before starting the next, working inward from the banks. Every gate that's still open keeps carrying its full share of Q, so the very last gate to seal must — for a moment — pass the entire flow through a single narrow slot, producing a sharp, localized velocity spike.
- Synchronized lowers every leaf together, so A(t) shrinks smoothly and evenly across the whole barrier width for the whole duration. The velocity still rises as closure completes, but the rise is shared across every gate rather than concentrated in one — this is why real barriers generally prefer graduated, coordinated closure over a single-file sequence.
Displayed velocity is clamped for readability as A(t) → 0 at the instant of full closure (the true value is unbounded in this idealized constant-Q model; in reality Q itself tapers as the head difference across the barrier equalizes near full closure).