A deep learning method amortizes probabilistic XCO2 retrieval from OCO-2 spectra via Laplace approximations and normalizing flows, trained on simulations with model errors to achieve faster inference and better-calibrated uncertainties than operational solvers.
Title resolution pending
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 4roles
background 1polarities
background 1representative citing papers
Introduces a scalable Bayesian inference framework for nonlinear conservation laws using Gaussian process priors and sparse approximations, enabling accurate forward simulations with UQ and fast posterior recovery on inverse problems.
A local L-order neighborhood approximation reduces Kriging prediction complexity from O(N n^3) to O(N log N + nM + M^3) for stationary Gaussian processes on regular grids while keeping pointwise errors around 10^{-5}.
An autoregressive Gaussian process transport-map construction factors spatio-temporal joint densities into conditional distributions with data-dependent sparsity to enable scalable generative modeling of non-Gaussian fields.
citing papers explorer
-
Amortized Probabilistic Retrieval of Atmospheric CO2 from OCO-2 Spectra Using Deep Learning with Laplace Approximations and Normalizing Flows
A deep learning method amortizes probabilistic XCO2 retrieval from OCO-2 spectra via Laplace approximations and normalizing flows, trained on simulations with model errors to achieve faster inference and better-calibrated uncertainties than operational solvers.
-
Scalable Bayesian Inference for Nonlinear Conservation Laws
Introduces a scalable Bayesian inference framework for nonlinear conservation laws using Gaussian process priors and sparse approximations, enabling accurate forward simulations with UQ and fast posterior recovery on inverse problems.
-
Rapid Approximation Prediction for Kriging
A local L-order neighborhood approximation reduces Kriging prediction complexity from O(N n^3) to O(N log N + nM + M^3) for stationary Gaussian processes on regular grids while keeping pointwise errors around 10^{-5}.
-
Scalable generative modeling of non-Gaussian spatio-temporal fields via autoregressive Gaussian processes
An autoregressive Gaussian process transport-map construction factors spatio-temporal joint densities into conditional distributions with data-dependent sparsity to enable scalable generative modeling of non-Gaussian fields.