arXiv:2609.23057
A Quantum-Inspired Two-Dimensional Lattice Polarized Radiative Transfer Simulator
Shan Zeng, Bing Lin, Ali Omar
Abstract
A quantum inspired framework was developed for polarized radiative transfer in the atmosphere to jointly model multiple scattering, absorption, and emission processes. The radiative field is represented by the Stokes vector and discretized using an n-directional two-dimensional spatial lattice on 3-dimensional uniformly distributed zenith and azimuth angular vectors. The formulation of this quantum computation system is based on the Lattice Boltzmann Method, which incorporates a physics-based polarized phase matrix. A classical lattice solver employing identical spatial and angular discretization and physical modeling as in quantum algorithm is implemented. The corresponding quantum algorithm is evaluated using the IBM Qiskit quantum simulator, enabling direct comparison of numerical accuracy to a classical simulation reference. The results demonstrate that fully physics-based polarized radiative transfer can be embedded within a quantum lattice framework, which provides an initial benchmark toward scalable quantum atmospheric radiative transfer and future quantum climate modeling
AI Summary
The paper investigates how polarized atmospheric radiative transfer—including scattering, absorption, and emission—can be represented in quantum circuits using a lattice-Boltzmann-style discretization. It extends a scalar one-dimensional lattice formulation to two spatial dimensions, with separately discretized three-dimensional propagation directions and the full Stokes vector $(I,Q,U,V)$ amplitude-encoded using a two-qubit polarization register. The construction combines direction-dependent Rayleigh Mueller matrices and polarization-frame rotations with ancilla-assisted unitary dilations and linear combinations of unitaries. Emission injects a persistent source, absorption implements Beer–Lambert attenuation, and fractional spatial streaming is represented by weighted integer lattice shifts reproducing bilinear semi-Lagrangian interpolation. Both periodic and absorbing boundaries are considered. Each timestep includes classical decoding and subsequent re-encoding. Validation uses noiseless Qiskit Aer statevector simulations against a classical solver with matching discretization and physical assumptions. The principal two-dimensional examples use a $16\times16$ lattice, eight angular directions, and 29 qubits. Reported classical–quantum differences are typically around $10^{-12}$ for intensity and degree of polarization across 64 timesteps; smaller angular-grid tests report absolute intensity differences around $10^{-13}$–$10^{-14}$. A reduced one-dimensional configuration is also compared with an earlier benchmark. The stated limitations include simulator-memory constraints, coarse angular sampling and associated interpolation artifacts, no hardware implementation, and omission of dichroic absorption. Spectral molecular absorption, aerosol/cloud scattering, and fully three-dimensional spatial transport remain future extensions. Fully coherent multistep evolution would require additional workspace; the proposed accounting gives 216 qubits for 64 steps. No runtime advantage is established.
Central claim may not hold: The polarization-frame formulas in Equations (13)–(15) spuriously rotate linear polarization during forward Rayleigh scattering at general azimuths, contradicting the claimed physical fidelity unless the implementation uses an undocumented correction. The classical–quantum comparisons use the same physical model and discretization, so their numerical agreement cannot independently establish that the polarized transport physics is correct. The per-step decoding and re-encoding scheme lacks a measurement procedure and resource accounting for recovering signed Stokes amplitudes, leaving practical accuracy and scalability unestablished. The decoding section writes the post-selection success probability as an L2 norm rather than its square, despite correctly describing a sum of squared magnitudes. The stated approximately 4 GB statevector memory requirement is inconsistent with dense double-precision complex storage, which requires 8 GiB for 29 qubits or 16 GiB for 30 qubits.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.