APPENDIX - Fringe Spacing Geometry Module¶

Purpose: Convert the “interferometry grid / invisible tripod” idea into a quantitative, falsifiable geometry claim. This appendix is not a claim that any specific emitter existed; it is a specification of what the model would have to match if an interference geometry is asserted.

Outputs: (1) fringe spacing as a function of crossing angle and wavelength, (2) reverse-calculated frequency bands for candidate feature scales, (3) explicit assumptions and falsifiers, (4) a minimal correlation test protocol.

0. Definitions¶

• Crossing angle (\(\theta\)): the angle between the two incoming wave vectors at the target region (equivalently, the angle between the two source bearings as seen from the target, under a plane-wave approximation).
• ENE directional proxy: a bookkeeping label for the east-northeast arrival sector at the WTC used by the reconstruction’s geometry (bearing ~79.3° from true north). It is not a site attribution.
• Erin-sector proxy: a bookkeeping label for the southeast arrival sector associated with Hurricane Erin’s carried role in the reconstruction’s geometry (bearing ~149.7° from true north), treated as an external input with bounded uncertainty.
• Wavelength (\(\lambda\)): the carrier wavelength of the two coherent fields.
• Fringe spacing (\(d\)): spacing between adjacent maxima (or minima) in the interference pattern in the plane normal to the bisector direction.

1. Fringe spacing for two coherent plane waves¶

For two equal-frequency plane waves with crossing angle \(\theta\), the interference fringe spacing is:

Equivalently:

Implication: once \(\theta\) is fixed, the geometry fixes the ratio \(d/\lambda\).

2. Example ratio for the ENE↔Erin-sector crossing angle (~70.4°, from geodetic bearings)¶

If \(\theta \approx 70.4^\circ\), then:

So, under this assumption:

Important: The crossing angle \(\theta \approx 70.4°\) is derived from geodetic bearings (ENE proxy at 79.3° and Erin-sector proxy at 149.7° from the WTC; difference = 70.4°). The ENE proxy bearing is fixed. The 149.7° Erin bearing is a proxy anchor for the Erin-sector arrival direction — reproducible from an Erin-sector proxy point (e.g., ~35°N, 70°W) but not currently traceably sourced to a specific storm-track fix inside the dossier. The model may use an effective scattering centroid with offset Δθ; Δθ is uncertainty to propagate, not tune (see sensitivity analysis in Section 4.1).

3. Reverse calculation: “What frequency gives a fringe spacing comparable to feature scale \(d\)?”¶

Given an observed feature scale \(d_{feature}\) and a chosen \(\theta\), solve for:

Convert to frequency using a phase velocity \(v_{ph}\) (do not silently assume \(c\) if your propagation model implies otherwise):

3.1 First-order table (\(\theta=70.4^\circ\) from ENE↔Erin-sector bearings; \(v_{ph}=c\) as free-space assumption)¶

With \(d \approx 0.866\,\lambda \Rightarrow \lambda \approx d/0.866\):

“Fringes” across a 100 m scale	Implied fringe spacing \(d\)	Implied \(\lambda\)	\(f\) (if \(v_{ph}=c\))
0.5 fringe	200 m	231 m	1.30 MHz
1 fringe	100 m	115 m	2.60 MHz
1.5 fringes	67 m	77 m	3.90 MHz
2 fringes	50 m	58 m	5.20 MHz
3 fringes	33 m	38 m	7.81 MHz

Caution: This table is not proof of an emitter band. It is a geometry-to-band mapping conditional on (i) two-source coherent interference, (ii) a valid \(\theta\), and (iii) a defensible mapping from “feature boundary scale” to fringe spacing.

3.2 Candidate “multi-feature” frequencies (reverse-calculated, conditional on geometry)¶

If one treats certain feature scales as candidate “boundary spacings” (e.g., ~100 m knife-edge boundary scale, ~63 m tower face width, ~25 m cylindrical/bisection features), then the reverse-calculation produces the following frequencies:

Frequency (if \(v_{ph}=c\))	Fringe spacing \(d\)	Candidate feature scale
2.6 MHz	100 m	WTC 4 boundary-scale spacing (1 fringe across 100 m)
4.1 MHz	63.5 m	Tower face width (~63 m)
5.2 MHz	50 m	“2 fringe” boundary-scale spacing across 100 m
10.0 MHz	26 m	WTC 6 borehole / WTC 3 bisection-scale features (~25 m)

Interpretive guardrail: “Feature scale match” is not evidence by itself. It becomes meaningful only if a computed node/fringe map at those frequencies (with explicit uncertainties) correlates with multiple independent boundary features better than chance.

4. Sensitivity (why uncertainty must be explicit)¶

Because \(d/\lambda = 1/(2\sin(\theta/2))\), small changes in \(\theta\) change the inferred \(\lambda\) and therefore the inferred \(f\).

Minimum deliverable for robustness: report \(\theta\) with an uncertainty band and propagate it through the reverse-calculated frequency estimates.

4.1 Computed sensitivity: ±5° in \(\theta\) (65°–75°)¶

Using \(v_{ph} = c\) and the four candidate feature scales from Section 3.2:

Feature scale	\(\theta=65°\)	\(\theta=68°\)	\(\theta=70.4°\) (central)	\(\theta=72°\)	\(\theta=75°\)	Spread
WTC 4 boundary (~100 m)	2.79 MHz	2.68 MHz	2.60 MHz	2.55 MHz	2.46 MHz	±6.3%
Tower face width (~63 m)	4.43 MHz	4.26 MHz	4.13 MHz	4.05 MHz	3.91 MHz	±6.3%
2-fringe / borehole 2× (~50 m)	5.58 MHz	5.36 MHz	5.20 MHz	5.10 MHz	4.93 MHz	±6.3%
WTC 6 / WTC 3 (~26 m)	10.74 MHz	10.32 MHz	10.01 MHz	9.82 MHz	9.48 MHz	±6.3%

Key observations:

1. Low sensitivity: ±5° in \(\theta\) shifts all implied frequencies by only ±6.3%. The multi-feature alignment is not a fine-tuned result — it survives substantial angle uncertainty.
2. HAARP band (2.8–10 MHz) overlap check: At the angle extremes, the 100 m feature drops just below 2.8 MHz (at \(\theta=75°\): 2.46 MHz) and the 26 m feature just grazes above 10 MHz (at \(\theta=65°\): 10.74 MHz). The middle two features (63 m and 50 m) remain solidly in-band regardless of angle choice.
3. Implication for the dossier: The frequency estimates have a natural ±6% uncertainty floor from the crossing-angle input alone. Any frequency claim should carry this band, not a point value.

5. Assumptions (must be stated)¶

This module requires explicit assumptions. If any are unacceptable, the fringe approach should be treated as non-operative.

1. Coherence: the two sources are sufficiently phase-stable over the relevant time window to produce a persistent interference structure.
2. Two-source dominance: other paths/sources do not wash out the pattern or dominate coupling.
3. Propagation model: the mapping from source vectors to a local crossing angle at ground/structure scale is valid (waveguide / refraction / multipath must be bounded).
4. Coupling threshold model: “damage boundaries” correspond to crossing a coupling threshold that can plausibly track a fringe intensity gradient.

6. Falsification protocol (minimal)¶

1. Lock inputs: specify source coordinates/timing and compute \(\theta\) with uncertainty.
2. Pick a band: select candidate frequencies from the reverse-calculation (include uncertainty band from \(\theta\) and from \(v_{ph}\)).
3. Compute a map: generate a fringe/node intensity map across the WTC complex (with stated approximations).
4. Define “match”: pre-register what constitutes a match (e.g., boundary alignment within X meters across Y independent features).
5. Fail fast: if the predicted map does not correlate with the damage boundary features beyond chance, the “interference geometry explains boundaries” claim is rejected.

7. Where this is referenced¶

• White Paper: prediction-facing summary and test protocol (to avoid burying falsifiability in appendices).
• Bridge Mechanism Physics Appendix, Section J: listed as an open engineering requirement + test plan module.

8. Notes on band statements (avoid overclaim)¶

The reverse-calculated frequency range (2.6–10 MHz) overlaps with the known operational band of HAARP-class ionospheric heaters (2.8–10 MHz; see also Bridge Appendix, Section I for facility-scale parameters). It can be tempting to treat this overlap as a "fingerprint." In this dossier posture, that overlap should be treated as a hypothesis generator, not as a proof:

• The geometry module can at most constrain "if interference geometry explains boundary placement, then candidate \(f\) lives near 2.6–10 MHz under stated propagation assumptions and crossing angle."
• Whether any real-world HF facility (HAARP or otherwise) can supply the required power-at-target, coherence, and collateral containment is a separate link-budget and feasibility question handled elsewhere (see Bridge Appendix, Section J).

8.1 Interpretive note: the band-placement finding¶

The finding¶

The ENE↔Erin-sector crossing angle (\(\theta \approx 70.4°\)) places the four reverse-calculated feature frequencies at 2.6–10 MHz — straddling the HAARP operational window (2.8–10 MHz) almost exactly:

Frequency	Fringe spacing \(d\)	Feature matched
~2.6 MHz	~100 m	WTC 4 damage boundary scale (knife-edge at ~⅔ of building length)
~4.1 MHz	~63 m	Tower face width (63 m square cross-section)
~5.2 MHz	~50 m	WTC 4 boundary at 2-fringe spacing; WTC 6 borehole at 2× scale
~10 MHz	~26 m	WTC 6 cylindrical void diameter (~25 m); WTC 3 bisection strip width (~25 m)

The feature scales span a 3.85× bandwidth ratio (100 m / 26 m). The HAARP operational band spans a 3.57× ratio (10 MHz / 2.8 MHz). These match to within 8% — close enough that the feature cluster straddles the HAARP window with only slight overflow at both edges (2.60 MHz just below 2.8; 10.01 MHz just at 10.0).

How the crossing angle controls band placement¶

The fringe equation (\(f = c / 2d\sin(\theta/2)\)) means the crossing angle acts as a tuning knob — it slides the entire feature cluster up or down the frequency axis without changing its bandwidth:

Crossing angle \(\theta\)	\(f\)(100 m)	\(f\)(26 m)	Cluster lands at
40°	4.38 MHz	16.85 MHz	above HAARP band
70.4° (ENE↔Erin-sector)	2.60 MHz	10.01 MHz	straddles HAARP band
120°	1.73 MHz	6.66 MHz	below HAARP band

The ENE↔Erin-sector angle is the closest possible fit — it centers the cluster so the overflow is nearly symmetric around the HAARP window.

Methodological precision: what is and is not angle-dependent¶

The fringe equation has an important property: for any two feature scales \(d_1\) and \(d_2\), their frequency ratio is \(f_1/f_2 = d_2/d_1\), which is independent of \(\theta\). This means:

• Not angle-dependent (trivially guaranteed): any crossing angle simultaneously produces fringe spacings matching all four feature scales, and they always cluster within the same 3.85× bandwidth. A null-hypothesis test (100,000 random angles) confirmed this: 81.8% of random angles produce 4+ matches in the 1–15 MHz range.
• Angle-dependent (the non-trivial part): where on the frequency axis the cluster lands. The ENE↔Erin-sector angle places it at the HAARP window. A different angle would place it elsewhere.

The null question (stated precisely)¶

The proper null question is: "Given the observed building dimensions in the WTC complex, how likely is it that their max/min scale ratio (\(\approx 3.85\times\)) approximates the bandwidth ratio of a specific real-world transmitter class (\(\approx 3.57\times\)) to within 8%?" This depends on the distribution of building dimensions and the number of candidate transmitter windows in the RF spectrum — a quantifiable but not yet computed probability.

Bottom line¶

The ENE↔Erin-sector crossing angle positions the feature-matched frequency cluster so that it straddles the HAARP operational window, with bandwidth ratios matching to within 8%. Whether this represents a geometric fingerprint of operating parameters or a numerical accident is a well-defined, testable question. The falsification protocol (Section 6) remains the primary test: if a computed fringe/node intensity map at these frequencies does not correlate with observed damage boundaries better than chance, the interference-geometry explanation is rejected.

8.2 Fringe orientation finding (zero free parameters)¶

The finding¶

The fringe node lines run parallel to the bisector of the two source bearings — at 114.5° from true north (roughly ESE–WNW). The WTC complex was rotated approximately 29° east of true north, so the buildings' "east-west" faces run at 119° from north.

The offset is 4.5°.

This means fringe node lines are nearly parallel to the buildings' E-W faces. Damage boundaries within individual buildings would appear as strips or divisions running roughly E-W — dividing buildings into north and south portions. This matches the FEMA 403 observations:

• WTC 4: knife-edge boundary dividing the south wing (destroyed) from the north wing (intact) — consistent with a node line running parallel to the E-W building axis.
• WTC 3 (Marriott): bisection strip running roughly along the E-W building axis — consistent with a node line passing through the building in the same direction.
• WTC 6: cylindrical excavation — no preferred orientation (consistent with any fringe direction).

What this depends on¶

This test has zero fitted parameters. It depends only on:

1. ENE proxy bearing from WTC: 79.3° (geodetically determined)
2. Erin bearing from WTC: 149.7° (Erin-sector proxy bearing used by the reconstruction; treated as an external input with bounded uncertainty, not optimized)
3. WTC complex rotation: ~29° from true north (from site plans / NIST NCSTAR 1)

It does not depend on: phase offset (φ₀), exact building coordinates, frequency choice, or any fitting or optimization.

Statistical significance¶

If the bisector bearing were uniformly random, the probability of aligning within 4.5° of either building face direction (two chances, separated by 90°) is:

This is borderline significant — not dispositive on its own, but notable in conjunction with the band-placement finding (§8.1).

Sensitivity to Erin's position¶

The orientation match holds to within 10° for Erin bearings from ~140° to ~160° (a ±10° range). This ±10° sweep corresponds to effective centroid uncertainty (storm center vs. aloft scattering centroid): the physically relevant source direction may be offset by Δθ from the storm's geometric center, and the sweep bounds that uncertainty rather than post-hoc tuning. The match is not fine-tuned to Erin's exact position. However, the closest match (4.5°) occurs specifically near the proxy bearing of 149.7°.

Why this is a second independent constraint¶

The band-placement finding (§8.1) depends on the crossing angle (Erin bearing − ENE proxy bearing = 70.4°). The orientation finding depends on the bisector angle ((Erin bearing + ENE proxy bearing) / 2 = 114.5°). These are algebraically independent: the crossing angle constrains the difference of the two bearings, while the bisector constrains their sum. Both matching simultaneously is more restrictive than either alone.

8.3 Null model & multiple comparisons¶

The orientation finding (§8.2) and band-placement finding (§8.1) both require careful null-model specification to avoid post-selection bias. This section makes the null models explicit and accounts for multiple comparisons.

8.3.1 Orientation finding null model¶

Null hypothesis: The bisector direction of two randomly chosen source bearings has no preferred alignment with any architectural feature of the WTC complex.

Null distribution: If the bisector bearing \(\beta\) is uniformly distributed on \([0°, 360°)\), the probability of \(\beta\) falling within \(\pm\delta\) of either building face direction (two orthogonal axes separated by 90°, giving two independent chances) is:

For \(\delta = 4.5°\): \(P = 4 \times 4.5° / 360° = 0.050\) (5.0%).

What the tolerance \(\delta\) means: The 4.5° value is not a chosen tolerance — it is the observed offset between the bisector (114.5°) and the nearest building face axis (119°). The p-value answers: "If the bisector were random, how often would it land at least this close to a building face?"

Multiple-comparison correction: The orientation test was formulated as a single pre-specified test (bisector vs. building faces). No scanning over feature sets was performed. However, the following potential look-elsewhere effects should be noted:

• Number of candidate source pairs: Only one bearing pair (ENE proxy, Erin-sector proxy) was tested. If \(N\) source pairs had been scanned, the adjusted p-value would be \(p_{\text{adj}} = 1 - (1 - 0.050)^N\).
• Axis choices: Both building face axes (N-S and E-W relative to building rotation) were included as valid targets, which is the correct treatment — the fringe direction has no reason to prefer one axis over the other. This doubles the "hit zone" and is already accounted for in the formula above.
• Feature set definition: The WTC 4 knife-edge and WTC 3 bisection were identified from FEMA 403 before the orientation test was computed. If they had been selected after seeing the test result, this would constitute post-hoc feature selection and would invalidate the p-value.

Pre-registration requirement: For this p-value to be fully rigorous, the feature set (which damage boundaries count as "orientation evidence") must be pre-registered before the test is run. The current status is that the features were identified from FEMA 403 independently of the fringe calculation, but the analysis was not formally pre-registered.

8.3.2 Band-placement null model¶

The band-placement finding (§8.1) has a more subtle null structure. The proper null question (stated in §8.1) is:

"Given the observed building dimensions in the WTC complex, how likely is it that their max/min scale ratio (\(\approx 3.85\times\)) approximates the bandwidth ratio of a specific real-world transmitter class (\(\approx 3.57\times\)) to within 8%?"

This requires specifying:

1. Feature-scale distribution: What is the distribution of "interesting feature scales" in a typical large building complex? If all urban complexes naturally produce feature clusters with ~3–4× bandwidth ratios, the coincidence with HAARP's bandwidth is less surprising.
2. Transmitter-class distribution: How many specific real-world transmitter classes exist with defined operational windows in the 1–30 MHz range? Each is an independent "target" that increases the look-elsewhere probability.
3. Tolerance definition: The 8% match tolerance should be treated as the observed discrepancy, not a chosen threshold.

Current status: This null-model quantification has not been performed. The band-placement finding is reported as an observation — the null probability remains unquantified. This is an explicit gap that should be closed before the finding can be treated as statistically significant.

8.3.3 Joint significance¶

If both findings were independent (which they approximately are, since one depends on the crossing angle and the other on the bisector), the joint p-value for both matching simultaneously would be:

Since \(P_{\text{band}}\) is currently unquantified, the joint significance is also unquantified. However, the algebraic independence of the two tests (one constrains the sum of bearings, the other constrains the difference) means the two findings do provide genuinely independent constraints on the source geometry.

8.4 Geometry drift → phase drift (sensitivity derivation)¶

The reconstruction requires Erin's position to remain stable enough that the fringe pattern stays registered on the WTC buildings. This section derives the sensitivity of fringe registration to Erin's effective scattering centroid position.

8.4.1 Setup¶

The fringe spacing at the target is \(d = \lambda / (2\sin(\theta/2))\), where \(\theta\) is the crossing angle between the two arrival directions. A fringe maximum (or minimum) is a surface of constant path-length difference \(\Delta L = n\lambda\) between the two sources. A shift in the fringe pattern by fraction \(\alpha\) of one period requires a path-length change:

At \(\theta = 70.4°\) and \(d = 26\) m (\(f \approx 10\) MHz):

At \(d = 100\) m (\(f \approx 2.6\) MHz):

8.4.2 Translating to Erin displacement¶

If Erin's effective scattering centroid shifts by \(\delta r\) in an arbitrary direction, the path-length change at WTC depends on the geometry of the ENE↔Erin-sector↔WTC triangle. For small displacements:

where \(\phi_E\) and \(\phi_B\) are the angles between the displacement direction and the Erin→WTC and ENE→Erin directions respectively. In the worst case (displacement along the fringe-variation axis), \(|\cos\phi_E - \cos\phi_B| \sim 1\), giving:

Frequency	\(\lambda\)	Half-fringe shift requires Erin displacement of
2.6 MHz	115 m	~58 m
5.2 MHz	58 m	~29 m
10 MHz	30 m	~15 m

8.4.3 Interpretation¶

• At 10 MHz, a ~15 m shift in Erin's effective scattering centroid shifts fringes by half a period. This is stringent — comparable to (or tighter than) typical uncertainty in publicly reported storm-center positions (though the "effective scattering centroid" is a different quantity that depends on the ionospheric conductivity distribution, not the storm center).
• At 2.6 MHz, the tolerance relaxes to ~58 m — still tight but more forgiving.
• During the event window (~102 minutes), Erin moved ~16 km (at ~6 mph). If the effective scattering centroid tracks the storm center, this represents many fringe periods of drift at 10 MHz but fewer at 2.6 MHz.

8.4.4 What this implies¶

The sensitivity analysis creates a tension: the highest-frequency fringes (10 MHz, \(d = 26\) m) require positional stability that may be difficult to maintain over 102 minutes, while the lowest-frequency fringes (2.6 MHz, \(d = 100\) m) are more robust. This could mean:

1. The effective fringe pattern is dominated by lower frequencies where positional tolerance is more forgiving.
2. The "effective scattering centroid" is NOT the storm center but rather a feature of the ionospheric conductivity structure (e.g., the centroid of the FAC-enhanced region) that may be more stable than the storm's surface position.
3. Higher-frequency features require tighter geometry lock — which may explain why Phase I preconditioning (Erin stalling) is a reconstruction prerequisite.

8.4.5 Open items¶

• The relationship between Erin's surface position and its "effective scattering centroid" for ionospheric re-radiation is unknown. This is a critical gap.
• The rate at which the fringe pattern drifts as Erin moves should be compared to the rate at which the coupling mechanism can produce irreversible damage (i.e., if damage accumulates faster than fringes drift, a slowly drifting pattern could still produce sharp boundaries via time-integrated exposure).
• Phase velocity \(v_{ph}\) uncertainty affects fringe spacing and therefore drift tolerance — if \(v_{ph} < c\) (waveguide propagation), fringes are tighter and tolerances are stricter.

8.5 Correlation protocol (upgraded)¶

The minimal falsification protocol (§6) specifies five steps. This section upgrades it with the specific definitions needed for a pre-registered test, as referenced by the reconstruction document.

8.5.1 Match metric¶

The spatial correlation test compares a predicted fringe intensity map to observed damage boundary data. The match metric is:

Distance-to-boundary residual: For each observed damage boundary feature (e.g., WTC 4 knife-edge, WTC 3 bisection strip edge), compute the perpendicular distance from the feature to the nearest predicted fringe node line. The residual for feature \(i\) is:

The aggregate match metric is:

where \(N\) is the number of independent features. Lower \(R\) = better match.

8.5.2 Pre-registered feature set¶

The following features are defined as the independent test targets, each contributing one residual:

Feature	Observable	Source
WTC 4 knife-edge	North-south boundary location (east-west position along building)	FEMA 403, Figure 5-16
WTC 3 bisection strip	Center-line of destruction strip (north-south position along building)	FEMA 403, Chapter 7
WTC 6 borehole center	Center of cylindrical excavation	FEMA 403, Figure 5-19
WTC 5 partial collapse boundary	Edge of collapse region on southern face	FEMA 403, Chapter 6

Exclusions: WTC 1 and WTC 2 are excluded because they are the primary targets (total destruction) and cannot provide boundary-contrast information. Structures beyond the WTC complex boundary are excluded because the fringe model predictions become less constrained at distance.

8.5.3 Success criterion¶

The test is:

The null distribution \(R_{\text{null}}\) is generated by:
1. Computing \(R\) for 10,000 random fringe maps (uniform random phase offset \(\phi_0 \in [0, 2\pi)\) and uniform random frequency within the 2.6–10 MHz band).
2. Computing the percentile of \(R_{\text{observed}}\) in the null distribution.

Success: \(R_{\text{observed}}\) falls below the 5th percentile of the null distribution (\(p < 0.05\)).

8.5.4 Fail-fast criterion¶

If, during the analysis, any of the following conditions are met, the test is stopped and the interference-geometry explanation is rejected:

1. No feature within \(2d\) of any node line: If any of the 4 pre-registered features is more than \(2d\) from the nearest predicted node line (where \(d\) is the fringe spacing at the tested frequency), the prediction has failed for that frequency.
2. Orientation mismatch exceeds 20°: If the predicted fringe orientation deviates from the observed damage boundary orientation by more than 20° for any feature with a clear directional signature, the geometry is rejected.
3. Best frequency requires \(v_{ph} < 0.5c\) or \(v_{ph} > 1.5c\): If phase-velocity tuning beyond this range is needed to fit the data, the propagation assumptions are implausible.

8.5.5 What "success" would and would not prove¶

• Success (\(p < 0.05\)) would mean: the observed damage boundaries are statistically inconsistent with random placement relative to the predicted fringe geometry. This supports the interference-geometry explanation but does not prove it (spatial correlation is necessary but not sufficient for causation).
• Failure (\(p \geq 0.05\)) would mean: the interference-geometry explanation of boundary placement is rejected. The fringe model could still be relevant for other aspects (e.g., total energy delivery) but could not explain the specific locations of damage boundaries.

9. How this connects to the feasibility choke points¶

This module most directly constrains:

• Emitter spec (partial): candidate frequency bands (conditional on geometry + propagation assumptions).
• Control/coherence: if boundaries are sharp, coherence/stability requirements must be consistent with propagation variability.
• Collateral containment: a computed node map must explain why adjacent structures sit in “safe zones” (anti-nodes) rather than being indiscriminately exposed.