Bayesian Positive Source Separation Research Articles

Raman spectra of Ni–Mg kerolite: effect of Ni–Mg substitution on O–H stretching vibrations

The Ni‐rich mineral phases constituting the Ni‐ores in New Caledonia are dominated in most cases by kerolite–pimelite series, also called talc‐like minerals. Despite their economic and geologic interests, the crystallographic structures of these minerals are not fully understood. In order to improve the knowledge of their crystallographic structure, a set of natural talc‐like minerals of varying Ni/Mg ratio was selected. A combination of SEM, EMPA, TEM, and Raman analysis shows that the studied mineral contains only one mineral phase identified as the kerolite–pimelite series. Although kerolite and pimelite are not currently recognized as distinct minerals by the IMA, this study shows that these minerals exist as distinct minerals without any mixed layering with other mineral phases such as serpentine or other sheet silicates. The solid solution between the Ni and Mg endmembers was complete without any gap, corresponding to a Ni–Mg substitution in the octahedral sheet. The Raman spectra in the OH stretching vibration region of representative samples covering the whole range of Ni–Mg substitution show a continuous and significant evolution from the Mg to the Ni endmember. The signal was decomposed into nine Gaussian–Lorentzian functions, but the large overlapping of the peaks and the complexity of the band structure prevent any interpretation of the spectra from a structural point of view. Finally, the problem is formalized as a multivariate curve resolution (MCR) which is solved using the Bayesian Positive Source Separation (BPSS) algorithm. This reveals that four possible arrangements of Ni and Mg in the octahedral sheet are encountered, and these arrangements are dependent on the Ni–Mg substitution rate. It also confirms that Raman microspectroscopy can be very efficient in quick evaluation of the Ni content of the mineral. Copyright © 2015 John Wiley & Sons, Ltd.

COSIMA calibration for the detection and characterization of the cometary solid organic matter

On the orbiter of the Rosetta spacecraft, the Cometary Secondary Ion Mass Analyser (COSIMA) will provide new in situ insights about the chemical composition of cometary grains all along 67P/Churyumov–Gerasimenko (67P/CG) journey until the end of December 2015 nominally. The aim of this paper is to present the pre-calibration which has already been performed as well as the different methods which have been developed in order to facilitate the interpretation of the COSIMA mass spectra and more especially of their organic content. The first step was to establish a mass spectra library in positive and negative ion mode of targeted molecules and to determine the specific features of each compound and chemical family analyzed. As the exact nature of the refractory cometary organic matter is nowadays unknown, this library is obviously not exhaustive. Therefore this library has also been the starting point for the research of indicators, which enable to highlight the presence of compounds containing specific atom or structure. These indicators correspond to the intensity ratio of specific peaks in the mass spectrum. They have allowed us to identify sample containing nitrogen atom, aliphatic chains or those containing polyaromatic hydrocarbons. From these indicators, a preliminary calibration line, from which the N/C ratio could be derived, has also been established. The research of specific mass difference could also be helpful to identify peaks related to quasi-molecular ions in an unknown mass spectrum. The Bayesian Positive Source Separation (BPSS) technique will also be very helpful for data analysis. This work is the starting point for the analysis of the cometary refractory organic matter. Nevertheless, calibration work will continue in order to reach the best possible interpretation of the COSIMA observations.

On the production of glycogen by Pseudomonas fluorescens during biofilm development: an in situ study by attenuated total reflection-infrared with chemometrics

Attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy was used to monitor Pseudomonas fluorescens biofilms in situ, non-destructively, in real time, and under fully hydrated conditions. Changes accompanying the metabolic evolution of the sessile bacterial cells from the nascent biofilm monolayer to the beginning of the multi-layered structure in the presence of nutrients were identified via the ATR-FTIR fingerprints of the young biofilm on the ATR crystal. The ATR-FTIR spectra were analysed by classical methods (time evolution of integrated intensities and profile evolution of specific bands), and also by a multivariate curve resolution, Bayesian positive source separation, to extract the pure component spectra and their change of concentration over time occurring during biofilm settlement. This work showed clearly the overproduction of glycogen by sessile P. fluorescens, which had not previously been described by other research groups.

Coronal Temperature Maps from Solar EUV Images: A Blind Source Separation Approach

Multi-wavelength solar images in the EUV are routinely used for analysing solar features such as coronal holes, filaments, and flares. However, images taken in different bands often look remarkably similar as each band receives contributions coming from regions with a range of different temperatures. This has motivated the search for empirical techniques that may unmix these contributions and concentrate salient morphological features of the corona in a smaller set of less redundant source images. Blind Source Separation (BSS) precisely does this. Here we show how this novel concept also provides new insight into the physics of the solar corona, using observations made by SDO/AIA. The source images are extracted using a Bayesian positive source separation technique. We show how observations made in six spectral bands, corresponding to optically thin emissions, can be reconstructed by linear combination of three sources. These sources have a narrower temperature response and allow for considerable data reduction since the pertinent information from all six bands can be condensed in only one single composite picture. In addition, they give access to empirical temperature maps of the corona. The limitations of the BSS technique and some applications are briefly discussed.

Hydrolysis of Uranyl(VI) in Acidic and Basic Aqueous Solutions Using a Noncomplexing Organic Base: A Multivariate Spectroscopic and Statistical Study

In the field of actinide aqueous chemistry, this work aims to resolve some controversy about uranyl(VI) hydroxide species present in basic aqueous solutions. We revisit the Raman, IR, and UV-visible spectra with two new approaches. First, Raman, IR and UV data were recorded systematically from aqueous solutions with the noncomplexing electrolyte (C(2)H(5))(4)NNO(3) at 25 °C and 0.1 MPa ([U(total)] = 0.005-0.105 M) in H(2)O and D(2)O over a wide range of -log mH(D)(+) between 2.92 and 14.50. Second, vibrational spectra (IR and Raman) of basic solutions in H(2)O and D(2)O were analyzed using the Bayesian Positive Source Separation method to estimate pure spectra of individual species. In D(2)O solutions, the new spectroscopic data showed the occurrence of the same species as those in H(2)O. As observed for the wavenumber of the symmetric stretching mode, the wavenumber characteristic of the O═U═O antisymmetric stretching mode decreases as the number of OH(D)(-) ligands increases. These kinds of data, completed by (1) analysis of the signal widths, (2) persistence of the apparent exclusion rule between IR and Raman spectra of the uranyl species stretching modes, and (3) interpretation of the absorption UV-visible spectra, allow discussion of the chemistry, structures, and polynuclearity of uranyl(VI) species. In moderate basic solutions, the presence of two trimers is suggested. In highly basic solutions ([OH(-)] ≈ 3 M), the two monomers UO(2)(OH)(4)(2-) and UO(2)(OH)(5)(3-) are confirmed to be in good agreement with earlier EXAFS and NMR results. The occurrence of the UO(2)(OH)(6)(4-) monomer is also suggested from the more basic solutions investigated.

Implementation Strategies for Hyperspectral Unmixing Using Bayesian Source Separation

Bayesian Positive Source Separation (BPSS) is a useful unsupervised approach for hyperspectral data unmixing, where numerical non-negativity of spectra and abundances has to be ensured, such in remote sensing. Moreover, it is sensible to impose a sum-to-one (full additivity) constraint to the estimated source abundances in each pixel. Even though non-negativity and full additivity are two necessary properties to get physically interpretable results, the use of BPSS algorithms has been so far limited by high computation time and large memory requirements due to the Markov chain Monte Carlo calculations. An implementation strategy which allows one to apply these algorithms on a full hyperspectral image, as typical in Earth and Planetary Science, is introduced. Effects of pixel selection, the impact of such sampling on the relevance of the estimated component spectra and abundance maps, as well as on the computation times, are discussed. For that purpose, two different dataset have been used: a synthetic one and a real hyperspectral image from Mars.

On the decomposition of Mars hyperspectral data by ICA and Bayesian positive source separation

The surface of Mars is currently being imaged with an unprecedented combination of spectral and spatial resolution. This high resolution, and its spectral range, gives the ability to pinpoint chemical species on the surface and the atmosphere of Mars more accurately than before. The subject of this paper is to present a method to extract informations on these chemicals from hyperspectral images. A first approach, based on independent component analysis (ICA) [P. Comon, Independent component analysis, a new concept? Signal Process. 36 (3) (1994) 287–314], is able to extract artifacts and locations of CO 2 and H 2 O ices. However, the main independence assumption and some basic properties (like the positivity of images and spectra) being unverified, the reliability of all the independent components (ICs) is weak. For improving the component extraction and consequently the endmember classification, a combination of spatial ICA with spectral Bayesian positive source separation (BPSS) [S. Moussaoui, D. Brie, A. Mohammad-Djafari, C. Carteret, Separation of non-negative mixture of non-negative sources using a Bayesian approach and MCMC sampling, IEEE Trans. Signal Process. 54 (11) (2006) 4133–4145] is proposed. To reduce the computational burden, the basic idea is to use spatial ICA yielding a rough classification of pixels, which allows selection of small, but relevant, number of pixels. Then, BPSS is applied for the estimation of the source spectra using the spectral mixtures provided by this reduced set of pixels. Finally, the abundances of the components are assessed on the whole pixels of the images. Results of this approach are shown and evaluated by comparison with available reference spectra.