[
  {
    "page": 1,
    "text": "SWSN SCIENTIFIC REVIEW SERIES SWSN-SCI/ORP-2035-04-R3\nNational Center for Scientific Integration for Earth-System Stability and National Continuity (SWSN)\nCross-domain residual coherence and inverse control of\nmonitored quantum trajectories\nA phenomenological framework for macroscopic causal-state\nrestoration\nGovernment scientific review manuscript - Revision 3\nAnan Wongkan1,2,∗, Emalin Wongkan1,2, Taechasit Kengkhuntod1,3\n1Complex Systems and Computational Physics Unit, SWSN, Thailand\n2Computational Physics Research Group, Suranaree University of Technology, Nakhon\nRatchasima, Thailand\n3Applied Mathematics and Control Systems Unit, SWSN, Thailand\n∗Corresponding investigator: A. Wongkan, Principal Scientist\nSubmitted: 17 August 2035 Document code:SWSN-SCI/ORP-2035-04-R3\nFunding: National Continuity Research Fund;Co-funded by CreativeLabTH Group.\nReview status:Not peer reviewed. Submitted for national scientific-resource allocation and\nfalsification testing.\nAbstract. We present a falsifiable two-layer framework motivated by persistent model resid-\nuals observed across heterogeneous monitoring systems and by recent advances in control-\nling the apparent arrow of time in monitored quantum systems. First, we test whether\nnormalized residuals from independent sensor families are better described by indepen-\ndent noise processes or by a shared latent factor with channel-dependent response. In the\nreproducibility stress test reported here, the leading principal component explains 60.1%\nof standardized variance and a one-factor model improves Bayesian information criterion\nby ∆BIC = 176 .5 relative to a diagonal Gaussian null. These statistics do not establish a\nnew spacetime effect; they define a detection target that conventional explanations must\nfirst attempt to absorb. Second, inspired by trajectory-level time-arrow control in moni-\ntored quantum systems, we formulate macroscopic causal-state restoration (MCSR) as an\ninverse-control problem rather than literal reversal of thermodynamic time. A conditional\nphase-noise model shows that access to an imperfect measurement record can raise mean\nstate-restoration fidelity from 0.550 without feedback to 0.854 at effective record efficiency\nη = 0 .85. We then introduce a phenomenological metric-overlap ansatz and derive obser-\nvational tests involving clock, gravimetric, orbital and gravitational-wave residuals. Finally,\nwe quantify the rotational-energy reservoir of a Kerr black hole: a10 M⊙ object with dimen-\nsionless spina∗ = 0 .98 contains an ideal extractable rotational reservoir of approximately\n4.0 × 1047 J. We emphasize that the coupling mechanism proposed here is conjectural. The\nmanuscript’s purpose is to justify a bounded national experiment capable of rejecting the\nhypothesis.\nKeywords: quantum control; arrow of time; inverse problem; complex systems; gravitational-\nwave residuals; Kerr black hole; causal-state restoration; scientific falsification\nCONTROLLED SCIENTIFIC DOCUMENT Internal review copy Page 1 of9"
  },
  {
    "page": 2,
    "text": "SWSN SCIENTIFIC REVIEW SERIES SWSN-SCI/ORP-2035-04-R3\n1 Introduction\nMicroscopic equations can possess time-reversal symmetries while macroscopic observa-\ntions exhibit an arrow of time. In monitored quantum systems, measurement backaction\nitself contributes to an operational arrow that can be quantified from trajectories. García-\nPintos, Liu and Gorshkov recently constructed Hamiltonian control tools capable of replicat-\ning monitored stochastic trajectories and, with feedback, producing dynamics more consis-\ntent with the time-reversed trajectory than the forward one [1]. This result is not a recipe\nfor reversing people, planets or entropy. It is nevertheless important for a different reason:\nit demonstrates that the direction assigned to a monitored trajectory can be reshaped by\ncontrol when the relevant measurement record and dynamics are known.\nOur national monitoring program faces a structurally different inverse problem. Residuals\nremain after accepted models are fitted separately to timing, gravimetric, orbital, geomag-\nnetic and environmental forecasting systems. The conventional interpretation is that these\nresiduals arise from calibration error, unmodelled local forcing, model mismatch and non-\nstationary infrastructure. That interpretation remains the null hypothesis of this manuscript.\nWe ask a narrower question:is there a common latent perturbation whose statistical footprint\nsurvives independent preprocessing pipelines?If so, the next task is not to declare a new uni-\nverse. It is to design observations that could falsify both conventional and nonconventional\nexplanations.\nFor internal work, the inverse-control family is namedORPHEUS. The name is deliberately\ncautionary. In the myth, Orpheus attempts to retrieve what has already been lost. MCSR\nhas the opposite objective: it does not seek to recover the dead or reconstruct a previous\nworld. It seeks, if the model survives testing, to restore a dynamically reachable boundary\ncondition from which forward evolution can continue.\n2 Epistemic status and scope\nThe framework contains three different levels of claim. They must not be conflated.\nLevel Status Use in this manuscript\nE0 Established or standard\nphysics\nOpen-system dynamics, conditional\nquantum trajectories, general rel-\nativity, Kerr rotational energy,\ngravitational-wave parameter esti-\nmation.\nH1 High-risk extrapolation Scaling inverse-control ideas from\nsmall monitored systems to a dis-\ntributed engineered control architec-\nture.\nH2 Phenomenological conjecture Inter-branch metric overlap (IBMO),\nnon-orientable boundary identifica-\ntion and a controllable causal bound-\nary defect.\nTable 1:Claim levels. H2 is not treated as established physics.\nThe central scientific requirement is therefore simple: every H2 claim must imply an obser-\nvation that could fail.\nCONTROLLED SCIENTIFIC DOCUMENT Internal review copy Page 2 of9"
  },
  {
    "page": 3,
    "text": "SWSN SCIENTIFIC REVIEW SERIES SWSN-SCI/ORP-2035-04-R3\n3 Cross-domain residual coherence\n3.1 Residual representation\nFor monitoring channelk, letyk(t) denote the observed quantity andˆyk(t|Mk) the prediction\nfrom the accepted domain modelMk. We define a standardized residual\nrk(t) = yk(t) − ˆyk(t|Mk)\nˆσk(t) . (1)\nThe working shared-factor model is\nrk(t) = akz(t − τk) + ϵk(t), (2)\nwhere z(t) is a latent process,ak is channel response,τk is a small lag andϵk is channel-specific\nnoise. This equation is intentionally agnostic about physical cause.\n3.2 Reproducibility stress test\nTo verify the analysis pipeline before restricted data are reprocessed, we generated a 108-\nmonth five-channel stress-test set containing an autoregressive latent process, three broad\ntransients, channel-specific lags and independent Gaussian noise. The code and exact seed\nare preserved in the technical supplement. Figure1 shows the standardized series. The\npoint of this exercise is computational reproducibility, not empirical proof.\nFigure 1:Normalized cross-domain residual stress-test series. Curves are vertically offset for visibil-\nity. Dashed lower trace is the latent reference used only for validation.\nA principal-component analysis yields a first-component explained variance fraction of 0.6005.\nA one-factor Gaussian model yields BIC= 1379 .4, compared with1555.9 for a diagonal Gaus-\nsian null, giving∆BIC = 176 .5 in favor of shared covariance in this controlled stress test.\nFigure 2 shows the residual correlation structure.\nCONTROLLED SCIENTIFIC DOCUMENT Internal review copy Page 3 of9"
  },
  {
    "page": 4,
    "text": "SWSN SCIENTIFIC REVIEW SERIES SWSN-SCI/ORP-2035-04-R3\nFigure 2:Correlation matrix of the reproducibility stress-test residuals. Sign differences are permit-\nted because different sensors may respond with opposite convention or transfer function.\nFor restricted operational data, the same pipeline must be challenged against known com-\nmon causes: shared time references, synchronized maintenance windows, common envi-\nronmental forcing, data-assimilation leakage and preprocessing artifacts. Failure to remove\nthese effects rejects the need for H2.\n4 Phenomenological inter-branch metric-overlap model\nWe introduce the metric ansatz\ngeff\nµν (x) = g(0)\nµν (x) + ε χ(x) h(B)\nµν (x), (3)\nwhere g(0)\nµν is the background metric,h(B)\nµν is an effective perturbation associated with a sec-\nond boundary branch,χ(x) is a compact overlap envelope andε is a dimensionless coupling\nparameter. No claim is made thath(B)\nµν corresponds to a physically realized parallel universe.\nThe ansatz is a bookkeeping device for deriving correlated signatures.\nThe strongest version of H2 further assumes that the overlap boundary is non-orientable\nin an effective higher-dimensional embedding. A Möbius strip is a two-dimensional anal-\nogy: following a locally continuous path can return to a position with reversed orientation.\nWe use this only as a topology metaphor. The physical model would require a well-defined\nmanifold, stress-energy conditions and causal structure that are not supplied by the analogy.\n4.1 Predicted observational signatures\nThe hypothesis is useful only if it predicts measurable inconsistencies. We therefore require\nat least three independently instrumented signatures:\n1. a coherent timing residual not removable by reference-clock or propagation corrections;\nCONTROLLED SCIENTIFIC DOCUMENT Internal review copy Page 4 of9"
  },
  {
    "page": 5,
    "text": "SWSN SCIENTIFIC REVIEW SERIES SWSN-SCI/ORP-2035-04-R3\n2. a correlated gravimetric or orbital residual with the predicted sign and lag structure;\n3. an astrophysical propagation anomaly whose detector-to-detector timing is consistent\nwith the same sky region but inconsistent with the best-fit general-relativistic source\nwaveform.\nGravitational-wave signals encode compact-object masses, spins, sky location and distance\nwhen compared with waveform models [7]. We therefore propose residual analysis after\nstandard parameter estimation, rather than treating a black hole as a continuously broad-\ncasting beacon.\nA minimal delayed-path phenomenology is\n˜hobs(f) = ˜hGR(f)\n[\n1 + αW (f)e2πif ∆t\n]\n+ ˜n(f), (4)\nwhere α, ∆t and W (f) are to be constrained by data. A statistically insignificantα across the\npredicted band counts against H2.\n5 From a quantum arrow of time to inverse control\nFor a Markovian open quantum system, a standard Lindblad description is\n˙ρ = − i\n¯h[H(t), ρ] +\n∑\nj\n(\nCjρC†\nj − 1\n2{C†\nj Cj, ρ}\n)\n, (5)\nwith density operatorρ, HamiltonianH and collapse operatorsCj. Numerical toolkits such\nas QuTiP implement these equations for open-system simulation [5, 6].\nThe 2026 result of García-Pintos et al. is trajectory-level and measurement-conditioned [1].\nOur extrapolation preserves that conceptual restriction. MCSR is formulated as an inverse-\ncontrol objective\nJ[u] = 1 − F\n(\nρu(T ), ρtarget\n)\n+ λ\n∫ T\n0\n∥u(t)∥2dt + µ R[u], (6)\nwhere F is state fidelity,u(t) is a control field andR penalizes unsafe or nonrobust control.\nThis is not a global inverse of entropy production. It is a search for a reachable state under\nconstrained dynamics.\n5.1 Conditional phase-noise toy model\nWe tested the simplest transparent case: an initial|+x⟩ qubit accumulates a stochastic phase\nϕ. Without access to the noise record, ensemble averaging dephases the state. With a con-\nditional estimate of the phase, feedback removes part of the accumulated rotation. Across\n60,000 trajectories, the uncontrolled mean fidelity is 0.550. At effective record efficiency\nη = 0 .85 and unit feedback gain, mean fidelity is 0.854. Perfect record access in this toy\nmodel returns the state exactly. Figure3 shows the full gain-efficiency surface.\nThe operational lesson is modest but important: trajectory information changes what control\nis possible. If a macroscopic restoration protocol exists at all, it must preserve a sufficiently\nrich measurement history rather than infer the past from a final state alone.\nCONTROLLED SCIENTIFIC DOCUMENT Internal review copy Page 5 of9"
  },
  {
    "page": 6,
    "text": "SWSN SCIENTIFIC REVIEW SERIES SWSN-SCI/ORP-2035-04-R3\nFigure 3: Mean restoration fidelity in the conditional phase-noise stress test. This is an illustrative\ncontrol model, not a reproduction of Ref. [1] and not evidence for macroscopic time reversal.\n6 Why a quantum computer is requested\nExact classical state-vector storage scales as2N . At complex128 precision, a state vector alone\nrequires approximately16 × 2N bytes: about 18 PB forN = 50 and 18 EB forN = 60 , before\noperators, trajectories, optimization states and error-correction overhead. Tensor-network\nand reduced-order methods can drastically reduce cost when structure permits; therefore\nexponential storage is not, by itself, proof that a quantum computer will provide advantage.\nThe resource request is narrower. We request controlled access to the national quantum sys-\ntem to test whether the ORPHEUS inverse-control subproblem can be encoded as a tractable\nquantum simulation or hybrid optimization task on a restricted model class. The success cri-\nterion is a measurable reduction in wall-clock or sample complexity against the best avail-\nable classical baseline. If no advantage appears, the quantum-compute branch of the pro-\ngram is to be closed.\n7 Astrophysical energy reservoir\nA rotating Kerr black hole contains extractable rotational energy. The irreducible-mass re-\nlation gives\nMirr\nM =\n√\n1 +\n√\n1 − a2∗\n2 , (7)\nso the ideal rotational reservoir is\nErot = ( M − Mirr) c2. (8)\nFor a∗ → 1, the ideal reservoir approaches 29.3% ofMc2. Penrose-process and related mecha-\nnisms show that energy extraction from a rotating black hole is permitted by general relativ-\nity in principle, although practical engineering at astrophysical scales is entirely unresolved\n[8, 9, 10].\nCONTROLLED SCIENTIFIC DOCUMENT Internal review copy Page 6 of9"
  },
  {
    "page": 7,
    "text": "SWSN SCIENTIFIC REVIEW SERIES SWSN-SCI/ORP-2035-04-R3\nFor a10M⊙ Kerr object, the ideal reservoir is2.73 × 1047 J ata∗ = 0 .90, 4.03 × 1047 J ata∗ = 0 .98\nand 4.96 × 1047 J ata∗ = 0 .999. Figure4 shows the spin dependence.\nFigure 4:Ideal rotational-energy fraction of a Kerr black hole. This is a reservoir calculation, not an\nengineering efficiency claim.\nWe therefore separate two questions: (i) whether sufficient ordered energy exists in princi-\nple, and (ii) whether a physically allowed coupling can transfer useful energy to the proposed\nboundary-control degree of freedom. Only the first question is established by the calculation\nabove.\n8 National falsification program\nWe propose a staged program with hard stop conditions.\nStage A - residual audit.Reprocess timing, gravimetric, orbital and geomagnetic residu-\nals using independent teams and blinded preprocessing. Reject H2 if common covariance\ndisappears after shared infrastructure and reference frames are removed.\nStage B - gravitational-wave residual search.Pre-register a delayed-path residual family\nand search only after standard general-relativistic parameter estimation. Reject the relevant\nH2 parameter region whenα is consistent with zero at the pre-registered sensitivity.\nStage C - quantum inverse-control benchmark.Implement the ORPHEUS toy and re-\nduced models on classical and quantum backends. Reject the resource case if the quantum\nimplementation does not outperform the best classical method on a task size where both are\nvalid.\nStage D - geometric consistency.If A-C survive, test whether one set of overlap parameters\npredicts all independent residual families. A model that needs separate free parameters for\neach sensor family is to be treated as curve fitting, not a new physical explanation.\n9 Risks, alternative explanations and limitations\nThe strongest risk is model overreach. Cross-domain correlation can be produced by com-\nmon clocks, shared data infrastructure, assimilation pipelines, solar activity, maintenance\nCONTROLLED SCIENTIFIC DOCUMENT Internal review copy Page 7 of9"
  },
  {
    "page": 8,
    "text": "SWSN SCIENTIFIC REVIEW SERIES SWSN-SCI/ORP-2035-04-R3\nschedules and human responses to disasters. A one-factor fit is therefore a diagnostic, not a\ndiscovery.\nThe second risk is category error. Reversing the statistical arrow of a monitored quantum\ntrajectory does not imply reversal of thermodynamic, biological or cosmological time. Our\nuse of the 2026 quantum-control result is motivational and algorithmic.\nThe third risk is energy rhetoric. A Kerr black hole’s rotational reservoir can be enormous\nwhile a useful extraction-and-coupling architecture remains impossible. The energy calcu-\nlation cannot rescue an invalid control mechanism.\nThe fourth risk is topology without dynamics. A Möbius-like non-orientable identification\nmay be mathematically describable yet physically inadmissible. The H2 branch must even-\ntually specify a metric, stress-energy source, stability conditions and causal structure or be\nabandoned.\n10 Conclusion\nThis manuscript does not claim that two universes are colliding. It identifies a sequence of\nincreasingly difficult tests that would have to fail before such a hypothesis deserves further\nattention. The reproducibility pipeline detects shared residual structure when it is present.\nConditional quantum control can, in a toy monitored system, recover trajectories that would\notherwise appear irreversible at the ensemble level. Kerr rotation provides a physically real\nbut technologically inaccessible energy reservoir. The conjectural step is the existence of a\nboundary degree of freedom that connects these facts.\nThe immediate request is therefore not authorization to manipulate spacetime. It is autho-\nrization to perform the cheapest experiment capable of proving us wrong.\nDeclarations\nFunding. This work was supported by the National Continuity Research Fund under SWSN\nprogram ORPHEUS.Co-funding was provided by CreativeLabTH Groupfor reproducible\nscientific-computing infrastructure and independent simulation verification. Funders had\nno authority to alter scientific conclusions.\nConflict of interest. The authors declare no financial conflict of interest relevant to the\nscientific claims in this manuscript.\nData and code availability.The reproducibility stress-test generator, fixed random seed,\nresidual table and Kerr-energy calculation are archived under package SWSN-SCI/ORP-2035-\n04-SUPP. Restricted operational sensor data are excluded from the public review copy.\nAuthor contributions.A.W. conceived the phenomenological framework and inverse prob-\nlem. E.W. designed the cross-domain statistical falsification strategy and reproducibility\npipeline. T.K. implemented control-system formulations, numerical benchmarks and robust-\nness tests. All authors reviewed the manuscript.\nA Reproducibility parameters\nThe cross-domain stress test contains 108 monthly samples and five standardized channels.\nA fixed pseudorandom seed of 20350817 is used. The latent process is an AR(1) sequence with\ncoefficient 0.86, three broad Gaussian transients and a weak sinusoidal component. Channel\nresponses use different signs, lags and independent Gaussian noise. The quantum-control\nCONTROLLED SCIENTIFIC DOCUMENT Internal review copy Page 8 of9"
  },
  {
    "page": 9,
    "text": "SWSN SCIENTIFIC REVIEW SERIES SWSN-SCI/ORP-2035-04-R3\nstress test draws 60,000 phase-noise trajectories with phase standard deviation 2.15 rad and\nevaluates fidelityF = (1+ cos ∆ϕ)/2 after efficiency-dependent feedback. The full executable\nscript is part of the technical supplement.\nReferences\n[1] García-Pintos LP, Liu YK, Gorshkov AV (2026) Reshaping the Quantum Arrow of Time. Phys Rev\nX 16:011028. https://doi.org/10.1103/l18s-9vmh\n[2] Dressel J, Chantasri A, Jordan AN, Korotkov AN (2017) Arrow of time for continuous quantum\nmeasurement. Phys Rev Lett 119:220507.\n[3] Harrington PM, Tan D, Naghiloo M, Murch KW (2019) Characterizing a statistical arrow of time\nin quantum measurement dynamics. Phys Rev Lett 123:020502.\n[4] Wiseman HM, Milburn GJ (2009) Quantum Measurement and Control. Cambridge University\nPress, Cambridge.\n[5] Johansson JR, Nation PD, Nori F (2012) QuTiP: An open-source Python framework for the dynam-\nics of open quantum systems. Comput Phys Commun 183:1760-1772.\n[6] Johansson JR, Nation PD, Nori F (2013) QuTiP 2: A Python framework for the dynamics of open\nquantum systems. Comput Phys Commun 184:1234-1240.\n[7] Aasi J et al. (2013) Parameter estimation for compact binary coalescence signals with the first\ngeneration gravitational-wave detector network. Phys Rev D 88:062001.\n[8] Penrose R (1969) Gravitational collapse: the role of general relativity. Riv Nuovo Cimento 1:252-\n276.\n[9] Christodoulou D (1970) Reversible and irreversible transformations in black-hole physics. Phys\nRev Lett 25:1596-1597.\n[10] Maeda K, Okabayashi K, Okawa H (2018) Maximal efficiency of the collisional Penrose process\nwith spinning particles. Phys Rev D 98:064027.\n[11] Maldacena J, Susskind L (2013) Cool horizons for entangled black holes. Fortschr Phys 61:781-\n811.\n[12] Jacobs K, Steck DA (2006) A straightforward introduction to continuous quantum measurement.\nContemp Phys 47:279-303.\nCONTROLLED SCIENTIFIC DOCUMENT Internal review copy Page 9 of9"
  }
]