Climate Summary — Gravitational Coupling, Not Insolation, Drives Earth’s Climate
Climate is forced by gravitational coupling among solar-system bodies. Solar insolation is one channel through which that coupling reaches Earth — but the rhythm itself, captured by the 8H lattice, is the more complete description.
This page synthesises the empirical case made across the Climate Formula, L1 Attribution, Insolation Null Test, and Related Work pages. Three headline numbers:
- R² = 0.87 — the 8H integer-divisor lattice (L1) alone, on post-MPT LR04 ice-volume variance
- R² = 0.05 — classical Berger 1978 insolation features alone, on the same record (17× less)
- ΔR² = 0 — adding Berger insolation features (with Laskar 2010’s wide-range eccentricity) to L1+L2+L3
And one structural finding from L1 Attribution: 0 of 32 L1 integers fully agree with Berger on both the planet name AND the mechanism. The two frameworks agree on which periods exist (the integer divisors of 8H) and disagree on which planets drive each beat. The lattice and the Berger parameterization are not independent forcings — they are different projections of the same physics, and the lattice is strictly more expressive.
1. The two paradigms
There are two competing answers to “what drives Earth’s climate variability over 10⁴–10⁷-year timescales?“
1.1 The classical Milankovitch paradigm (Berger 1978, Laskar 2004)
Climate is driven by changes in solar insolation received at high northern latitudes during summer. Insolation is parameterized by Earth’s three slowly-varying orbital elements:
- ε(t) — obliquity (axial tilt, ~41 kyr period)
- e(t) — eccentricity (~95 / 405 kyr beats)
- ϖ(t) — longitude of perihelion (~23 kyr climatic precession)
These three are themselves derived from the gravitational coupling among the planets — Laskar 2004 publishes them as sums of fundamental-frequency beats g_j ± g_k, k ± s_j, etc. But in the classical paradigm, the climate-relevant quantity is the insolation those orbital elements produce.
1.2 The 8H-lattice paradigm (this framework)
Climate is driven by the gravitational rhythm of the entire solar system, parameterized as integer divisors of the 8H = 2,682,536 yr Solar System Resonance Cycle. Each integer corresponds to a specific planet–planet beat or a direct planet-cycle harmonic. The full L1 set is 32 integers (L1 Attribution gives per-integer attribution).
The two paradigms are not contradictory — they describe the same gravitational physics. They differ in:
- Granularity — L1 enumerates 32 integer-divisor frequencies; Berger collapses these into 3 time-domain functions
- Channel — Berger’s reduction implicitly assumes insolation is the sole transmission mechanism; the lattice paradigm is agnostic about transmission
- Expressivity — adding Berger insolation features to L1+L2+L3 yields ΔR² ≈ 0 (see Insolation Null Test); the lattice contains the variance the projection produces; the reverse is not true
2. Why the lattice is more complete
Berger’s three insolation features ε(t), e(t), ϖ(t) are themselves defined as Fourier projections of the same secular gravitational coupling that L1 enumerates. For example:
| Berger insolation feature | Decomposition into Laskar beats | Corresponding L1 integer |
|---|---|---|
| Eccentricity 95-kyr beat | g₄ − g₅ (Mars-Jupiter) | n = 28 |
| Eccentricity 405-kyr line | g₂ − g₅ (Venus-Jupiter) | off-lattice → L2 |
| Climatic precession 23.7 kyr | k + g₅ (Jupiter) | n = 113 |
| Climatic precession 22 kyr | k + g₂ (Venus) | n = 120 |
| Obliquity 41 kyr | k + s₃ (Earth nodal) | n = 65 |
| Obliquity 41.3 kyr | k + s₄ (Mars sub-peak) | n = 68 |
Every Berger insolation peak is somewhere in the L1 lattice (see L1 Attribution for the full mapping). The lattice contains all of Berger, plus integer-divisor structure that Berger’s reduction does not surface (planet-planet beats not historically considered insolation-relevant, like 8H/16 Mars Axial, 8H/35 Earth-Mercury-Saturn 3-term beat, etc.).
This is why adding Berger insolation features to L1+L2+L3 yields ΔR² ≈ 0: the information is already there, parameterized at a finer-grained level.
3. The empirical case in one table
All numbers from the canonical regression scripts (see Climate Formula for architecture and Insolation Null Test for the augmented-regression detail):
| Test | LR04 (post-MPT, 0–1000 kyr) | LR04 (0–500 kyr, Laskar window) | EPICA CO₂ (0–800 kyr) |
|---|---|---|---|
| L1 alone (32 lattice integers) | R² = 0.870 | R² ≈ 0.93 | R² ≈ 0.80 |
| L1+L2+L3 (canonical formula) | R² = 0.8735 | R² = 0.9424 | R² = 0.8452 |
| Berger insolation alone (model e/ϖ) | R² = 0.049 | R² = 0.188 | R² = 0.096 |
| Berger insolation alone (Laskar e/ϖ) | — | R² = 0.293 | R² = 0.172 |
| L1+L2+L3 + Berger insolation (model) | R² = 0.8776 → ΔR² = +0.0041 | R² = 0.9436 → +0.00123 | R² = 0.8494 → +0.0042 |
| L1+L2+L3 + Berger insolation (Laskar) | — | R² = 0.9424 → +0.00000 | R² = 0.9230 → +0.00001 |
Three lines tell the story:
- The 8H lattice (L1) carries the variance. R² = 0.87 on post-MPT LR04 — the lattice alone, 32 sinusoids at fixed gravitational-rhythm frequencies.
- Classical insolation alone explains very little. R² = 0.05 (with our model’s e) → R² = 0.29 (with Laskar’s full-range e). Significant absolute, but a fraction of what L1 captures.
- The two parameterizations carry overlapping, not independent, information. Adding Laskar’s better-parameterized insolation to L1+L2+L3 yields ΔR² = 0.000 on LR04 and ΔR² = 0.00001 on EPICA. The lattice already contains all of it.
The third line is the crux. Berger insolation’s R² = 0.29 (Laskar e) is real — it’s just already inside L1.
The canonical formula on LR04 δ¹⁸O over the past 700 kyr. R² = 0.87 — the 32-integer 8H lattice + 3-line carbon thermostat + 6-step transitions captures most of the post-MPT climate signal. Lattice positions are fixed by orbital geometry; only the amplitudes are fitted to the data.
4. What this conclusion does — and does not — claim
4.1 What this claim says
- The 8H integer-divisor lattice is a more complete regression basis for LR04 / EPICA / CenCO2PIP variance than the classical Berger insolation parameterization
- Climate is forced by the gravitational coupling rhythm of the solar system — a many-body phenomenon involving Mercury through Neptune, not Earth-Sun geometry alone
- The classical Milankovitch paradigm is empirically correct within its scope (its named peaks are real lattice integers), but it is a projection, not the fundamental description
4.2 What this claim does NOT say
This claim is explicitly bounded:
- ❌ It does NOT say insolation is irrelevant. Berger 1978’s quantitative theory of insolation at any latitude/day is correct — its peaks correspond to specific lattice integers
- ❌ It does NOT say Earth-Sun geometry is unimportant. ε(t), e(t), ϖ(t) are real orbital elements that do shape Earth’s seasonal heating
- ❌ It does NOT propose a new transmission mechanism between gravity and climate. The rhythm is what L1 captures, and the rhythm explains the variance regardless of channel mix
- ❌ It does NOT claim universal explanatory power. R² collapses across regime boundaries (pre-MPT vs post-MPT). The framework is descriptive within regimes, not predictive across boundary-condition shifts.
4.3 Conclusion
Climate is forced by the gravitational coupling among solar-system bodies. Insolation is one channel through which that coupling reaches Earth. The 8H integer-divisor lattice is a more complete description of the rhythm than the classical Berger insolation parameterization — strictly more expressive, with no information lost relative to the 4-feature parameterization.
5. An observation about the current era
A consequence visible directly in the Climate Formula Explorer modal: the L2 405-kyr carbon-thermostat layer in the canonical formula sits near a warming peak around the present era. This is a fitted observation from LR04 — the layer’s phase is set by the regression, and the cycle position then falls where it falls.
Quantitatively:
| Quantity | Value | Source |
|---|---|---|
| L2 fundamental period | 405 kyr | g₂ − g₅ (Venus–Jupiter eccentricity beat) |
| L2 amplitude (post-MPT fit) | ~0.5–1.0 °C peak-to-trough | Climate Formula |
| Current phase | Near warm peak | Climate Formula Explorer |
| L2 rate of change at peak | < 0.005 °C / century | Derivative of fitted sinusoid |
What this observation does and does not say — to preempt misreading in either direction:
| Statement | Status |
|---|---|
| The model’s L2 layer is currently near a warm peak | ✅ Directly visible in fitted output |
| The natural pre-industrial baseline (~1850 CE) was slightly elevated (~0.3–0.5 °C) above the long-term natural mean | ✅ Implied by L2 phase position |
| L2 explains the ~1.2 °C industrial-era warming | ❌ False — L2’s rate is ~300× too slow (industrial-era ~0.7 °C / century vs L2 maximum ~0.005 °C / century) |
| Anthropogenic CO₂ warming is overstated | ❌ Not implied — L2 affects baseline level, not recent rate |
Two distinct natural-cycle signals point opposite directions for the current era:
- Precession-driven NH summer insolation (~23 kyr) is in a declining phase since the Holocene Climate Optimum (~10 kyr BP) → mainstream view: natural trend is cooling
- 405-kyr eccentricity carbon thermostat (our L2) is near peak → our framework: natural baseline slightly warmer than long-term mean
Both signals are real; they sit at different timescales. Mainstream Holocene-attribution discussions emphasise the first (precession-cooling); the second (405-kyr-warming-baseline) is rarely cited outside cyclostratigraphy.
This is connected to the “Holocene Temperature Conundrum” — an active scientific debate (Nature 2022 ) about why proxy records and climate models disagree on the magnitude and direction of Holocene-era temperature trends. See Related Work §7.4 for the literature positioning.
The observation is offered here as an empirical fact about the model’s fitted output — not as a contribution to the anthropogenic-warming debate.
6. Empirical anchors
| Quantity | Value | Source |
|---|---|---|
| 8H Solar System Resonance Cycle | 2,682,536 yr | Fundamental Cycles |
| L1 lattice integers | 32 | L1 Attribution |
| L1 alone, post-MPT LR04 | R² = 0.870 | Climate Formula |
| L1+L2+L3, post-MPT LR04 | R² = 0.8735 | same |
| L1+L2+L3, EPICA CO₂ | R² = 0.8452 | same |
| L1+L2+L3, CenCO2PIP (0–66 Ma) | R² = 0.7626 | same |
| Berger insolation alone, LR04 0–500 kyr | R² = 0.293 (Laskar e) | Insolation Null Test |
| Berger insolation added to L1+L2+L3 | ΔR² = 0.00000 (LR04), 0.00001 (EPICA CO₂) | same |
| L1 dual-attribution rate | 32 / 32 integers | L1 Attribution |
All numbers reproducible with python3 scripts/milankovitch_*.py in the 3d repository . Deterministic, no random seeds.
See also
- Climate Formula — architecture, per-regime fits, forward projection
- L1 Attribution — per-integer Berger vs Holistic top-1 attribution
- Insolation Null Test — empirical basis for the ΔR² = 0 claim
- Related Work — position of this framework in the 2024 climate-forcing literature
- Fundamental Cycles — what each L1 integer is
- Fibonacci Laws — the structural identities the integer divisors encode