Inference over radiative transfer models using variational and expectation maximization methods
Entity
UAM. Departamento de Ingeniería InformáticaPublisher
SpringerDate
2021-06-08Citation
10.1007/s10994-021-05999-4
Machine Learning 112 (2023): 921-937
ISSN
0885-6125DOI
10.1007/s10994-021-05999-4Funded by
Research partly funded by the ERC under the ERC-CoG-2014 project (grant agreement 647423) and the ERC-SyG-2019 USMILE project (grant agreement 855187), the Ministerio de Ciencia e Innovación project PID2019-106827GB-I00/AEI/10.13039/501100011033, and the projects TEC2016-77741-R, DPI2017-89867-C2-2-R, RTI2018-096765-A-I00, PID2019- 109026RB-I00 and PID2019-105032GB-I00Project
info:eu-repo/grantAgreement/EC/ERC/647423; info:eu-repo/grantAgreement/EC/ERC/855187; Gobierno de España. PID2019-106827GB-I00; Gobierno de España. TEC2016-77741-R; Gobierno de España. DPI2017-89867-C2-2-R; Gobierno de España. RTI2018-096765-A-I00; Gobierno de España. PID2019- 109026RB-I00; Gobierno de España. PID2019-105032GB-I00Editor's Version
https://doi.org/10.1007/s10994-021-05999-4Subjects
Density estimation; Expectation maximization; Inverse modeling; Radiative transfer model; Variational autoencoder; InformáticaRights
© The Author(s), under exclusive licence to Springer Science+Business Media LLC, part of Springer Nature 2021Abstract
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
Files in this item
Google Scholar:Svendsen, Daniel Heestermans
-
Hernández Lobato, Daniel
-
Martino, Luca
-
Laparra, Valero
-
Moreno Martínez, Álvaro
-
Camps Valls, Gustau
This item appears in the following Collection(s)
Related items
Showing items related by title, author, creator and subject.