Exploiting sparse conditional dependence in remote sensing observing systems
Abstract
We develop a novel filtering algorithm for remote sensing inverse problems that exploits sparsity in the precision (inverse covariance) matrix arising from conditional independence structures in spatio-temporal satellite data. Motivated by NASA's Orbiting Carbon Observatory-2 (OCO-2), Earth Surface Mineral Dust Source Investigation (EMIT), and upcoming Explorer for Artemis Geology Lunar and Earth (EAGLE) missions, we formulate a Bayesian inference framework for retrieving atmospheric and surface parameters from hyperspectral radiance measurements. Unlike traditional methods that operate on covariances and ignore spatio-temporal dependencies, our approach models the full spatio-temporal state using sparse Gaussian graphical models, which corresponds to a block-sparse precision matrix. We derive an information filter based on this structure, enabling scalable joint inference over large spatial grids with uncertainty propagation. We demonstrate the effectiveness of the algorithm in two distinct remote sensing contexts. First, we apply the method to synthetic data configured to mimic the OCO-2 mission, focusing on retrievals of atmospheric CO2, and observe a 50% reduction in CO2 retrieval error compared to single-pixel methods. Then, we evaluate the algorithm using imaging spectroscopy data from the Surface Biology and Geology (SBG) High Frequency Time Series (SHIFT) campaign to retrieve surface reflectance, achieving a 20% reduction in retrieval error relative to baseline approaches.
Related
- Observing hyperbolic systems with source and dynamic boundary disturbances
- Exploiting Pinching Antenna Systems in NOMA Covert Communication Networks
- Regime-Aware Estimation of Large Equity Correlation Matrices: Tail-Pairwise Dependence and Dynamic Conditional Correlation Decomposition for Water and Energy Sectors
- Conditional Fluctuations of Mean-Field Systems with Stable Common Noise
- Admissibility-Preserving Constraint Lifting for Conditional Deadbeat Control of Discrete-Time Nonlinear Systems
BibTeX
@misc{leung2026exploiting,
title = {{Exploiting sparse conditional dependence in remote sensing observing systems}},
author = {Kelvin M. Leung and Jouni Susiluoto and Regina Eckert and Youssef Marzouk},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.24710},
eprint = {2610.24710},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.24710},
}
RIS
TY - UNPB
TI - Exploiting sparse conditional dependence in remote sensing observing systems
AU - Kelvin M. Leung
AU - Jouni Susiluoto
AU - Regina Eckert
AU - Youssef Marzouk
PY - 2026
DA - 2026/10/17
PB - arXiv
DO - 10.48550/arxiv.2610.24710
UR - https://arxiv.org/abs/2610.24710
AB - We develop a novel filtering algorithm for remote sensing inverse problems that exploits sparsity in the precision (inverse covariance) matrix arising from conditional independence structures in spatio-temporal satellite data. Motivated by NASA's Orbiting Carbon Observatory-2 (OCO-2), Earth Surface Mineral Dust Source Investigation (EMIT), and upcoming Explorer for Artemis Geology Lunar and Earth (EAGLE) missions, we formulate a Bayesian inference framework for retrieving atmospheric and surface parameters from hyperspectral radiance measurements. Unlike traditional methods that operate on covariances and ignore spatio-temporal dependencies, our approach models the full spatio-temporal state using sparse Gaussian graphical models, which corresponds to a block-sparse precision matrix. We derive an information filter based on this structure, enabling scalable joint inference over large spatial grids with uncertainty propagation. We demonstrate the effectiveness of the algorithm in two distinct remote sensing contexts. First, we apply the method to synthetic data configured to mimic the OCO-2 mission, focusing on retrievals of atmospheric CO_2, and observe a 50% reduction in CO_2 retrieval error compared to single-pixel methods. Then, we evaluate the algorithm using imaging spectroscopy data from the Surface Biology and Geology (SBG) High Frequency Time Series (SHIFT) campaign to retrieve surface reflectance, achieving a 20% reduction in retrieval error relative to baseline approaches.
ER -