This note is based on the exact-author corpus in output/realclimate_whut_comments.json and output/realclimate_whut_comments.md, not the earlier over-collected archive. The current corpus contains 931 comments from 2016-2026, almost all under Paul Pukite (@whut), with a few Paul Pukite / Paul Pukite (@WHUT) variants.
All comment references below point into the local archive: output/realclimate_whut_comments.md.
The recurring model package is:
- ENSO, QBO, and related indices are not primarily chaotic autonomous oscillators.
- They are better treated as forced, non-autonomous responses to lunisolar/tidal inputs.
- The observable patterns are standing-wave or sloshing behaviors constrained by equatorial geometry and boundary conditions.
- The right way to test these models is historical cross-validation, not just waiting years for forward predictions.
- The same forcing logic may extend to mean sea level, AMO, delta-LOD, Chandler wobble, and other geophysical indices.
This is the most persistent claim in the archive. Pukite repeatedly argues that ENSO and QBO should not be modeled as self-excited internal oscillators that spontaneously emerge from a chaotic background. Instead, they should be treated more like tides: externally forced, phase-sensitive responses.
- ENSO and QBO are non-autonomous systems.
- Their apparent irregularity reflects complicated forcing and response, not irreducible chaos.
- Mainstream GCM treatments too often assume autonomous oscillation where explicit forcing should be included.
- Comment 652930 says QBO is likely forced by seasonally aliased monthly tidal cycles and argues that both QBO and ENSO need explicit forcing inputs.
- Comment 660487 argues that ENSO has deterministic properties and cites signal-processing work against the "completely unpredictable" framing.
- Comment 682647 explicitly compares ENSO to ocean tides and says boundary conditions and forcing matter more than butterfly-effect sensitivity.
- Comment 686879 says ENSO should be tested as a deterministic process by feeding in lunisolar forcing directly.
The central modeling move is to replace internal variability first with forced response first. That is the conceptual bridge connecting the RealClimate comments to the later LTE / reduced-order modeling work elsewhere in this repo.
Pukite's preferred forcing is not vague "natural variability." It is much more specific: lunisolar gravitational/tidal forcing, often acting through thermocline motion, subsurface waves, or seasonally aliased harmonics.
- QBO can be explained by seasonally aliased lunar tidal forcing.
- ENSO switching is linked to subsurface waves and tidal forcing.
- Lunar/solar cycles are measurable in other geophysical observables such as delta-LOD.
- The thermocline is the especially sensitive part of the ocean-atmosphere system.
- Comment 660635 lays out the QBO case most explicitly: wind acceleration, latitudinal displacement tied to tidal elevation, and seasonal aliasing of lunar tractive forces to match the observed QBO period.
- Comment 680663 says lunisolar gravitational forcing is emerging as a mechanism for ENSO and other flow patterns.
- Comment 721359 cites JPL work on lunar tidal forcing in ocean winds, rainfall, OLR, wave height, and tropical instability waves.
- Comment 760628 argues that delta-LOD shows a faithful lunar-cycle response, including the 18.6-year nodal envelope.
- Comment 788212 ties AMO/rotation-rate variations to the 14.765-day Msf tidal factor and quotes a Perigaud/JPL line of argument.
This is not a generic "the Moon matters somehow" position. The comments repeatedly narrow the mechanism to specific harmonic tidal factors, aliasing, subsurface thermocline response, and angular-momentum exchange.
Another strong theme is that these climate behaviors should be understood geometrically. Pukite often frames ENSO as a Pacific standing wave or sloshing thermocline dipole, and QBO as a highly symmetric equatorial behavior with fewer degrees of freedom than people assume.
- ENSO is a standing-wave or sloshing dipole tied to Pacific boundary conditions.
- QBO is simpler than ENSO because it has stronger symmetry and fewer incommensurate forcings.
- Equatorial structure can insulate these modes from turbulence.
- Hemispheric asymmetry and nodal crossings matter for how forcing appears.
- Comment 652930 calls ENSO a single standing-wave behavior over the equatorial Pacific.
- Comment 686879 describes ENSO as a sloshing dipole in the equatorial Pacific thermocline.
- Comment 704565 argues that ENSO and QBO are standing waves protected by equatorial/topological structure rather than dominated by turbulence.
- Comment 820565 says an idealized QBO can be captured by a mathematical construction with the right topology, then mapped to empirical observations.
This is where the comments most clearly anticipate the repo's later emphasis on equatorial reduction, standing-wave structure, and low-order forced dynamics.
Pukite is not arguing for a black-box model. The methodological preference is closer to: use machine learning or signal processing to discover candidate forcings, then validate reduced physical models by cross-validation on long time series.
- Machine learning is useful for discovering forcing terms or dead ends, not as a substitute for mechanism.
- Curve fitting is acceptable when it extracts physical parameters rather than merely interpolating.
- Historical out-of-band cross-validation is the right standard in Earth science where controlled experiments are impossible.
- Overfitting is a risk, but the answer is stronger cross-validation, not abandoning reduced models.
- Comment 652930 presents machine learning and data mining as tools for discovering QBO/ENSO forcings.
- Comment 743109 describes cross-validation of a fluid-dynamics-based model of cyclic climate indices.
- Comment 788923 says there is already enough data for detailed cross-validation and that the bottleneck is methodological adoption.
- Comment 792784 describes train-on-one-interval, test-on-out-of-band intervals as the practical antidote to overfitting.
- Comment 793206 argues that for slow oscillations like ENSO, validation by future prediction alone is too slow and must be supplemented by intensive cross-validation.
This is important because the archive does not just say "fit cycles." It says: use structure-aware fits, discover forcing harmonics, and validate rigorously on held-out intervals.
By the later comments, the same framework is being extended far beyond ENSO and QBO.
- Mean sea level residuals may encode tidal/common-mode forcing plus links to ENSO and NAO.
- AMO may contain a tidal component.
- Earth rotation and Chandler wobble provide cleaner geophysical evidence for torque/forcing.
- A unified forcing picture may connect multiple geophysical indices.
- Comment 783391 says Chandler wobble should be viewed as a forced response to lunar and solar nodal cycles.
- Comment 822693 discusses mean sea level as a 19-year averaging problem because of tidal phase and amplitude cycles.
- Comment 837451 describes cross-validating residual mean-sea-level oscillations at dozens of long-record ports.
- Comment 839465 argues that multi-year sea-level oscillations and ocean indices can be unified by a common tidal-forcing mechanism.
- Comment 840971 links internal waves, ENSO, sea-level variation, and NAO-correlated Baltic sea-level cycles in one forcing chain.
The later comments broaden the proposal from "tidal forcing might matter for ENSO/QBO" to "a common forcing vocabulary may explain multiple coupled geophysical observables."
Across the corpus, Pukite is usually arguing against four opposing ideas:
- ENSO/QBO are inherently chaotic and not worth deterministic modeling.
- Wind anomalies are the primary initiator rather than a downstream response.
- Tidal/lunar inputs are too weak or too slow to matter at interannual scales.
- Validation should rely mainly on future prediction rather than historical cross-validation.
The corrected RealClimate corpus supports a much sharper summary than the earlier mistaken archive:
- Pukite's core proposal is a forced-response model of interannual climate variability.
- The preferred forcing source is lunisolar/tidal, often acting through subsurface ocean dynamics.
- ENSO and QBO are treated as deterministic standing-wave or sloshing responses, not primarily chaotic autonomous oscillators.
- Cross-validation on long historical time series is his preferred validation standard.
- The same framework is repeatedly extended to sea level, AMO, delta-LOD, and Chandler wobble.
These are the external links that most clearly define the mature version of the model package described in the comments.
| Link | Why it matters |
|---|---|
| https://www.nature.com/articles/s41598-019-49678-w | The most-cited third-party support link in the archive: subsurface ocean waves and likely lunar tidal forcing for ENSO switching. |
| https://agupubs.onlinelibrary.wiley.com/doi/10.1002/9781119434351.ch11 | Pukite's QBO chapter in Mathematical Geoenergy; cited as the long-form model for QBO. |
| https://agupubs.onlinelibrary.wiley.com/doi/10.1002/9781119434351.ch12 | Pukite's ENSO chapter in Mathematical Geoenergy; cited as the long-form model for ENSO. |
| https://esd.copernicus.org/preprints/esd-2020-74/ | "Unforced Variations should be a Forced Response"; used in comments on thermocline sensitivity and Chandler wobble forcing. |
| https://geoenergymath.com/2024/03/25/proof-for-allowed-modes-of-an-ideal-qbo/ | Mature statement of the topological / allowed-modes argument for an idealized QBO. |
| https://geoenergymath.com/2022/01/14/sea-level-height-as-a-proxy-for-enso/ | Repeatedly cited for using sea-level height as an ENSO proxy and for tying sea-level variability back to climate indices. |
| https://pukite.substack.com/p/mean-sea-level-models | Used for the expanded mean-sea-level modeling program and the claim that tidal forcing synchronizes multiple observables. |
| https://geoenergymath.com/2024/09/23/amo-and-the-mt-tide/ | Representative later link extending the same framework to AMO. |
| https://geoenergymath.com/2024/11/10/lunar-torque-controls-all/ | Representative late-stage synthesis link for the broad torque/forcing view. |
| https://github.com/pukpr/GEM-LTE | Code/data repository linked in later comments for the broader LTE-style modeling framework. |
If you want to re-read the archive in a high-signal order, start here:
- Comment 652930 — QBO tidal forcing, ENSO biennial forcing, non-autonomous systems, ML
- Comment 660635 — compact QBO derivation and seasonal aliasing of lunar tractive forces
- Comment 682647 — Hawkmoth vs Butterfly framing; ENSO as boundary-forced like tides
- Comment 686879 — ENSO as deterministic sloshing thermocline dipole
- Comment 743109 — cross-validation of cyclic climate-index models
- Comment 760628 — delta-LOD and the 18.6-year nodal envelope
- Comment 783391 — Chandler wobble as a forced response
- Comment 820565 — allowed modes / topology argument for ideal QBO
- Comment 839465 — mean sea level, tidal factors, and common-mode forcing