Development of Robust Ratio Linear Fitting Method of Temperature and Emissivity Separation for High-Temperature Data
Sensors · 14.08.2026 · 26 (16) · 5151
DOI: 10.3390/s26165151 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 14.08.2026. Verfasst von Mitchell Manzardo und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Engineering Physics, Air Force Institute of Technology, 2950 Hobson Way, Wright-Patterson AFB, Dayton, OH 45433, USA. Die Arbeit nennt 9 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: U.S. Air Force Institute of Technology.
Abstract (Originalsprache)
Accurate temperature and emissivity separation from thermal infrared radiance is essential for characterizing materials under high-temperature laboratory conditions. Existing temperature and emissivity separation methods have largely been developed for multispectral remote sensing applications, where long atmospheric path lengths require extensive atmospheric compensation. In contrast, the current work considers hyperspectral laboratory measurements acquired over a short optical path, where atmospheric effects are comparatively small but increased measurement uncertainty remains within portions of the measured spectrum. The ABB MR304 FTIR spectrometer used in this study exhibits reduced optical transmission below approximately 2.5 μm, producing increased measurement uncertainty within the spectral region containing much of the temperature information. To address these conditions, a modified Gray Body Emissivity method, referred to as the Robust Ratio Linear Fitting method, was developed using robust linear regression, spectral masking, and iterative temperature refinement. The algorithm was validated by comparing the retrieved temperatures with pyrometer measurements and the retrieved spectral emissivities with a high-accuracy spectral emissivity database collected using a SOC-100 hemispherical directional reflectometer. When applied to radiance measurements of a carbon phenolic sample heated using a plasma torch and measured with an ABB MR304 FTIR spectrometer, the algorithm retrieved temperatures with a mean absolute percentage error of 3.05% and spectral emissivities with a mean absolute percentage error of 3.13% relative to the SOC-100 reference measurements. Although the method is ineffective at lower temperatures where the peak of the Planck radiance lies within excluded spectral regions, the results demonstrate that the proposed approach provides accurate temperature and emissivity retrieval for high-temperature laboratory FTIR measurements acquired under these experimental conditions.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Engineering Physics, Air Force Institute of Technology, 2950 Hobson Way, Wright-Patterson AFB, Dayton, OH 45433, USA.
- Los Alamos National Laboratory, Los Alamos, NM 87545, USA.
ORCID-Profile: Mitchell Manzardo, Michael Dexter, Shannon Young, John Bowlan, Anthony Franz
Förderung
- U.S. Air Force Institute of Technology
Zitieren
APA
Manzardo, M., Dexter, M., Young, S., Bowlan, J., & Franz, A. (2026). Development of Robust Ratio Linear Fitting Method of Temperature and Emissivity Separation for High-Temperature Data. Sensors, 26(16), 5151. https://doi.org/10.3390/s26165151
Vancouver
Manzardo M, Dexter M, Young S, Bowlan J, Franz A. Development of Robust Ratio Linear Fitting Method of Temperature and Emissivity Separation for High-Temperature Data. Sensors. 2026;26(16):5151. doi:10.3390/s26165151
BibTeX
@article{Manzardo2026,
author = {Manzardo, Mitchell and Dexter, Michael and Young, Shannon and Bowlan, John and Franz, Anthony},
title = {Development of Robust Ratio Linear Fitting Method of Temperature and Emissivity Separation for High-Temperature Data},
journal = {Sensors},
year = {2026},
volume = {26},
number = {16},
pages = {5151},
doi = {10.3390/s26165151}
}
EUVIMED