Quantum Bayesian Inference for Atmospheric Biosignature and Technosignature Characterization in Exoplanetary Systems

Master's Thesis in Quantum Computing (Academic Year 2025/2026)
Escuela Politécnica Superior — Universidad Antonio de Nebrija

     

Author: Daniel Rivero Losa  |  Supervisor: Roberto Campos Ortiz
Benchmark Target: Exoplanet K2-18b (JWST Transmission Spectroscopy)

---

🌌 Executive Abstract

Next-generation space observatories—chiefly the James Webb Space Telescope (JWST) and the future Habitable Worlds Observatory (HWO)—produce rich, coupled transmission telemetry of exoplanetary atmospheres. Evaluating the causal likelihood of candidate biosignatures (e.g., DMS) and ultra-rare technosignatures (e.g., CFCs at $p \approx 10^{-6}$) using classical probabilistic graphical models encounters insurmountable computational bottlenecks:

1. Exact Inference: Clique trees (Junction Tree) suffer an exponential memory ceiling $\mathcal{O}(N \cdot d^{w+1})$ driven by high network treewidth $w$.
2. Stochastic Sampling: Markov Chain Monte Carlo (MCMC) and Rejection Sampling are constrained to $\mathcal{O}(1/\sqrt{M})$, succumbing to L'Ecuyer's variance divergence ($\mathrm{RE} \to \infty$) and Kac's recurrence trap ($\mathbb{E}[\tau] \approx 10^6$ steps).

This work establishes an end-to-end Quantum Bayesian Network (QBN) architecture coupled with Quantum Amplitude Estimation (QAE) in Qiskit to overcome both boundaries with bounded in-degree state preparation $\mathcal{O}(N \cdot 2^{k_{\max}})$:

Figure: Master 6-Stage Quantum Bayesian Inference Architecture — from transit observation and JWST spectral telemetry to NISQ Zero-Noise Extrapolation and final astrobiological hypothesis resolution.

---

⚖️ Tripartite Benchmark: Classical vs. FTQC vs. NISQ

Feature / Metric	Classical Baseline (pgmpy / MCMC)	Canonical QAE (Fault-Tolerant Horizon)	Error-Mitigated NISQ (Physical Viability)
Spatial Scaling	$\mathcal{O}(N \cdot d^{w+1})$ RAM explosion in clique potentials.	$\mathcal{O}(N)$ Qubits: 9 qubits encode all $2^9 = 512$ states.	5 physical qubits encode core inference kernel.
Query Scaling	$\mathcal{O}(1/\sqrt{M})$ Monte Carlo rate; diverges on rare states.	$\mathcal{O}(1/M)$ Heisenberg speedup via Grover rotation.	$\mathcal{O}(1)$ queries with noise-scaled folding factors.
Empirical Results ($p \sim 10^{-6}$)	• MCMC: $0$ hits in $100{,}000$ steps (false negative). • Rejection: $90.42%$ discarded samples under evidence.	• Phase Quantization: $100.0%$ shots at 10000 ($y=16$). • Centroid: Sub-resolution phase $\theta_a \ll \Delta\theta$ projects onto central basis state.	• Raw Noise: $+137.21%$ distortion ($6.30% \to 14.94%$). • ZNE Mitigated: $7.47%$ recovered ($86.44%$ error cancelled).
Physical Hardware	Exact matrix math on CPU / RAM.	Master circuit: $52{,}393$ depth, $53{,}456$ gates (FTQC horizon).	Kernel: Depth $D = 39$, $57$ native gates ($23$ CNOTs at $\lambda=1$).
Reference Section	Chapter 2 & Notebook 01	Chapter 3 & Notebook 02	Chapter 4 & Notebook 03

---

🧭 Project Navigation & Deep Dives

Explore dedicated components across the repository:

Section	Description	Key Deliverables
📓 Interactive Notebooks	Executed Jupyter pipelines for classical stress tests, ideal QAE, and ZNE.	01_Classical_Limits • 02_QAE_Ideal • 03_NISQ_ZNE
🖼️ Pipeline Architecture	Master 6-step animated quantum Bayesian pipeline, interactive web dashboard, and vector assets.	croquis_keynote_2row.svg • interactive_overview.html • croquis_keynote_2row_static.svg
📊 Data & Priors	Structured exoplanetary priors, JWST transmission spectra, and benchmark logs.	k218b_cpt_priors.json • transmission_spectrum.csv • data/README.md
📚 Scientific Library	Archival repository containing 21 peer-reviewed open-access PDFs with DOIs.	Quantum Algorithms • Error Mitigation • Classical Complexity • papers/README.md
📑 Thesis LaTeX Source	Full academic monograph source code for print, digital, and Overleaf editions.	Crown Quarto (main.tex) • Digital A4 (main_digital.tex) • Overleaf (main_overleaf.tex)
🎯 Conclusions & Horizons	Comprehensive synthesis, NISQ-to-FTQC roadmap, and HWO telescope prospects.	Iterative QAE (IQAE without QFT) • PEC mitigation • Multi-planetary scaling
📋 Appendix A (CPTs)	Exhaustive combinatorial CPTs ($2^k$ rows), analytical marginals, and derivations.	Prior $P(X_3) = 0.0763$, $P(X_8) = 10^{-6}$, JWST posterior $P(\mathbf{e}) = 0.095804$
🛠️ Appendix B (Transpilation)	100% verified Qiskit transpiler metrics, gate inventories, and CNOT scaling.	17Q Master Profile ($53{,}456$ ops) • ZNE CNOT linearity ($23 \to 69 \to 115$) • $n_E \in [3, 10]$ scaling

---

🚀 Quickstart & Reproduction

# 1. Clone the repository and set up environment
git clone https://github.com/danielriverolosa/quantum-bayesian-inference-astrobiology.git
cd quantum-bayesian-inference-astrobiology
python3.11 -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt

# 2. Run classical baseline (Junction Tree & MCMC)
python src/chapter_2_classical/build_and_run_01_classical.py

# 3. Run quantum simulations (Ideal QAE & NISQ ZNE)
python src/chapter_4_quantum/build_and_run_02_qae_ideal.py
python src/chapter_4_quantum/build_and_run_03_nisq_zne.py

---

📜 Academic Citation

@mastersthesis{riverolosa2026quantum,
  author       = {Daniel Rivero Losa},
  title        = {Quantum Bayesian Inference for Atmospheric Biosignature and Technosignature Characterization in Exoplanetary Systems},
  school       = {Escuela Politécnica Superior, Universidad Antonio de Nebrija},
  year         = {2026},
  month        = {February},
  type         = {Master's Thesis},
  address      = {Madrid, Spain},
  note         = {Supervisor: Roberto Campos Ortiz},
  url          = {https://github.com/danielriverolosa/quantum-bayesian-inference-astrobiology}
}

---

⚖️ License

Code released under the GNU General Public License v3.0 (GPLv3). Thesis manuscript and figures are published under academic research fair use.