skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: DO WE REALLY KNOW THE DUST? SYSTEMATICS AND UNCERTAINTIES OF THE MID-INFRARED SPECTRAL ANALYSIS METHODS

Abstract

The spectral region around 10 {mu}m, showing prominent emission bands from various dust species is commonly used for the evaluation of the chemical composition of protoplanetary dust. Different methods of analysis have been proposed for this purpose, but so far, no comparative test has been performed to test the validity of their assumptions. In this paper, we evaluate how good the various methods are in deriving the chemical composition of dust grains from infrared spectroscopy. Synthetic spectra of disk models with different geometries and central sources were calculated, using a two-dimensional radiative transfer code. These spectra were then fitted in a blind test by four spectral decomposition methods. We studied the effect of disk structure (flared versus flat), inclination angle, size of an inner disk hole, and stellar luminosity on the fitted chemical composition. Our results show that the dust parameters obtained by all methods deviate systematically from the input data of the synthetic spectra. The dust composition fitted by the new two-layer temperature distribution method, described in this paper, differs the least from the input dust composition and the results show the weakest systematic effects. The reason for the deviations of the results given by the previously used methodsmore » lies in their simplifying assumptions. Due to the radial extent of the 10 {mu}m emitting region there is dust at different temperatures contributing to the flux in the silicate feature. Therefore, the assumption of a single averaged grain temperature can be a strong limitation of the previously used methods. The continuum below the feature can consist of multiple components (e.g., star, inner rim, and disk midplane), which cannot simply be described by a Planck function at a single temperature. In addition, the optically thin emission of 'featureless' grains (e.g., carbon in the considered wavelength range) produces a degeneracy in the models with the optically thick emission of the disk. The effect of different noise levels on the results has also been tested. We find that for a signal-to-noise ratio (S/N) of 100 one can expect an absolute uncertainty in the value of the crystallinity of about 11% using ground-based observations (8-13 {mu}m). For space-based observations (7-17 {mu}m) the expected uncertainty is about 5% for the same S/N value. Moreover, the average value of the estimated crystallinity increases toward lower S/N in general. On the basis of our results, we propose a recipe for the analysis and interpretation of dust spectroscopy data in the mid-infrared which should be especially valuable for analyzing Spitzer spectroscopy data and ground-based infrared spectroscopy data in the 10 {mu}m window.« less

Authors:
; ; ;  [1];  [2]
  1. Max-Planck-Institut fuer Astronomie, Koenigstuhl 17, D-69117 Heidelberg (Germany)
  2. Johns Hopkins University, 3400 N. Charles St., Baltimore, MD 21218 (United States)
Publication Date:
OSTI Identifier:
21296234
Resource Type:
Journal Article
Journal Name:
Astrophysical Journal
Additional Journal Information:
Journal Volume: 695; Journal Issue: 2; Other Information: DOI: 10.1088/0004-637X/695/2/1024; Country of input: International Atomic Energy Agency (IAEA); Journal ID: ISSN 0004-637X
Country of Publication:
United States
Language:
English
Subject:
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; ABSORPTION SPECTROSCOPY; CARBON; CHEMICAL COMPOSITION; DECOMPOSITION; DUSTS; INFRARED SPECTRA; LUMINOSITY; PROTOPLANETS; RADIANT HEAT TRANSFER; SIGNAL-TO-NOISE RATIO; SILICATES; STARS; TEMPERATURE DISTRIBUTION

Citation Formats

Juhasz, A, Henning, Th, Bouwman, J, Dullemond, C P, Pascucci, I, and Apai, D. DO WE REALLY KNOW THE DUST? SYSTEMATICS AND UNCERTAINTIES OF THE MID-INFRARED SPECTRAL ANALYSIS METHODS. United States: N. p., 2009. Web. doi:10.1088/0004-637X/695/2/1024; COUNTRY OF INPUT: INTERNATIONAL ATOMIC ENERGY AGENCY (IAEA).
Juhasz, A, Henning, Th, Bouwman, J, Dullemond, C P, Pascucci, I, & Apai, D. DO WE REALLY KNOW THE DUST? SYSTEMATICS AND UNCERTAINTIES OF THE MID-INFRARED SPECTRAL ANALYSIS METHODS. United States. doi:10.1088/0004-637X/695/2/1024; COUNTRY OF INPUT: INTERNATIONAL ATOMIC ENERGY AGENCY (IAEA).
Juhasz, A, Henning, Th, Bouwman, J, Dullemond, C P, Pascucci, I, and Apai, D. Mon . "DO WE REALLY KNOW THE DUST? SYSTEMATICS AND UNCERTAINTIES OF THE MID-INFRARED SPECTRAL ANALYSIS METHODS". United States. doi:10.1088/0004-637X/695/2/1024; COUNTRY OF INPUT: INTERNATIONAL ATOMIC ENERGY AGENCY (IAEA).
@article{osti_21296234,
title = {DO WE REALLY KNOW THE DUST? SYSTEMATICS AND UNCERTAINTIES OF THE MID-INFRARED SPECTRAL ANALYSIS METHODS},
author = {Juhasz, A and Henning, Th and Bouwman, J and Dullemond, C P and Pascucci, I and Apai, D.},
abstractNote = {The spectral region around 10 {mu}m, showing prominent emission bands from various dust species is commonly used for the evaluation of the chemical composition of protoplanetary dust. Different methods of analysis have been proposed for this purpose, but so far, no comparative test has been performed to test the validity of their assumptions. In this paper, we evaluate how good the various methods are in deriving the chemical composition of dust grains from infrared spectroscopy. Synthetic spectra of disk models with different geometries and central sources were calculated, using a two-dimensional radiative transfer code. These spectra were then fitted in a blind test by four spectral decomposition methods. We studied the effect of disk structure (flared versus flat), inclination angle, size of an inner disk hole, and stellar luminosity on the fitted chemical composition. Our results show that the dust parameters obtained by all methods deviate systematically from the input data of the synthetic spectra. The dust composition fitted by the new two-layer temperature distribution method, described in this paper, differs the least from the input dust composition and the results show the weakest systematic effects. The reason for the deviations of the results given by the previously used methods lies in their simplifying assumptions. Due to the radial extent of the 10 {mu}m emitting region there is dust at different temperatures contributing to the flux in the silicate feature. Therefore, the assumption of a single averaged grain temperature can be a strong limitation of the previously used methods. The continuum below the feature can consist of multiple components (e.g., star, inner rim, and disk midplane), which cannot simply be described by a Planck function at a single temperature. In addition, the optically thin emission of 'featureless' grains (e.g., carbon in the considered wavelength range) produces a degeneracy in the models with the optically thick emission of the disk. The effect of different noise levels on the results has also been tested. We find that for a signal-to-noise ratio (S/N) of 100 one can expect an absolute uncertainty in the value of the crystallinity of about 11% using ground-based observations (8-13 {mu}m). For space-based observations (7-17 {mu}m) the expected uncertainty is about 5% for the same S/N value. Moreover, the average value of the estimated crystallinity increases toward lower S/N in general. On the basis of our results, we propose a recipe for the analysis and interpretation of dust spectroscopy data in the mid-infrared which should be especially valuable for analyzing Spitzer spectroscopy data and ground-based infrared spectroscopy data in the 10 {mu}m window.},
doi = {10.1088/0004-637X/695/2/1024; COUNTRY OF INPUT: INTERNATIONAL ATOMIC ENERGY AGENCY (IAEA)},
journal = {Astrophysical Journal},
issn = {0004-637X},
number = 2,
volume = 695,
place = {United States},
year = {2009},
month = {4}
}