Filter-Based Classification of Training Image Patterns for Spatial Simulation

Multiple-point simulation, as opposed to simulation one point at a time, operates at the pattern level using a priori structural information. To reduce the dimensionality of the space of patterns we propose a multi-point filtersim algorithm that classifies structural patterns using selected filter statistics. The pattern filter statistics are specific linear combinations of pattern pixel values that represent directional mean, gradient, and curvature properties. Simulation proceeds by sampling from pattern classes selected by conditioning data.     

A statistical method for estimating rates of soil development and ages of geologic deposits: A design for soil-chronosequence studies

A statistical method for estimating rates of soil development in a given region based on calibration from a series of dated soils is used to estimate ages of soils in the same region that are not dated directly. The method is designed specifically to account for sampling procedures and uncertainties that are inherent in soil studies. Soil variation and measurement error, uncertainties in calibration dates and their relation to the age of the soil, and the limited number of dated soils are all considered. Maximum likelihood (ML) is employed to estimate a parametric linear calibration curve, relating soil development to time or age on suitably transformed scales. Soil variation on a geomorphic surface of a certain age is characterized by replicate sampling of soils on each surface; such variation is assumed to have a Gaussian distribution. The age of a geomorphic surface is described by older and younger bounds. This technique allows age uncertainty to be characterized by either a Gaussian distribution or by a triangular distribution using minimum, best-estimate, and maximum ages. The calibration curve is taken to be linear after suitable (in certain cases logarithmic) transformations, if required, of the soil parameter and age variables. Soil variability, measurement error, and departures from linearity are described in a combined fashion using Gaussian distributions with variances particular to each sampled geomorphic surface and the number of sample replicates. Uncertainty in age of a geomorphic surface used for calibration is described using three parameters by one of two methods. In the first method, upper and lower ages are specified together with a coverage probability; this specification is converted to a Gaussian distribution with the appropriate mean and variance. In the second method, absolute older and younger ages are specified together with a most probable age; this specification is converted to an asymmetric triangular distribution with mode at the most probable age. The statistical variability of the ML-estimated calibration curve is assessed by a Monte Carlo method in which simulated data sets repeatedly are drawn from the distributional specification; calibration parameters are reestimated for each such simulation in order to assess their statistical variability. Several examples are used for illustration. The age of undated soils in a related setting may be estimated from the soil data using the fitted calibration curve. A second simulation to assess age estimate variability is described and applied to the examples.     

Re-Os同位素定年方法进展及ICP-MS精确定年测试关键技术

本文介绍了Re-Os同位素定年的基本原理、技术发展及应用现状;综述了样品分解和Re-Os分离富集的主要方法,重点对ICP-MS法进行Re-Os同位素定年做了较详尽的介绍,包括质量分馏校正、干扰校正、含量初测、取样量的确定、稀释剂的稀释比及稀释剂加入量等,以确保高精度测试;评述了ICP-MS最常见的测定对象-辉钼矿中Re-Os的失耦现象及降低其对Re-Os同位素定年影响的对策,文中描述了由测定同位素比值计算含量时的误差传递公式并重申了最佳稀释比。最后,指出了Re-Os同位素定年方法研究中应该关注的工作方向。     

A prior probability method for smoothing discriminant analysis classification maps

A statistical method is presented for smoothing discriminant analysis classification maps by including pixel-specific prior probability estimates that have been determined from the frequency of tentative class assignments in a window moving across an initial per-point classification map. The class at the center of the window is reevaluated using the data for that location and the prior probability estimates obtained from the window area. An example using Landsat spectral data demonstrates the effectiveness of the method and shows an increase in classification accuracy after smoothing.     

Temporal distributions of anthropogenic Al, Zn and Pb in Hong Kong Porites coral during the last two centuries

A 182-year long record of trace metal concentrations of aluminum, zinc and lead was reconstructed from a massive Porites coral skeleton from southeastern Hong Kong to evaluate the impacts of anthropogenic activity on the marine environment. Zn/Ca and Pb/Ca ratios fluctuate synchronously from the early 19th century to the present, indicating that the marine environment has been anthropogenically influenced since industrialization. Additionally, land reclamation, mining, and ship building activities are recorded by elevated Al/Ca ratios from 1900 to 1950. The coral record indicates that high levels of Zn, Pb and Al occur coincidentally with local wars, and may have contributed to partial colony mortality. Pb/Ca does not correlate well with hemispheric proxy records after 1950, indicating that coastal corals may be recording local rather than hemispheric contamination. Pb/Ca levels in Hong Kong, Guangdong and Hainan corals imply a continuous supply of Pb-based contamination to southern China not reflected in hemispheric signals.     

Estimation of Geological Attributes from a Well Log: An Application of Hidden Markov Chains

Geophysical well logs used in petroleum exploration consist of measurements of physical properties (such as radioactivity, density, and acoustic velocity) that are digitally recorded at a fixed interval (typically half a foot) along the length of the exploratory well. The measurements are informative of the unobserved rock type alternations along the well, which is critical for the assessment of petroleum reservoirs. The well log data that are analyzed here are from a North Sea petroleum reservoir where two distinct strata have been identified from large scale seismic data. We apply a hidden Markov chain model to infer properties of the rock type alternations, separately for each stratum. The hidden Markov chain uses Dirichlet prior distributions for the Markov transition probabilities between rock types. The well log measurements, conditional on the unobserved rock types, are modeled using Gaussian distributions. Our analysis provides likelihood estimates of the parameters of the Dirichlet prior and the parameters of the measurement model. For fixed values of the parameter estimates we calculate the posterior distributions for the rock type transition probabilities, given the well log measurement data. We then propagate the model parameter uncertainty into the posterior distributions using resampling from the maximum likelihood model. The resulting distributions can be used to characterize the two reservoir strata and possible differences between them. We believe that our approach to modeling and analysis is novel and well suited to the problem. Our approach has elements in common with empirical Bayes methods in that unspecified parameters are estimated using marginal likelihoods. Additionally, we propagate the parameter uncertainty into the final posterior distributions.     

福建平潭岛晚第四纪沉积序列 及MIS 5海侵旋回特征

选择位于福建平潭岛北部芦洋埔海积平原的PT01钻孔,岩芯长度28.3 m,采用加速器碳同位素（AMS 14C）和光释光（OSL）测年建立地层年代框架,认为晚更新世富含有孔虫等海相生物化石的地层应归属于MIS 5高海面期,即MIS 5e的潮下带滨海相沉积。结合岩性特征、古生物化石、沉积物色度、粒度分析及其参数等指标,进行了沉积单元的划分,重建了MIS 6末期以来的沉积相变化过程,揭示出MIS 5期间的多次海平面波动变化,即MIS 5高海面期沉积环境仍存在滨海、潮坪、潟湖、风成砂等沉积相之间的转换,并可能存在短暂的陆相或风化过程。而该区整个末次冰期的沉积缺失或较薄,上部全新世沉积也不发育,厚度仅为3.79 m,以沼泽相―风成砂相为主。     

Numerical classification applied to certain Jamaican eocene nummulitids

Eighty-eight specimens of Eocene nummulitids from the Yellow Limestone Formation of northwestern Jamaica are classified according to quantitative measurements of morphologic parameters that are generally considered to be taxonomically useful. The specimens are grouped into homogeneous classes by the computer screening of differently oriented data projections. By this method, the use of similarity coefficients and the question of a priori weighting of characters, for which numerical taxonomy has been heavily criticized, are both avoided. The stability of the classes thus obtained is validated by discriminant analysis. These techniques provide an objective view of phenetic differences among specimens and show how the measured characters produce those differences. Tightness of coiling and total number of whorls, prove to be the most useful features in discriminating between groups but seem to have taxonomic value only at the specific and not at the generic level. This suggests that the generaOperculinoides andNummulites are synonymous.     

Longitude distribution of solar flares

Statistical tests, based on the maximum-likelihood method, have been performed on flare series extending over several years. If all flares in each plage region are taken into account, a rich spectrum is obtained. If one carries out similar analysis of a reduced flare series, which includes at most one flare from each plage region, the spectrum is almost devoid of structure, and what structure does remain is not statistically significant. The inference is that solar activity does not display genuine rigid-rotation modulation, but that repeated events from individual centers of activity yield modulation which may be mistaken for rigid-rotation modulation.A test for correlation between reduced flare series for the northern and southern hemispheres gives no significant correlation. This test therefore yieds no support to the hypothesis that solar activity is modulated by planetary effects.Department of Statistics.