Temporal Fourier Likelihoods with Spatial Hilbert-Space Gaussian Process Approximations
Abstract
Reconstructing stationary space-time Gaussian processes at unobserved locations is costly when many sites share regular temporal records. We develop a spectral likelihood combining a temporal discrete Fourier transform (DFT) with a Hilbert-space Gaussian process (HSGP) representation of frequency-specific spatial covariance. We derive the exact covariance of the finite-record DFT coefficients and use a Whittle likelihood that approximates distinct frequencies as independent spatial problems. At each temporal frequency, HSGP approximates spatial covariance by evaluating the sampled spectral multiplier at retained Laplacian eigenfrequencies. For models specified by a joint spectral density whose half-spectrum lacks a convenient closed form, this construction avoids repeated Fourier inversion. The fixed spatial basis also permits cached feature projections to be reused in likelihood fitting and held-site reconstruction, with approximation accuracy depending on domain extension and basis size. In a simulation study of such a model, HSGP achieved reconstruction accuracy comparable to a high-accuracy quadrature reference while reducing mean fitting time by 69%. Additional applications to wind-field reconstruction tasks examine the effects of basis rank, temporal record length, and separability, and demonstrate accurate inference on held-out test sites when sufficiently rich bases are used. Taken together, the results indicate that computational savings are attainable when the spectral multiplier can be evaluated directly, numerical spatial inversion is costly, and an adequate basis has rank lower than the number of fitting sites.
Related
- A nonlocal nonlinear Schrödinger model: well-posedness, local limit, and structure-preserving asymptotically compatible Fourier approximations
- Para-Gaussian Approximations of Random Deep Neural Networks
- Joint Source-Channel Coding of Gaussian Sources over Block Erasure Channels: Nonasymptotic Bounds and Channel-Uniform Normal Approximations
- Beyond Laplace: Closed-form wrapped Gaussian posterior approximations on statistical manifolds
- Higher-Order Gaussian Approximations on Path Space for the Simple Symmetric Exclusion Process
BibTeX
@misc{huang2026temporal,
title = {{Temporal Fourier Likelihoods with Spatial Hilbert-Space Gaussian Process Approximations}},
author = {Xin Huang and Jia Li and Jun Yu},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.13441},
eprint = {2609.13441},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.13441},
note = {CC BY-NC-SA 4.0},
}
RIS
TY - UNPB
TI - Temporal Fourier Likelihoods with Spatial Hilbert-Space Gaussian Process Approximations
AU - Xin Huang
AU - Jia Li
AU - Jun Yu
PY - 2026
DA - 2026/09/11
PB - arXiv
DO - 10.48550/arxiv.2609.13441
UR - https://arxiv.org/abs/2609.13441
AB - Reconstructing stationary space-time Gaussian processes at unobserved locations is costly when many sites share regular temporal records. We develop a spectral likelihood combining a temporal discrete Fourier transform (DFT) with a Hilbert-space Gaussian process (HSGP) representation of frequency-specific spatial covariance. We derive the exact covariance of the finite-record DFT coefficients and use a Whittle likelihood that approximates distinct frequencies as independent spatial problems. At each temporal frequency, HSGP approximates spatial covariance by evaluating the sampled spectral multiplier at retained Laplacian eigenfrequencies. For models specified by a joint spectral density whose half-spectrum lacks a convenient closed form, this construction avoids repeated Fourier inversion. The fixed spatial basis also permits cached feature projections to be reused in likelihood fitting and held-site reconstruction, with approximation accuracy depending on domain extension and basis size. In a simulation study of such a model, HSGP achieved reconstruction accuracy comparable to a high-accuracy quadrature reference while reducing mean fitting time by 69%. Additional applications to wind-field reconstruction tasks examine the effects of basis rank, temporal record length, and separability, and demonstrate accurate inference on held-out test sites when sufficiently rich bases are used. Taken together, the results indicate that computational savings are attainable when the spectral multiplier can be evaluated directly, numerical spatial inversion is costly, and an adequate basis has rank lower than the number of fitting sites.
ER -