Learning Structures in Earth Observation Data with Gaussian Processes

TL;DR

This paper reviews recent advances in Gaussian Processes for geoscience, highlighting new algorithms that improve modeling of earth observation data with respect to signal, noise, and uncertainty quantification.

Contribution

It provides a comprehensive review of recent theoretical and algorithmic developments in Gaussian Processes applied to geoscience and remote sensing.

Findings

New algorithms respect signal and noise characteristics

Automatic feature ranking methods introduced

Uncertainty intervals enable space-time GP applications

Abstract

Gaussian Processes (GPs) has experienced tremendous success in geoscience in general and for bio-geophysical parameter retrieval in the last years. GPs constitute a solid Bayesian framework to formulate many function approximation problems consistently. This paper reviews the main theoretical GP developments in the field. We review new algorithms that respect the signal and noise characteristics, that provide feature rankings automatically, and that allow applicability of associated uncertainty intervals to transport GP models in space and time. All these developments are illustrated in the field of geoscience and remote sensing at a local and global scales through a set of illustrative examples.