Prediction of Stellar Atmospheric Parameters using Instance-Based Machine Learning and Genetic Algorithms

In this article we present a method for the automated prediction of stellar atmospheric parameters from spectral indices. This method uses a genetic algorithm (GA) for the selection of relevant spectral indices and prototypical stars and predicts their properties, using the k-nearest neighbors method (KNN). We have applied the method to predict the effective temperature, surface gravity, metallicity, luminosity class and spectral class of stars from spectral indices. Our experimental results show that the feature selection performed by the genetic algorithm reduces the running time of KNN up to 92%, and the predictive accuracy error up to 35%. This revised version was published online in July 2006 with corrections to the Cover Date.     

Numerical modelling of permafrost in bedrock in northern Fennoscandia during the Holocene

The occurrence of permafrost in bedrock in northern Fennoscandia and its dependence on past and presently ongoing climatic variations was investigated with one- (1D) and two-dimensional (2D) numerical models by solving the transient heat conduction equation with latent heat effects included. The study area is characterized by discontinuous permafrost occurrences such as palsa mires and local mountain permafrost. The ground temperature changes during the Holocene were constructed using climatic proxy data. This variation was used as a forcing function at the ground surface in the calculations. Several versions of the present ground temperature were applied, resulting in different subsurface freezing–thawing conditions in the past depending on the assumed porosity and geothermal conditions.Our results suggest that in high altitude areas with a cold climate (present mean annual ground temperature between 0°C and −3°C), there may have been considerable variations in permafrost thickness (ranging from 0 to 150 m), as well as periods of no permafrost at all. The higher is the porosity of bedrock filled with ice, the stronger is the retarding effect of permafrost against climatic variations.Two-dimensional models including topographic effects with altitude-dependent ground temperatures and slope orientation and inclination dependent solar radiation were applied to a case of mountain permafrost in Ylläs, western Finnish Lapland, where bedrock permafrost is known to occur in boreholes to a depth of about 60 m. Modelling suggests complicated changes in permafrost thickness with time as well as contrasting situations on southern and northern slopes of the mountain.Extrapolating the climatic warming of the last 200 years to the end of the next century when the anticipated increase in the annual average air temperature is expected to be about 2 K indicates that the permafrost occurrences in bedrock in northern Fennoscandia would be thawing rapidly in low-porosity formations. However, already a porosity of 5% filled with ice would retard the thawing considerably.     

On model predictions of the power spectral density of radial solar p modes

We investigate the frequency dependence of the power spectral density of low-degree solar p modes by comparing measurements with the results of a stochastic-excitation model. In the past it was common practice to use the total power in such investigations. Using the maximum of the power spectral density instead provides a direct comparison with the measured mode heights in the observed power spectrum. This method permits a more careful calibration of the adjustable parameters in the excitation model, a model which we present here, for the first time, in a format that precisely and unambiguously relates the amplitudes of the modes of oscillation to the Reynolds stress in the equilibrium model. We find that errors in the theory of the linear mode damping rates, particularly at low frequency, have a dramatic impact on the predictions of the mode heights in the spectral density, whereas parameter changes in the stochastic excitation model, within a plausible domain of parameter space, have a comparatively small effect.     

Radar observations of Itokawa in 2004 and improved shape estimation

Abstract— We present June 2004 radar images of asteroid 25143 Itokawa (1998 SF36) that improve upon the longitude‐latitude coverage of images obtained in 2001 by Ostro et al. (2004) and use the 2001–2004 data to refine that paper's constraints on Itokawa's shape. The 2004 images, the first of the asteroid's southern side, look distinctly different from the 2001 images, revealing leading edges that are much more curved and rugged than the nearly convex leading edges seen at northern latitudes in 2001. Itokawa is shaped like a slightly asymmetrical, bent, lumpy ellipsoid with dimensions along the principal axes within 10% of 594 times 320 times 288 m. To illustrate the uncertainty space associated with shape reconstruction from images with suboptimal orientational coverage, we present two alternative three‐dimensional models of the object.     

Profiles of spectral lines of an accretion ring around a Kerr black hole

Profiles of spectral lines emitted from an accretion ring around an object with strong gravitational field should be affected by Doppler shift, gravitational redshift, and deflection of light. Taking these effects into account, precise line profile of a Keplerian ring around a Kerr black hole for a distant observer is obtained by solving the kinetic equation of photons.     

Observations on the ecological importance of salt marshes in the Cumberland Basin,a macrotidal estuary in the Bay of Fundy

The Cumberland Basin, a 118 km² estuary at the head of the Bay of Fundy which has an average tidal range of about 11m, contains large tracts of salt marsh (15% of the area below highest high water). Low marsh (below about 0·9 m above mean high water) is composed almost exclusively of Spartina alterniflora while the vegetation on high marsh is more diverse but dominated by Spartina patens. Because of its higher elevation, high marsh is flooded infrequently for short periods by only extreme high tides. Low marsh is inundated much more frequently by water as much as 4m deep for periods as long as 4 h per tide. Temporal variability in the occurrence of extreme tides influences the flooding frequency of high marsh for any given month and year. Using a modification of Smalley's method, the mean annual net aerial primary production (NAPP) of low and high marsh is estimated to be 272 and 172 g C m^?2, respectively. Vegetation turnover times average 1·0 and 2·0 y for low and high marsh, respectively. Because of abundant tidal energy, much of the low marsh production appears to be exported and distributed widely about the estuary. Since high levels of turbidity suppress phytoplankton production, salt marshes produce approximately half of the carbon fixed photosynthetically in the Cumberland Basin. It is concluded that salt marshes play a major ecological role in the Cumberland Basin.     

Contact Binaries: A Study of the Proximity Effects and Gravity Darkening Based on Kopal’s Fourier Method

A new method for the determination of the proximity effects and gravity darkening exponents in contact binaries of W UMa type is presented. The method is based on Kopal’s method of Fourier analysis of the light changes of eclipsing variables in the Frequency Domain. The method was applied to 36 W UMa systems for which geometric and photometric elements have been derived by the most powerful techniques. The derived values are very close to those predicted by the existing theory of radiative transfer or convective equilibrium.