Estimation for partially observed Markov processes

Many stochastic process models for environmental data sets assume a process of relatively simple structure which is in some sense partially observed. That is, there is an underlying process (X_n, n 0) or (X_t, t 0) for which the parameters are of interest and physically meaningful, and an observable process (Y_n, n 0) or (Y_t, t 0) which depends on the X process but not otherwise on those parameters. Examples are wide ranging: the Y process may be the X process with missing observations; the Y process may be the X process observed with a noise component; the X process might constitute a random environment for the Y process, as with hidden Markov models; the Y process might be a lower dimensional function or reduction of the X process. In principle, maximum likelihood estimation for the X process parameters can be carried out by some form of the EM algorithm applied to the Y process data. In the paper we review some current methods for exact and approximate maximum likelihood estimation. We illustrate some of the issues by considering how to estimate the parameters of a stochastic Nash cascade model for runoff. In the case of k reservoirs, the outputs of these reservoirs form a k dimensional vector Markov process, of which only the kth coordinate process is observed, usually at a discrete sample of time points.     

Crystallization processes in an artificial magma: variations in crystal shape,growth rate and composition with melt cooling history

A large (4.8 m³, 1.3x10⁷ g) artificial mafic melt with a bulk composition similar to that of a basalt (but with a high CaO content of 17 wt%) was generated during a demonstration of in situ vitrification and was allowed to cool naturally. During the melting process, convection was vigorous, resulting in a chemically and thermally homogeneous melt body. Once heating was complete, the cooling rate was rapid with the temperature dropping from 1500°C to 500°C in 6 days within the interior of the 3 m diameter, 1.5 m thick body. A 20 h period of constant temperature (1140°C) observed during colling was the result of latent heat released by widespread crystallization. The final crystalline assemblage consists of diopsidic to hedenbergitic pyroxene and anorthitic feldspar, with a subordinate amount of potassic feldspar, plus a small amount of evolved glass. The compositions and proportions of phases agree well with those predicted by the MELTS thermodynamic model. Thermal and textural evidence suggest that convection within the melt ceased coincident with formation of the first crystals. Textural investigation of core samples reveals large (up to 1 cm in length) acicular diopsidic pyroxenes in a matrix of smaller feldspar and zoned pyroxene crystals (500 m in length). Crystal shape and pyroxene composition vary as a function of position within the solidified body, as a function of cooling rate. Both crystal size and degree of crystallinity are highest in the central, most slowly-cooled parts of the rock. Crystal shape is characterized by tabular, equilibrium-growth forms in the slowly-cooled areas, grading to highly skeletal, dendritic forms at the rapidly-cooled edges of the body. The pyroxene crystals are dominantly homogeneous diopside, but crystals are characterized by thin Fe-rich hedenbergitic rims. These rims were deposited when Mg in the melt was depleted by diopside growth, and melt temperature had cooled sufficiently to allow Fe-rich pyroxene growth. Crystal growth rates can be calculated based on thermal behavior of the melt, reinforced by thermodynamic modelling, and are determined to be between 10^-7 and 10^-8 cm/s in the central part of the melt. These estimates agree well with growth rates in natural systems with similar cooling rates. Pyroxene crystals that formed at a higher cooling rates are characterized by higher Al and lower Mg contents relative to tabular equilibrium crystalline forms, presumably as a result of disequilibrium melt compositions at the crystal-melt interface.New Mexico Bureau of Mines and Mineral Resources, New Mexico Institute of Mining and Technology, Socorro, NM 87801, USA     

合肥遥测台网地震速报系统

论述了适用于遥测地震台网大震速报的一套计算机处理系统,系统设计以《全国遥测台网观测技术规范》和《近震分析》理论为依据,实现了数据输入,震相选择,数据处理,地震定位,震级计算,结果输出及贮存等多项功能,达到了快速确定地震三要素,提高速报速度和精度之目的,是一个较完善的,实用的速报系统。     

Sedimentary Features of Shallow Ancient River Channels on the Northern Shelf of the South China Sea

Quaternary buried ancient river channels are widespread in the shallow-level sediments of the northern shelf of the South China Sea. The sedimentary sequence mainly of fluvial deposits comprise an important component part of the low-stand system tract and transgressive system tract in the study region. The plannar variation and spatial association of the sedimentary features such as incised valley fillings, deltaic foreset wedges and block slides of shelf-marginal fans reflect the palaeogeographic environment during the fall of the regional sea level in the northern part of the South China Sea. Based on the high-resolution seismic reflection data and gelogical data from boreholes, the present paper makes an integrated interpretation of the Quaternary ancient river channels in the shallow sediments of the study area, studies the sedimentary features of the ancient channels such as their spatial distribution, seismic facies reflection indicators, sedimentary facies and sand -body types, and discusses thei     

湘中“桃江式”锰矿含锰岩系的定量研究及其找矿意义

本文从分析地质体数学特征的思想出发,来研究“桃江式”锰矿的含矿母岩－“含锰岩系”的数学特征。含锰岩系主要由五种岩石组成：粘土岩、黑色页岩、条带状页岩、含锰灰岩和碳酸锰矿层,其数学特征研究包括其统计特征分析、结构特征分析和空间特征分析。文最后讨论了这些数学特征在寻找同类锰矿中的意义。     

A topological approach to understand a multiple-loop solar flare

We analyze the UV and X-ray data obtained by the SMM satellite for the flare starting at 02:36 UT on November 12, 1980 in AR 2779. From a detailed revision of the Ov emission, we find that the observations are compatible with energy being released in a zone above the magnetic inversion line of the AR intermediate bipole. This energy is then transported mainly by conduction towards the two distant kernels located in the AR main bipole. One of these kernels is first identified in this paper. Accelerated particles contribute to the energy transport only during the impulsive phase.We model the observed longitudinal magnetic field by means of a discrete number of subphotospheric magnetic poles, and derive the magnetic field overall topology. As in previous studies of chromospheric flares, the Ov kernels are located along the intersection of the computed separatrices with the photosphere. Especially where the field-line linkage changes discontinuously, these kernels can be linked in pairs by lines that extend along separatrices. Our results agree with the hypothesis of magnetic energy released by magnetic reconnection occurring on separatrices.     

Energy release in a prominence-loaded flaring loop

Zaitsev and Stepanov (1991, 1992) proposed a mechanism for energy release in solar flares that involves the intrusion of dense prominence material into a coronal loop. The resulting non-steady state conditions are claimed to increase the resistance of the loop by 8–10 orders of magnitude. It is shown here that the dramatic increase in resistance calculated by Zaitsev and Stepanov depends on a gross overestimate of the of the magnitude of the magnetic force in the loop prior to the flare trigger. A more realistic estimate of the increase due to the mechanism suggests that it is by no more than about four orders of magnitude. As a consequence, the prominence-loading mechanism does not provide a tenable flare model.     

Universal birefringent filter with a new double passband mode

This article describes the universal birefringent filter (UBF) which will be mounted at Taiwan. The UBF permits observations of solar images, vector magnetic fields and/or line-of-sight velocity fields in any Fraunhofer line in the wavelength region from 4000 to 7000 with half width from 0.05 to 0.14 . We have tested it at ten Fraunhofer lines under a spectral telescope in Huairou Station of Beijing Astronomical Observatory and obtained the passband profiles. The principles of magnetic field and velocity field measurement with the UBF are also described.     

An examination of directional discontinuities and magnetic polarity changes around interplanetary sector boundaries usingE > 2 keV electrons

Past studies of interplanetary magnetic sector boundaries have been based on the assumption that one can determine the field polarities by comparing the field directions with those of the nominal Parker spiral angles. Previous investigators have found evidence for decreases of B, the magnitude of the magnetic fieldB, and increases of , the angle betweenB and the ecliptic plane, at sector boundaries. Others have argued that the characteristic thickness of sector boundaries exceeds that of tangential discontinuities, making sector boundaries a separate class of structures.We use a simple technique for inferring the polarities of interplanetary magnetic fields based on the assumption thatE > 2 keV electrons are always flowing along the magnetic field away from the Sun. Electron data from the UC Berkeley experiment on the ISEE-3 spacecraft are used to examine periods around several apparent sector boundaries in 1978 and 1979. We compare properties of (a) boundaries with field polarity changes and (b) large-angle ( > 60°) directional discontinuities with no field polarity changes. We find no significant differences between the sector boundaries and the directional discontinuities in terms of associated decreases in B or of values of . These results suggest no significant difference between sector boundaries and directional discontinuities other than the change in field polarities. Within limited statistics we find that about half the polarity changes would not have been identified using a requirement that > 90° and that half of the > 120° discontinuities would have been misidentified as polarity changes.