Inference over radiative transfer models using variational and expectation maximization methods

DOI10.1007/S10994-021-05999-4MaRDI QIDQ6106441zbMATH OpenOpenAlexWikidataFDO

Authors Daniel Heestermans Svendsen, Daniel Hernández-Lobato, Luca Martino, Valero Laparra, Álvaro Moreno-Martínez, Gustau Camps-Valls

Publication date 27 June 2023

Published in Machine Learning (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2204.03346

zbMATH Keywords

density estimationexpectation maximizationinverse modelingradiative transfer modelvariational autoencoder

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05)

Abstract: Earth observation from satellites offers the possibility to monitor our planet with unprecedented accuracy. Radiative transfer models (RTMs) encode the energy transfer through the atmosphere, and are used to model and understand the Earth system, as well as to estimate the parameters that describe the status of the Earth from satellite observations by inverse modeling. However, performing inference over such simulators is a challenging problem. RTMs are nonlinear, non-differentiable and computationally costly codes, which adds a high level of difficulty in inference. In this paper, we introduce two computational techniques to infer not only point estimates of biophysical parameters but also their joint distribution. One of them is based on a variational autoencoder approach and the second one is based on a Monte Carlo Expectation Maximization (MCEM) scheme. We compare and discuss benefits and drawbacks of each approach. We also provide numerical comparisons in synthetic simulations and the real PROSAIL model, a popular RTM that combines land vegetation leaf and canopy modeling. We analyze the performance of the two approaches for modeling and inferring the distribution of three key biophysical parameters for quantifying the terrestrial biosphere.

Recommendations

• scientific article; zbMATH DE number 5584247
• Variational Bayesian Inference of Line Spectra
• Optimal prediction for radiative transfer: a new perspective on moment closure
• Markov chain Monte Carlo solutions for radiative transfer problems
• Approximate models for radiative transfer
• Machine Learning Approaches for the Inversion of the Radiative Transfer Equation
• Distributed parameter estimation for the time-dependent radiative transfer equation
• A posteriori error analysis in radiative transfer
• Numerical methods for parameter identification in stationary radiative transfer

Cites work

• A Stochastic Approximation Method
• scientific article; zbMATH DE number 3567782 (Why is no real title available?)
• scientific article; zbMATH DE number 1557074 (Why is no real title available?)
• Inverse Problem Theory and Methods for Model Parameter Estimation
• Layered adaptive importance sampling
• Machine learning of linear differential equations using Gaussian processes
• MCMC using Hamiltonian dynamics
• On convergence properties of the Monte Carlo EM algorithm
• The design and analysis of computer experiments.

Cited in

(3)

• Machine learning moment closure models for the radiative transfer equation. I: Directly learning a gradient based closure
• Machine Learning Approaches for the Inversion of the Radiative Transfer Equation
• scientific article; zbMATH DE number 5584247 (Why is no real title available?)

This page was built for publication: Inference over radiative transfer models using variational and expectation maximization methods

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6106441)