Automated Kerogen Classification in Microscope Images of Dispersed Kerogen Preparation

We develop the classification part of a system that analyses transmitted light microscope images of dispersed kerogen preparation. The system automatically extracts kerogen pieces from the image and labels each piece as either inertinite or vitrinite. The image pre-processing analysis consists of background removal, identification of kerogen material, object segmentation, object extraction (individual images of pieces of kerogen) and feature calculation for each object. An expert palynologist was asked to label the objects into categories inertinite and vitrinite, which provided the ground truth for the classification experiment. Ten state-of-the-art classifiers and classifier ensembles were compared: Naïve Bayes, decision tree, nearest neighbour, the logistic classifier, multilayered perceptron (MLP), support vector machines (SVM), AdaBoost, Bagging, LogitBoost and Random Forest. The logistic classifier was singled out as the most accurate classifier, with an accuracy greater than 90. Using a 10 times 10-fold cross-validation provided within the Weka software, we found that the logistic classifier was significantly better than five classifiers (p<0.05) and indistinguishable from the other four classifiers. The initial set of 32 features was subsequently reduced to 6 features without compromising the classification accuracy. A further evaluation of the system alerted us to the possible sensitivity of the classification to the ground truth that might vary from one human expert to another. The analysis also revealed that the logistic classifier made most of the correct classifications with a high certainty.     

Permeability development in vesiculating magmas: implications for fragmentation

 Fragmentation, or the "coming apart" of magma during a plinian eruption, remains one of the least understood processes in volcanology, although assumptions about the timing and mechanisms of fragmentation are key parameters in all existing eruption models. Despite evidence to the contrary, most models assume that fragmentation occurs at a critical vesicularity (volume percent vesicles) of 75–83%. We propose instead that the degree to which magma is fragmented is determined by factors controlling bubble coalescence: magma viscosity, temperature, bubble size distribution, bubble shapes, and time. Bubble coalescence in vesiculating magmas creates permeability which serves to connect the dispersed gas phase. When sufficiently developed, permeability allows subsequent exsolved and expanded gas to escape, thus preserving a sufficiently interconnected region of vesicular magma as a pumice clast, rather than fully fragmenting it to ash. For this reason pumice is likely to preserve information about (a) how permeability develops and (b) the critical permeability needed to insure clast preservation. We present measurements and calculations that constrain the conditions (vesicularity, bubble size distribution, time, pressure difference, viscosity) necessary for adequate permeability to develop. We suggest that magma fragments explosively to ash when and where, in a heterogeneously vesiculating magma, these conditions are not met. Both the development of permeability by bubble wall thinning and rupture and the loss of gas through a permeable network of bubbles require time, consistent with the observation that degree of fragmentation (i.e., amount of ash) increases with increasing eruption rate. Received: 5 July 1995 / Accepted: 27 December 1995     

Connection between X-ray and optical variability in NGC 4151

We have used long light curves in the X-ray (ARIEL V, EXOSAT, Ginga and others), UV (IUE), and optical (data of intensive ground-based monitoring) ranges of NGC 4151 to investigate connections between variations in these spectral fields. Applying Fourier (CLEAN algorithm) and modern cross-correlation analysis, we revealed the different character of the variability and the cross-correlation between high and low states of the nucleus. In contrast to earlier results, we found that X-ray and optical fluxes correlate in the active state without any apparent delay.In the power spectrum of the optical and X-ray light curves in the high state we found the existence of the same period of about 65 days, which disappeared in the low state.We confirmed previous data that X-ray variations correlate more strongly with UV and optical continuum variability during the low state, but it is possible that the optical variations appear to lag behind those involving X-rays by several tens of days.Strong variations in H line profile and [Fe X]6374 line intensity variations correlate with the observed X-ray covering dip in 1990 (Yaqoobet al., 1993).We discuss briefly the implications of these results.     

Resonance line radiation originating from a region with well-developed plasma turbulence

This study considers the influence of the effects of scattering due to Langmuir turbulent pulsations in the transfer of radiation in the spectral lines. The transfer equation of radiation in spectral lines, by taking into account scattering due to Langmuir turbulent pulsations, is written in a form convenient for application by numerical methods.The profile's intensity for a plane-parallel finite isothermal slab of a turbulent plasma in the case of complete redistribution of scattering by an atom are obtained. Numerical studies show that in this case with the broadening of spectral lines and the decreasing of self-reversal, the Langmuir frequency _pe is of the same order as the electronic Doppler width _De. Creation of the line satellites when _pe is larger than the line width is shown with the aid of numerical methods.     

The inverse planetary problem: a numerical treatment

The so-called inverse planetary problem can be stated as follows: given the distances from the centre, masses, and radii of (say) three planets of a planetary system, find the optimum polytropic index, mass, and radius of their star, and also other quantities of interest, which depend either explicitly or implicitly on the foregoing ones (e.g., central and mean density, central and mean pressure, central and mean temperature, etc.). It is hereafter tacitly assumed that the system is opaque with respect to observations concerning periods of planetary otbits; hence, we cannot have any relevant estimates due to the well-known period laws. In the present paper, the inverse planetary problem is treated numerically on the basis of the so-called global polytropic model, developed recently by the first author.     

The effect of non-thermal excitation and ionization on the hydrogen emission in impulsive solar flares

This is a quantitative investigation of the electron beam effect on the hydrogen line profiles and continuum intensity distribution during the impulsive phase of flares. The flaring atmosphere is suggested to be a hydrogenic one and its physical condition corresponds to the gas dynamics problem solution. The radiative transfer, steady-state and particle conservation equations are solved for the three-level hydrogen model atoms with continua. Return-current losses were neglected. Hydrogen line profiles are found to be slightly sensitive to nonthermal impacts with beam electrons in the cores and more sensitive in the wings. With the initial energy flux,F ₀, rising and energy spectral index, , decreasing, the wing intensities begin to increase, and the H lines are shown to have rather extended wings as is often observed. The hydrogen continua are shown to be strongly affected by nonthermal impacts. The bigger the value ofF ₀ and the smaller the value of , the greater absolute intensities of the hydrogen continua heads. This effect is more noticeable for the Balmer and Paschen continua. The head intensity slopes of them can be used for determination of these electron beam parameters on depths of the hydrogen emission origin and their following comparison with the same parameters for the coronal heights from the X-ray observations.     

Adsorption of fluoride on goethite surfaces-implications on dental epidemiology

Fluoride ion interaction with synthetically prepared goethite has been investigated over a range of pH values (4–9) and F^– concentrations (10^–3–10^–5 M). The amount of F^– retained by goethite suspensions was found to be a function of pH, media ionic strength, F^– concentration, and goethite concentration. The lowest ionic strength (0.001 M KNO₃) gave the highest adsorption medium. Uptake was minimal at pH >7 and increased with decreasing pH. Thermodynamic properties for fluoride adsorption at 298 K and 323 K were investigated. The isosteric heat of adsorption (H _r) was calculated and the heterogeneity and homogeneity of the surface examined for goethite. In view of the importance of fluoride in dental health, the interaction of fluoride on goethite in the physical environment has important implications on dental epidemiology.     

Search for low energyν e,μ,τ andv e,μ,τ in coincidence with BATSE gamma ray bursts

A search for low energy neutrinos of all flavours in correlation with 553 ray bursts detected by BATSE aboard the Compton Observatory has been performed by the LSD (Liquid Scintillator Detector) neutrino telescope. No excess ofe,, orv _e,, candidate has been detected by LSD during the time interval in which BATSE detected the 90% of the photon flux for any of the GRBs analyzed. Upper limits on the neutrino fluxes are given in the paper.     

The fir emission of dust particles around c-rich IRAS sources

The spectra of 26 IRAS sources have been fitted using experimental optical data obtained in our laboratory for different types of carbon and silicon carbide (SiC) submicronic particles. Laboratory data are the main input for a radiative transfer model, which, taking into account several important parameters of the circumstellar matter and of the central star, evaluates the temperature gradient across the envelope and hence the emission best-fit spectrum. In this way it has been possible to fit, with good accuracy, most of the examined IRAS spectra in the range 8 - 100µm. The sources have been chosen among those showing SiC feature around 11µm and the best fits have been obtained by carefully assessing the amount and the type ( &) of SiC particles present in the grain mixtures.