Five load-bearing claims · Discriminant-oriented assessment
The apparent "two annual flood waves" in parts of the Amazon system are not a stable semiannual basin-scale regime at major downstream gauges; rather, they are shoulders / multi-peak structures arising from (i) non-sinusoidal seasonal forcing (higher harmonics) and (ii) distributed routing + storage with time-varying kernels, with strong interannual modulation by ENSO/Atlantic variability.
Primary gauges for basin-integrated behaviour: Óbidos (Amazon mainstem) and Manaus (Rio Negro stage; confluence influence). Óbidos metadata and summary statistics from UNH/GRDC composite station page [1].
A year y is classified as "two-maxima" if the seasonal component of \(Q_y(t)\) has two distinct local maxima within one annual period after removing weather-band variability (e.g., monthly means or low-passed daily series).
In \(t \in [0,T)\), \(\mathrm{d}Q/\mathrm{d}t = 0\) has \(\geq 4\) roots and \(\geq 2\) of them satisfy \(\mathrm{d}^2Q/\mathrm{d}t^2 < 0\) (two maxima).
Mathematical limitation (EC-gate)
Pass (formal)Claim: A strictly single-frequency harmonic forcing through an LTI kernel cannot generate two maxima per period.
If \(Q(t) = B\cos(\omega t - \psi)\), then \(\mathrm{d}Q/\mathrm{d}t = 0\) yields exactly two roots per period → one maximum, one minimum.
Consequence: Any "two floods per year" explanation must invoke (a) higher harmonics in forcing, or (b) nonlinearity / time variance in routing, or both.
Corrected minimal seasonal model (1st + 2nd harmonic sufficiency)
Pass (formal)Claim: A corrected reduced model with annual + semiannual components is the minimal linear model that can exhibit two maxima per year.
Define \(r = B_2/B_1\), \(\delta = \psi_2 - 2\psi_1\). Two maxima can occur only if \((r, \delta)\) lies in a "multi-root" region of \(\sin x + 2r\sin(2x + \delta) = 0\).
Notes: This is a structural discriminant; thresholds are \(\delta\)-dependent and should be assessed numerically for root counts when annual-by-annual fits exist. See Model for the full derivation.
Óbidos is dominantly annual; semiannual term is small
Provisional-PassClaim: At Óbidos, the climatological seasonal discharge is well described by annual + semiannual harmonics with small r, consistent with a unimodal annual hydrograph.
Evidence anchor: UNH/GRDC station summary [1] provides long-period discharge statistics and mainstem length (used for routing scale).
Verifier-derived parameters (from monthly climatology, 1928–1996):
\(B_1 \approx 6.4 \times 10^4 \;\mathrm{m^3/s}\), \(B_2 \approx 8.3 \times 10^3 \;\mathrm{m^3/s}\), \(r \approx 0.13\), \(\delta \approx 0.9 \;\mathrm{rad}\)
With \(r \ll 1\), the derivative equation is generically in the unimodal regime for most \(\delta\).
Upgrade gate: Requires reproducible fit archive with station ID, time window, harmonic decomposition code, and output \((r, \delta)\) table — tagged as dependency of F-1. Staleness policy: re-evaluate after 6 months or upon availability of new gauge data, whichever comes first.
ENSO/Atlantic variability modulates flood characteristics strongly
Pass (as stated)Claim: Modes of climate variability (ENSO type and tropical Atlantic variability) produce significant changes in flood magnitude and duration across the Amazon basin, with timing effects more heterogeneous.
Evidence anchor: Hydrology and Earth System Sciences study [2] explicitly reports significant changes in flood magnitude and duration, with dependence on ENSO "type."
Distributed routing + backwater/floodplain storage matters
Pass (as stated)Claim: Hydrologic/hydrodynamic routing with floodplains and backwater effects is required to reproduce observed lags/attenuation; simple routing misses key physics.
Evidence anchor: Large-scale hydrologic/hydrodynamic modelling of the Amazon using MGB-IPH [3] with validation against observations and remotely sensed inundation.
At Óbidos, the climatological discharge is strongly annual; a small semiannual component exists but is not indicative of a stable "two floods per year" regime. [1]
Explaining true two-maxima behaviour requires at least annual + semiannual structure (or nonlinearity / time variance); single-harmonic LTI cannot do it.
ENSO/Atlantic variability significantly modulates flood characteristics; therefore year-to-year variability in any "secondary peak / shoulder" is expected. [2]
Distributed routing with floodplains/backwater is load-bearing for timing/attenuation; kernel drift is plausible and should be parameterised if long-term change is claimed. [3]
⚠ Citation anchors marked TODO require resolution to stable URIs or DOIs
before public commit. See
Falsification for the upgrade path.