九道岭煤矿贯通测量方案设计与新方法研究

设计了两套九道岭煤矿副井与西风井之间-700m水平大巷贯通测量方案,并从精度和工程预算等方面分析了其优缺点,选出最优方案。讨论了井下导线加测陀螺边的优点,提出测陀螺边的最佳位置为2i/（2m＋1）L （L为导线全长,适用等边直伸型导线）,最佳条数为1～2条。     

线性模型的泛最小二乘法

本文首先针对线性模型提出了泛最小二乘法,在设计矩阵不加限制的情形下,得到了参数的泛最小二乘估计量。该方法既发扬了最小二乘法的优点,又克服了它的一些不足,它包含了常见的岭估计和最小二乘估计法;其次讨论了泛最小二乘法的理论依据;接着研究了泛最小二乘估计量的一些统计性质,并与最小二乘估计进行比较,在一定意义上前者优于后者;然后讨论了平衡参数的选取问题;最后,给出一个应用,说明了泛最小二乘法的有效性和可行性。     

基于相态识别的S波段双线偏振雷达最优化定量降水估测方法研究

为了提高雷达定量降水估测的精度,本文利用雨滴谱数据、实际雨量以及不同偏振参量建立起的降水估测公式,参考CSU-ICE算法降水估测模型,建立了一种基于相态识别的S波段双线偏振雷达最优化定量降水估测算法(简称HCALIQ)。利用广东珠海S波段双线偏振雷达数据和华南密集的地面自动站网的雨量,以两次华南夏季典型的降水过程为例,对该最优化算法的估测效果进行了统计分析,并进一步与CSU-ICE方法、传统的R(Z_H)方法进行了比较。结果表明:以地面雨量计为标准,HCA-LIQ最优化算法表现出与雨量计较强的相关性且有着较好的稳定性。雷达相对地面雨量计小时雨量估测的偏差分布与离雷达的距离关系不大。按过程分类的结果显示,强对流云降水时,两种最优化算法要明显优于传统的R(Z_H)方法;混合云降水时,最优化算法的效果并不比R(Z_H)方法好;HCA-LIQ最优化算法比CSU-ICE算法效果更佳。按雨强分类统计时发现,与传统的R(Z_H)方法相比,HCA-LIQ最优化算法对小雨的估测偏差降低了23%,对中雨的估测偏差相当,对大雨、暴雨的估测偏差分别降低了71%、68%。     

Comparing an estimate of seabirds at risk to a mortality estimate from the November 2004 Terra Nova FPSO oil spill

On 21 November 2004, about 1000 barrels of crude oil were accidentally released from the Terra Nova FPSO (floating production, storage and offloading) onto the Grand Banks, approximately 340 km east-southeast of St. John's, Newfoundland. We estimated the number of vulnerable seabirds (murres (Uria spp.) and dovekies (Alle alle)) at risk from this incident by multiplying observed densities of seabirds with the total area covered by the slick, estimated at 793 km(2). A mean density of 3.46 murres/km(2) and 1.07 dovekies/km(2) on the sea surface was recorded during vessel-based surveys on 28 and 29 November 2004, with a mean density of 6.90 murres/km(2) and 13.43 dovekies/km(2) combining those on the sea and in flight. We calculated a mean of 9858 murres and dovekies were at risk of being oiled, with estimates ranging from 3593 to 16,122 depending on what portion of birds in flight were assumed to be at risk. A mortality model based on spill volume was derived independently of the risk model, and estimated that 4688 (CI 95%: 1905-12,480) birds were killed during this incident. A low mortality estimate based strictly on spill volume would be expected for this incident, which occurred in an area of relatively high seabird densities. Given that the risk and mortality estimates are statistically indistinguishable, we estimate that on the order of 10,000 birds were killed by the Terra Nova spill.     

大气环境氮同位素示踪及生物监测

大气中氮的转化、迁移和沉降等大气过程与地表氮的生物地球化学循环密切相关。氮同位素对大气氮源和氮的大气化学过程的定量研究已引起广泛关注。本文对大气不同氮源的同位素特征、氮在大气过程中的同位素分馏机制以及如何定量化大气氮过程的方法进行了简要阐述,并介绍了苔藓氮同位素作为一种新的研究手段在示踪大气氮源和大气氮过程等方面取得的新进展。对国内开展的有关大气氮同位素以及苔藓氮同位素方面的工作进行了重点介绍。     

内蒙古兴和县二道河治理工程设计

本文通过工程实例分析了内蒙古兴和县前、后河和二道河的气象水文工程地质环境及其特点,对河道的治理进行了论述,提出了治理设计、施工技术要点,水土保持和环境保护的保障措施,得出投资概算与防洪效益是合理的结论。     

Bayesian estimation of isotopic age differences

Isotopic dating is subject to uncertainties arising from counting statistics and experimental errors. These uncertainties are additive when an isotopic age difference is calculated. If large, they can lead to no significant age difference by classical statistics. In many cases, relative ages are known because of stratigraphic order or other clues. Such information can be used to establish a Bayes estimate of age difference which will include prior knowledge of age order. Age measurement errors are assumed to be log-normal and a noninformative but constrained bivariate prior for two true ages in known order is adopted. True-age ratio is distributed as a truncated log-normal variate. Its expected value gives an age-ratio estimate, and its variance provides credible intervals. Bayesian estimates of ages are different and in correct order even if measured ages are identical or reversed in order. For example, age measurements on two samples might both yield 100 ka with coefficients of variation of 0.2. Bayesian estimates are 22.7 ka for age difference with a 75% credible interval of [4.4, 43.7] ka.This paper was presented at Emerging Concepts, MGUS-87 Conference, Redwood City, California, 13–15 April 1987.     

Oil and gas in offshore tracts: Bias in estimates of reserves

Two independent data sets show that current means of estimating reserves of oil and gas in offshore fields are biased. The bias which is established here is that small fields (with less than about 10 million barrels of oil equivalent) are overestimated, and larger fields (at least up to 200 MMBOE) are underestimated. This systematic bias occurs with or without a significant overall bias, and in addition to a large scatter. Because bias (as opposed to scatter) has serious implications for national energy policy, and in particular for conduct of offshore lease sales, these results indicate that evaluation methods should be revised, or that ad hoc adjustments (based on historical data) should be applied.     

由GMS卫星资料获取我国夏季地表辐射收支

地表辐射收支是一个重要的基本辐射参数，本文导得卫星可见光和红外波段测值与地表辐射收支的基本关系，据此利用GMS静止气象卫星可见光和红外通道数值化资料和地面辐射收支观测资料，建立了几个由卫星资料估计地面辐射收支的模式，选择效果最佳的一种模式，通过内插方法求取全国辐射收支分布，回归效果与实测结果比较表明，方法有实际应用价值。     

导线网方差分量估计的综合研究

提出方差分量估计的多余观测分量平均匹配方式；总结出导线网方差分量估计定权的基本规律，并给出其权增量循环算法和特征根法以及将两个实对称矩阵同时对角化的实用方法。