108 Matching Results

Search Results

Advanced search parameters have been applied.

Detection of greenhouse-gas-induced climatic change. Progress report, 1 December 1991--30 June 1994

Description: In addition to changes due to variations in greenhouse gas concentrations, the global climate system exhibits a high degree of internally-generated and externally-forced natural variability. To detect the enhanced greenhouse effect, its signal must be isolated from the ``noise`` of this natural climatic variability. A high quality, spatially extensive data base is required to define the noise and its spatial characteristics. To facilitate this, available land and marine data bases will be updated and expanded. The data will be analyzed to determine the potential effects on climate of greenhouse gas concentration changes and other factors. Analyses will be guided by a variety of models, from simple energy balance climate models to ocean General Circulation Models. Appendices A--G contain the following seven papers: (A) Recent global warmth moderated by the effects of the Mount Pinatubo eruption; (B) Recent warming in global temperature series; (C) Correlation methods in fingerprint detection studies; (D) Balancing the carbon budget. Implications for projections of future carbon dioxide concentration changes; (E) A simple model for estimating methane concentration and lifetime variations; (F) Implications for climate and sea level of revised IPCC emissions scenarios; and (G) Sulfate aerosol and climatic change.
Date: July 1, 1994
Creator: Wigley, T. M. L. & Jones, P. D.
Partner: UNT Libraries Government Documents Department

Maximum likelihood estimation and the multivariate Bernoulli distribution: An application to reliability

Description: We investigate systems designed using redundant component configurations. If external events exist in the working environment that cause two or more components in the system to fail within the same demand period, the designed redundancy in the system can be quickly nullified. In the engineering field, such events are called common cause failures (CCFs), and are primary factors in some risk assessments. If CCFs have positive probability, but are not addressed in the analysis, the assessment may contain a gross overestimation of the system reliability. We apply a discrete, multivariate shock model for a parallel system of two or more components, allowing for positive probability that such external events can occur. The methods derived are motivated by attribute data for emergency diesel generators from various US nuclear power plants. Closed form solutions for maximum likelihood estimators exist in many cases; statistical tests and confidence intervals are discussed for the different test environments considered.
Date: August 1, 1994
Creator: Kvam, P. H.
Partner: UNT Libraries Government Documents Department

Probability computations using the SIGMA-PI method on a personal computer

Description: The SIGMA-PI ({Sigma}{Pi}) method as implemented in the SIGPI computer code, is designed to accurately and efficiently evaluate the probability of Boolean expressions in disjunctive normal form given the base event probabilities. The method is not limited to problems in which base event probabilities are small, nor to Boolean expressions that exclude the compliments of base events, nor to problems in which base events are independent. The feasibility of implementing the {Sigma}{Pi} method on a personal computer has been evaluated, and a version of the SIGPI code capable of quantifying simple Boolean expressions with independent base events on the personal computer has been developed. Tasks required for a fully functional personal computer version of SIGPI have been identified together with enhancements that could be implemented to improve the utility and efficiency of the code.
Date: September 30, 1990
Creator: Haskin, F. E.; Lazo, M. S. & Heger, A. S.
Partner: UNT Libraries Government Documents Department

Multivariate diagnostics and anomaly detection for nuclear safeguards

Description: For process control and other reasons, new and future nuclear reprocessing plants are expected to be increasingly more automated than older plants. As a consequence of this automation, the quantity of data potentially available for safeguards may be much greater in future reprocessing plants than in current plants. The authors first review recent literature that applies multivariate Shewhart and multivariate cumulative sum (Cusum) tests to detect anomalous data. These tests are used to evaluate residuals obtained from a simulated three-tank problem in which five variables (volume, density, and concentrations of uranium, plutonium, and nitric acid) in each tank are modeled and measured. They then present results from several simulations involving transfers between the tanks and between the tanks and the environment. Residuals from a no-fault problem in which the measurements and model predictions are both correct are used to develop Cusum test parameters which are then used to test for faults for several simulated anomalous situations, such as an unknown leak or diversion of material from one of the tanks. The leak can be detected by comparing measurements, which estimate the true state of the tank system, with the model predictions, which estimate the state of the tank system as it ``should`` be. The no-fault simulation compares false alarm behavior for the various tests, whereas the anomalous problems allow one to compare the power of the various tests to detect faults under possible diversion scenarios. For comparison with the multivariate tests, univariate tests are also applied to the residuals.
Date: August 1, 1994
Creator: Burr, T.; Jones, J. & Wangen, L.
Partner: UNT Libraries Government Documents Department

Chemical recognition software

Description: We have developed a capability to make real time concentration measurements of individual chemicals in a complex mixture using a multispectral laser remote sensing system. Our chemical recognition and analysis software consists of three parts: (1) a rigorous multivariate analysis package for quantitative concentration and uncertainty estimates, (2) a genetic optimizer which customizes and tailors the multivariate algorithm for a particular application, and (3) an intelligent neural net chemical filter which pre-selects from the chemical database to find the appropriate candidate chemicals for quantitative analyses by the multivariate algorithms, as well as providing a quick-look concentration estimate and consistency check. Detailed simulations using both laboratory fluorescence data and computer synthesized spectra indicate that our software can make accurate concentration estimates from complex multicomponent mixtures, even when the mixture is noisy and contaminated with unknowns.
Date: June 1, 1994
Creator: Wagner, J. S.; Trahan, M. W.; Nelson, W. E.; Hargis, P. H. Jr. & Tisone, G. C.
Partner: UNT Libraries Government Documents Department

Multivariate methods in nuclear waste remediation: Needs and applications

Description: The United States Department of Energy (DOE) has developed a strategy for nuclear waste remediation and environmental restoration at several major sites across the country. Nuclear and hazardous wastes are found in underground storage tanks, containment drums, soils, and facilities. Due to the many possible contaminants and complexities of sampling and analysis, multivariate methods are directly applicable. However, effective application of multivariate methods will require greater ability to communicate methods and results to a non-statistician community. Moreover, more flexible multivariate methods may be required to accommodate inherent sampling and analysis limitations. This paper outlines multivariate applications in the context of select DOE environmental restoration activities and identifies several perceived needs.
Date: May 1, 1992
Creator: Pulsipher, B. A.
Partner: UNT Libraries Government Documents Department

Multivariate optimization of production systems: The time dimension

Description: Traditional analysis of oil and gas production systems treats individual nodes one at a time. This only calculates a feasible solution which is not necessarily optimal. Multivariate optimization is able to determine the most profitable configuration, including all variables simultaneously. The optimization can also find the optimal recovery over a period of time, rather than just at a single instant as in traditional methods. This report describes the development of multivariate optimization for situations in which the decision variables may change as a function of time. For example, instead of estimating a tubing size which is optimal over the life of the project, this approach determines a series of optimal tubing sizes which may change from year to year. Examples show that under an optimal strategy, tubing size can be changed only infrequently while still increasing profitability of a project. The methods used in this work considered the special requirements of objectives which are not smooth functions of their decision variables. The physical problems considered included artificial lift production systems.
Date: April 1, 1993
Creator: Ravindran, N. & Horne, R. N.
Partner: UNT Libraries Government Documents Department

Multivariable and distributed control of nonlinear chemical processes using adaptive methods. Final report, February 1, 1985--January 31, 1988

Description: In this work we studied the application of adaptive learning and optimization to chemical process control. The work covered theory as well as practical applications of adaptive and nonlinear control, including multivariable periodic control The main findings were: 1. Linear adaptive control systems may display chaotic behavior. The chaos has small amplitude if the algorithm is properly implemented. 2. Stability theory for nonlinear adaptive control has been developed. 3. Experimental evaluation of predictive control was performed. 4. A theory for periodic control and adaptive periodic control of chemical processes was developed.
Date: December 31, 1988
Creator: Ydstie, B. E.
Partner: UNT Libraries Government Documents Department

Parallel computation of automatic differentiation applied to magnetic field calculations

Description: The author presents a parallelization of an accelerator physics application to simulate magnetic field in three dimensions. The problem involves the evaluation of high order derivatives with respect to two variables of a multivariate function. Automatic differentiation software had been used with some success, but the computation time was prohibitive. The implementation runs on several platforms, including a network of workstations using PVM, a MasPar using MPFortran, and a CM-5 using CMFortran. A careful examination of the code led to several optimizations that improved its serial performance by a factor of 8.7. The parallelization produced further improvements, especially on the MasPar with a speedup factor of 620. As a result a problem that took six days on a SPARC 10/41 now runs in minutes on the MasPar, making it feasible for physicists at Lawrence Berkeley Laboratory to simulate larger magnets.
Date: September 1, 1994
Creator: Hinkins, R. L.
Partner: UNT Libraries Government Documents Department

Exploring phase space concepts in the forecasting of time series with artificial neural networks

Description: The authors study the performance of feedforward artificial neural networks in forecasting future values of several different time series. They explore both short- and long-term prediction of several periodic time series. They find that a significant source of error in long-term prediction of time series is introduced by a phase shift between the network output and the time series. They explore the origin of this phase shift and suggest strategies for minimizing its effect. They find that the phase diagrams of the time series and the neural network forecast contain useful diagnostic information.
Date: September 13, 1993
Creator: Rogers, R. D. & Vemuri, V.
Partner: UNT Libraries Government Documents Department

Control of DWPF (Defense Waste Processing Facility) melter feed composition

Description: The Defense Waste Processing Facility will be used to immobilize Savannah River Site high-level waste into a stable borosilicate glass for disposal in a geologic repository. Proper control of the melter feed composition in this facility is essential to the production of glass which meets product durability constraints dictated by repository regulations and facility processing constraints dictated by melter design. A technique has been developed which utilizes glass property models to determine acceptable processing regions based on the multiple constraints imposed on the glass product and to display these regions graphically. This system along with the batch simulation of the process is being used to form the basis for the statistical process control system for the facility. 13 refs., 3 figs., 1 tab.
Date: January 1, 1990
Creator: Edwards, R.E. Jr.; Brown, K.G. & Postles, R.L.
Partner: UNT Libraries Government Documents Department

Multivariate optimization of production systems: The time dimension

Description: Traditional analysis of oil and gas production systems treats individual nodes one at a time. This only calculates a feasible solution which is not necessarily optimal. Multivariate optimization is able to determine the most profitable configuration, including all variables simultaneously. The optimization can also find the optimal recovery over a period of time, rather than just at a single instant as in traditional methods. This report describes the development of multivariate optimization for situations in which the decision variables may change as a function of time. For example, instead of estimating a tubing size which is optimal over the life of the project, this approach determines a series of optimal tubing sizes which may change from year to year. Examples show that under an optimal strategy, tubing size can be changed only infrequently while still increasing profitability of a project. The methods used in this work considered the special requirements of objectives which are not smooth functions of their decision variables. The physical problems considered included artificial lift production systems.
Date: April 1, 1993
Creator: Ravindran, N. & Horne, R.N.
Partner: UNT Libraries Government Documents Department

Statistical techniques applied to aerial radiometric surveys (STAARS): principal components analysis user's manual. [NURE program]

Description: A Principal Components Analysis (PCA) has been written to aid in the interpretation of multivariate aerial radiometric data collected by the US Department of Energy (DOE) under the National Uranium Resource Evaluation (NURE) program. The variations exhibited by these data have been reduced and classified into a number of linear combinations by using the PCA program. The PCA program then generates histograms and outlier maps of the individual variates. Black and white plots can be made on a Calcomp plotter by the application of follow-up programs. All programs referred to in this guide were written for a DEC-10. From this analysis a geologist may begin to interpret the data structure. Insight into geological processes underlying the data may be obtained.
Date: January 1, 1981
Creator: Koch, C.D.; Pirkle, F.L. & Schmidt, J.S.
Partner: UNT Libraries Government Documents Department

1987 wet deposition temporal and spatial patterns in North America

Description: The focus of this report is on North American wet deposition temporal patterns from 1979 to 1987 and spatial patterns for 1987. The report investigates the patterns of annual precipitation-weighted average concentration and annual deposition for nine ion species: hydrogen, sulfate, nitrate, ammonium, calcium, chloride, sodium, potassium, and magnesium. Data are from the Acid Deposition System (ADS) for the statistical reporting of North American deposition data which includes the National Atmospheric Deposition Program/National Trends Network (NADP/NTN), the MAP3S precipitation chemistry network, the Utility Acid Precipitation Study Program (UAPSP), the Canadian Precipitation Monitoring Network (CAPMoN), and the daily and 4-weekly Acidic Precipitation in Ontario Study (APIOS-D and APIOS-C). Mosaic maps, based on surface estimation using kriging, display concentration and deposition spatial patterns of pH, hydrogen, sulfate, nitrate, ammonium, and calcium ion species for 1987 annual, winter, and summer periods. The temporal pattern analyses use a subset of 39 sites over a 9-year (1979--1987) period and an expanded subset of 140 sites with greater spatial coverage over a 6-year (1982--1987) period. 68 refs., 15 figs., 15 tabs.
Date: March 1, 1990
Creator: Simpson, J.C. & Olsen, A.R.
Partner: UNT Libraries Government Documents Department

Investigation of an empirical probability measure based test for multivariate normality

Description: Foutz (1980) derived a goodness of fit test for a hypothesis specifying a continuous, p-variate distribution. The test statistic is both distribution-free and independent of p. In adapting the Foutz test for multivariate normality, we consider using chi/sup 2/ and rescaled beta variates in constructing statistically equivalent blocks. The Foutz test is compared to other multivariate normality tests developed by Hawkins (1981) and Malkovich and Afifi (1973). The set of alternative distributions tested include Pearson type II and type VII, Johnson translations, Plackett, and distributions arising from Khintchine's theorem. Univariate alternatives from the general class developed by Johnson et al. (1980) were also used. An empirical study confirms the independence of the test statistic on p even when parameters are estimated. In general, the Foutz test is less conservative under the null hypothesis but has poorer power under most alternatives than the other tests.
Date: January 1, 1984
Creator: Booker, J.M.; Johnson, M.E. & Beckman, R.J.
Partner: UNT Libraries Government Documents Department

Geochemical and physical properties of soils and shallow sediments at the Savannah River Site

Description: A program to characterize the geochemical and physical properties of the unimpacted soils and shallow sediments at the Savannah River Site (SRS) has been completed. The maximum, minimum, median, standard deviation, and mean values for metals, radionuclides, inorganic anions, organic compounds, and agricultural indicator parameters are summarized for six soil series that were identified as representative of the 29 soil series at SRS. The soils from unimpacted areas of SRS are typical of soils found in moderately aggressive weathering environments, including the southeastern United States. Appendix 8 organic compounds were detected in all samples. Since these constituents are not generally present in soil, this portion of the investigation was intended to assess possible laboratory artifacts. An additional objective of the SRS Soil Study was to determine if the composition of the split spoon sampler biased chemical analysis of the soils. Twenty-five duplicate samples were analyzed for a number of metals, radiological and agricultural parameters, and organics by two laboratories currently contracted with to analyze samples during waste site characterization. In all cases, the absolute values of the average differences are relatively small compared to the overall variability in the population. 31 refs., 14 figs., 48 tabs.
Date: August 31, 1990
Creator: Looney, B.B.; Eddy, C.A.; Ramdeen, M.; Pickett, J. (Savannah River Lab., Aiken, SC (USA)); Rogers, V. (Soil Conservation Service, Aiken, SC (USA). Savannah River Site Savannah River Lab., Aiken, SC (USA)); Scott, M.T. et al.
Partner: UNT Libraries Government Documents Department

The comparative uptake and interaction of several radionuclides in the trophic levels surrounding the Los Alamos Meson Physics Facility (LAMPF) waste water ponds

Description: A study was undertaken to examine the uptake, distribution, and interaction of five activation products (Co-57, Be-7, Cs-134, Rb-83, and Mn-54) within the biotic and abiotic components surrounding the waste treatment lagoons of the Los Alamos Meson Physics Facility (LAMPF). The study attempted to ascertain where, and what specific interactions were taking place among the isotopes and the biotic/abiotic components. A statistical approach, utilizing Multivariate Analysis of Variance (MANOVA), was conducted testing the radioisotopic concentrations by (1) the trophic levels (TROPLVL) in each position sampled on the grid, (2) where sampled on the grid (TRAN), (3) where sampled with-in each grid line (PLOT), and (4) the side with which sampled (SIDE). This provided both the dependent and independent variables that would be tested. The Null Hypothesis (Ho) tested the difference in the mean values of the isotopes within/between each of the four independent variables. The Rb-83 statistic indicated an accumulation within the TRAN and PLOT variables within the sampled area. The Co-57 test statistic provided a value which indicated that accumulation of this isotope within TROPLVL was taking place. Mn-54 test values indicated that accumulation was also taking place at the higher trophic levels within the PLOT, TRAN, and SIDE positions. Cs-134 was found to accumulate to third level in this trophic level structure (TROPLVL-(vegetation)), and then decrease from there. The Be-7 component provided no variance from known compartmental transfers. 210 refs., 17 figs., 4 tabs.
Date: August 1, 1989
Creator: Brooks, G.H. Jr.
Partner: UNT Libraries Government Documents Department

Modeling and generating input processes

Description: This tutorial paper provides information relevant to the selection and generation of stochastic inputs to simulation studies. The primary area considered is multivariate but much of the philosophy at least is relevant to univariate inputs as well. 14 refs.
Date: January 1, 1987
Creator: Johnson, M.E.
Partner: UNT Libraries Government Documents Department

A method for evaluating the effectiveness of site characterization measurements

Description: A quantitative approach for evaluating the effectiveness of site characterization measurement activities is developed and illustrated with an example application to hypothetical measurement schemes at a potential geologic repository site for radioactive waste. The method is a general one and could also be applied at sites for underground disposal of hazardous chemicals. The approach presumes that measurements will be undertaken to support predictions of the performance of some aspect of a constructed facility or natural system. It requires a quantitative performance objective, such as groundwater travel time or contaminant concentration, against which to compare predictions of performance. The approach recognizes that such predictions are uncertain because the measurements upon which they are based are uncertain. The effectiveness of measurement activities is quantified by a confidence index, ..beta.., that reflects the number of standard deviations separating the best estimate of performance from the predetermined performance objective. Measurements that reduce the uncertainty in predictions lead to increased values of ..beta... 5 refs., 4 figs.
Date: January 1, 1987
Creator: Ditmars, J.D.
Partner: UNT Libraries Government Documents Department

Stepwise Calculation for the Determination of the Number of Transfer Units in Countercurrent Extraction Columns

Description: Report discussing a method for the stepwise calculation of the number of transfer units utilized in a given solvent extraction operation. "Use of the method results in a very appreciable saving in time of calculation with an error of 1.5% for the runs tested. The error tends to become smaller with increasing total numbers of transfer units involved."
Date: September 12, 1949
Creator: Burns, W. A. & Eschbach, E. A.
Partner: UNT Libraries Government Documents Department

Multivariable and distributed control of nonlinear chemical processes using adaptive methods

Description: In this work we studied the application of adaptive learning and optimization to chemical process control. The work covered theory as well as practical applications of adaptive and nonlinear control, including multivariable periodic control The main findings were: 1. Linear adaptive control systems may display chaotic behavior. The chaos has small amplitude if the algorithm is properly implemented. 2. Stability theory for nonlinear adaptive control has been developed. 3. Experimental evaluation of predictive control was performed. 4. A theory for periodic control and adaptive periodic control of chemical processes was developed.
Date: January 1, 1988
Creator: Ydstie, B.E.
Partner: UNT Libraries Government Documents Department

Multivariate methods in nuclear waste remediation: Needs and applications

Description: The United States Department of Energy (DOE) has developed a strategy for nuclear waste remediation and environmental restoration at several major sites across the country. Nuclear and hazardous wastes are found in underground storage tanks, containment drums, soils, and facilities. Due to the many possible contaminants and complexities of sampling and analysis, multivariate methods are directly applicable. However, effective application of multivariate methods will require greater ability to communicate methods and results to a non-statistician community. Moreover, more flexible multivariate methods may be required to accommodate inherent sampling and analysis limitations. This paper outlines multivariate applications in the context of select DOE environmental restoration activities and identifies several perceived needs.
Date: May 1, 1992
Creator: Pulsipher, B.A.
Partner: UNT Libraries Government Documents Department

Distribution selection in statistical simulation studies

Description: The statistics profession has been remiss in exploiting the numerous advances in simulation methodology. The purpose of this article is to outline progress in variate generation relevant to the conduct of statistical simulation studies. The emphasis is on multivariate distributions, a thriving area of research. 11 refs.
Date: January 1, 1986
Creator: Johnson, M.E.
Partner: UNT Libraries Government Documents Department

Identification of multivariate linear systems

Description: Multivariate identification problems are treated with a least-squares approach. A chapter on scalar problems focuses attention on the classical parameter-estimate bias problem caused by measurement noise and develops a straightforward and effective way to remove the bias. A chapter on multivariate problems generalizes the bias removal method and develops a form selection procedure. The form selection procedure generally ensures more accurate identification than is possible with identification methods which rely on a fixed form. The concept of a form selection procedure is new to this work. A results chapter presents four example problems. Each example illustrates specific features of the identification technique. As a collection the examples emphasize the complexity of system identification and demonstrate that identification techniques will perform well when carefully applied.
Date: April 1, 1982
Creator: Griffith, J.M.
Partner: UNT Libraries Government Documents Department