Dealing with Zeros and Missing Values in Compositional Data Sets Using Nonparametric Imputation

The statistical analysis of compositional data based on logratios of parts is not suitable when zeros are present in a data set. Nevertheless, if there is interest in using this modeling approach, several strategies have been published in the specialized literature which can be used. In particular, substitution or imputation strategies are available for rounded zeros. In this paper, existing nonparametric imputation methods—both for the additive and the multiplicative approach—are revised and essential properties of the last method are given. For missing values a generalization of the multiplicative approach is proposed.     

Methodology for Landslide Susceptibility Mapping by Means of a GIS. Application to the Contraviesa Area (Granada,Spain)

This article presents a method to map landslide susceptibility in rock massifs using Geographical Information Systems (GIS). The method is based on making an inventory of rupture zones of different types of slope movements and then analysing the bivariate correlation of these with the factors that determine instability. After determining the factors that present the highest correlation with each type of movement, a matrix is created to combine these factors and to determine the percentage of the rupture zone in each combination, which provides an expression of the susceptibility of the terrain. The map thus obtained is divided into susceptibility classes. The susceptibility maps (made in 1995) for each type of movement are first calibrated with the inventory of the movements from which they are derived (previous to 1995), and subsequently validated by another inventory elaborated after the susceptibility maps (in 1997). In both cases, significant correlation coefficients were obtained (the Goodman–Kruskal coefficients were over 0.8 and sometimes exceeded 0.9). The relative error (degree of accumulated fit for very low to low susceptibility classes) was always less than 5%,while the relative success rate was always above 50%. These resultsillustrate the adequacy of the method and of the maps obtained.     

The effect of water on the viscosity of a haplogranitic melt under P-T-X conditions relevant to silicic volcanism

 The viscosities of hydrous haplogranitic melts synthesized by hydrothermal fusion at 2 kbar pressure and 800 to 1040° C have been measured at temperatures just above the glass transition and at a pressure of 1 bar using micropenetration techniques. The micropenetration viscometry has been performed in the viscosity range of 10⁹ Pa s to 10¹² Pa s. The samples ranged in water content from 0.4 to 3.5 wt%. For samples with up to 2.5 wt% H₂O, the water contents have been determined using infrared spectroscopy obtained before and after each viscometry experiment to be constant over the duration of the measurements. Above this water content a measurable loss of water occurs during the viscometry. The viscosity data illustrate an extremely nonlinear decrease in viscosity with added water. The viscosity drops drastically with the addition of 0.5 wt% of water and then shallows out at water contents of 2 wt%. An additional viscosity datum point obtained from the analysis of fluid inclusions in a water-saturated HPG8 confirms a near invariance of the viscosity with the addition of water between 2 and 6 wt%. These measurements may be compared directly with the data of Hess et al. (1995, in press) for the effects of excess alkali and alkaline earth oxides on the viscosity of HPG8 (also obtained at 1 bar). The viscosity of the melts, compared on an equivalent molar basis, increases in the order H₂O<(Li₂O<Na₂O< K₂O<Rb₂O,Cs₂O<BaO<SrO<CaO<MgO< BeO). The extraordinary decrease in melt viscosity with added water is poorly reproduced by the calculation scheme of Shaw (1972) for the range of water contents investigated here. The speciation of water in the quenched glasses can be used to quantify the dependence of the viscosity on hydroxyl content. Considering only the hydroxyl groups as active fluidizers in the hydrous melts the nonlinearity of the viscosity decrease and the difference with the effects of the alkali oxides becomes larger. Consequences for degassing calcalkaline rhyolite are discussed. Received: 17 August 1995/Accepted: 8 January 1996     

End-states of short-period comets and their role in maintaining the zodiacal dust cloud

Physical lifetimes and end-states of short-period comets are analysed in connection with the problem of the maintainance of the zodiacal dust cloud. In particular, the problem of the comet-asteroid relationship is addressed. Recent studies of the physical properties of Apollo-Amor asteroids and short-period comets (e.g., Hartmann et al., 1987) show significant differences between them, suggesting that they are distinct classes of objects. A few percent of the active SP comets might become asteroidal-like bodies in comet-type orbits due to the buildup of dust mantles. The remainder probably disintegrate as they consume their volatile content so their debris can only be observed as fireballs when they meet the Earth. Unobservable faint SP comets — i.e., comets so small (m 10¹⁴ g) that quickly disintegrate before being detected, might be a complementary source of dust material. They might be completely sublimated even at rather large heliocentric distances (r - 3 AU). Yet the released dust grains can reach the vicinity of the Sun by Poynting-Robertson drag. The mass associated with unobservable SP comets with perihelion distances q 3 AU might be comparable to that computed for the sample of observed SP co-mets with q 1.5 AU. It is concluded that SP comets (from the large to the unobservable small ones) may supply an average of several tons/sec of meteoric matter to the zodiacal dust cloud.     

An outline of paleolimnology in Argentina

Argentine literature from different disciplines dealing with lake sediments is presented in chronological order and briefly discussed. Quaternary and pre-Quaternary deposits are considered. Papers reveal five periods of production and three major regions of paleolimnological research.This is the first in a series of papers published in this issue on Paleolimnology in Southern South America. Dr C. A. Fernández served as guest editor for these papers.     

Mesoscale Convective Complexes over the Western Mediterranean area during 1990–1994

Summary Mesoscale Convective Systems (MCSs) data registers from June to December during 1990–94 were obtained from the Spanish National Meteorological Institute (INM). Fifteen Mesoscale Convective Complexes (MCCs) were identified through this database. Most of the MCCs developed during the last week of September. The dominant synoptic patterns related to the mesoscale systems were cold fronts at the surface with warm and moist low-level cores, and cut-off low or deep trough throughout the middle and upper levels. These synoptic patterns were found in all the fifteen cases studied.The hourly centroid location of each MCC was used to trace their tracks, which followed a general direction towards the E or NE in almost all cases. These trajectories are clearly related to the synoptic patterns found. Finally, two MCCs chosen as representative of their evolution are described and the related physical processes are discussed.With 14 Figures