Share Email Print

Proceedings Paper

HT-FRTC: a fast radiative transfer code using Gaussian processes
Author(s): Stephan Havemann; Gerald Wong; Warren Lewis
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

The Havemann-Taylor Fast Radiative Transfer Code (HT-FRTC) is a principal component based fast radiative transfer code that can be used across the whole electromagnetic spectrum to calculate transmittance, radiance and flux spectra. The principal components cover the spectrum at a very high spectral resolution, which allows very fast line-by-line, hyperspectral and broadband simulations for satellite-based, airborne and ground-based sensors. The principal components are derived during a code training phase from line-by-line simulations for a diverse set of atmospheric and surface conditions. The derived principal components are sensor independent, i.e. no extra training is required to include additional sensors. During the training phase, predictors that are required by the fast radiative transfer code to determine the principal component scores from the monochromatic radiances (or fluxes, transmittances) are also derived. These predictors are calculated for each training profile at a small number of frequencies, which are selected by a k-means cluster algorithm during the training phase. The predictors are calculated using Gaussian Processes, which is more accurate than a linear regression algorithm. The HT-FRTC code is an integral part of the Met Office’s Tactical Decision aids such as Neon and IRVIS and as such plays an important part in the prediction of environmental impacts on various sensors such as IR cameras and night-vision goggles. Moreover, the HT-FRTC has been incorporated into a onedimensional variation (1D-Var) retrieval system that also works solely in principal component space. This keeps the dimensions of the matrices involved small which is important for computational efficiency.

Paper Details

Date Published: 5 September 2017
PDF: 8 pages
Proc. SPIE 10402, Earth Observing Systems XXII, 1040212 (5 September 2017); doi: 10.1117/12.2276937
Show Author Affiliations
Stephan Havemann, Met Office (United Kingdom)
Gerald Wong, Met Office (United Kingdom)
Warren Lewis, Met Office (United Kingdom)

Published in SPIE Proceedings Vol. 10402:
Earth Observing Systems XXII
James J. Butler; Xiaoxiong (Jack) Xiong; Xingfa Gu, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?