Search NASA⌕ Search

SEARCH · Search NASA

Results for “forensics”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 955 records · Page 53

Variation of lead isotopic composition and atomic weight in terrestrial materials (IUPAC Technical Report)

The isotopic composition and atomic weight of lead are variable in terrestrial materials because its three heaviest stable isotopes are stable end-products of the radioactive decay of uranium ( 238 U to 206 Pb; 235 U to 207 Pb) and thorium ( 232 Th to 208 Pb). The lightest stable isotope, 204 Pb, is primordial. These variations in isotope ratios and atomic weights provide useful information in many areas of science, including geochronology, archaeology, environmental studies, and forensic science. While elemental lead can serve as an abundant and homogeneous isotopic reference, deviations from the isotope ratios in other lead occurrences limit the accuracy with which a standard atomic weight can be given for lead. In a comprehensive review of several hundred publications and analyses of more than 8000 samples, published isotope data indicate that the lowest reported lead atomic weight of a normal terrestrial materials is 206.1462 ± 0.0028 (k = 2), determined for a growth of the phosphate mineral monazite around a garnet relic from an Archean high-grade metamorphic terrain in north-western Scotland, which contains mostly 206 Pb and almost no 204 Pb. The highest published lead atomic weight is 207.9351 ± 0.0005 (k = 2) for monazite from a micro-inclusion in a garnet relic, also from a high-grade metamorphic terrain in north-western Scotland, which contains almost pure radiogenic 208 Pb. When expressed as an interval, the lead atomic weight is [206.14, 207.94]. It is proposed that a value of 207.2 be adopted for the single lead atomic-weight value for education, commerce, and industry, corresponding to previously published conventional atomic-weight values.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Using Hygrothermal Modeling to Resolve Practical Low-Slope Roofing Issues

Within building envelope science, the use of tools to model performance of the building envelope has evolved, and designers and scientists today have a myriad of tools available. Hygrothermal modeling is one such tool and, when used correctly, offers designers and forensic investigators important information about the movement of moisture within the building envelope. This traditional use of hygrothermal modeling is maturing as models are validated with field experience, creating a confidence in the modeling tools and their predictive capabilities. Based on this confidence, application of hygrothermal modeling to a more practical application is an evolution whose time has arrived. With changes in materials and methods, the performance of roofing systems can often times no longer be based on experience alone. Roofing contractors have been confounded by the unexplained and unplanned-for presence of moisture in roofing systems. Although hygrothermal modeling can be used to explain where the moisture came from, we propose that it should be used to prevent its occurrence. Unfortunately, more than 70 % of the roofing market is in reroofing, and oftentimes, a designer is not engaged in the roofing system design. Without the expertise of a designer and given that hygrothermal modeling is not well understood by the roofing industry, its adoption as a tool that can prevent problems has not been widely seen. This paper presents an extensive summary of hygrothermal modeling performed for roofing systems and then shows how the modeling can be used to create simple resources for roofing contractors to reduce the potential for moisture condensation at critical locations within a roofing system. The systems modeled have shown a propensity for moisture accumulation, and the use of hygrothermal modeling to eliminate these problems is a practical application of a building science tool that can greatly improve the durability of roofing systems.

Desjarlais, Andre Omer↗

Improved Parametric Models for Explosion Pressure Signals Derived From Large Datasets

Accurate recording and characterization of explosion-induced pressure signals are key components of the forensic analysis of explosion events in the atmosphere. Parametric overpressure models based on several key waveform features (peak overpressure, positive pulse duration, and impulse) are widely used to estimate explosion energy in terms of trinitrotoluene equivalent yield. However, those models are often developed by a limited dataset, including only a few events or recordings at relatively short propagation distances. Here, we develop empirical waveform-parameter models based on a regression analysis of a large set of data curated from four chemical explosion experiments including 16 detonations. We measured peak overpressure and impulse for positive and negative phases from 1000 pressure signals recorded at local ranges (<20 km) with scaled distance up to 8000 m/kg 1/3 . Additionally, the measured waveform parameters showed large variation with respect to observing distances indicating the effects of atmospheric propagation. In this study, a second-order polynomial model was used in a least-squares regression to account for those propagation effects and to improve data fitting. In addition to model parameters for waveform features, we also determined range-dependent model uncertainties based on data variance. The model uncertainty represents the prediction error of our models and can be critical to evaluating the uncertainty of yield estimate.

58 GEOSCIENCES↗

The Seismic Signature of a High-Energy Density Physics Laboratory and Its Potential for Measuring Time-Dependent Velocity Structure

The Z Machine at Sandia National Laboratories is a pulsed power facility for high-energy density physics experiments that can shock materials to extreme temperatures and pressures through a focused energy release of up to ~ 25 MJ in < 100 nanoseconds. It has been in operation for more than two decades and conducts up to ~ 100 experiments, or “shots,” per year. Based on a set of 74 known shot times from 2018, we determined that Z Machine shots produce detectable ~ 3–17 Hz ground motion 12 km away at the Albuquerque Seismological Laboratory, New Mexico (ANMO), borehole seismograph, with peak signal at ~ 7 Hz. The known shot waveforms were used to create a three-component template, leading to the detection of 2339 Z Machine shots since 1998 through single-station cross-correlation. Local seismic magnitude estimates range from local magnitude (M L ) —2 to —1.3 and indicate that only a small fraction of the shot energy is transmitted by seismic phases observable at 12 km distance. The most recent major facility renovation, which was intended to decrease mechanical dissipation, is associated with an abrupt decrease in observed seismic amplitudes at ANMO despite stable maximum shot energy. The highly repetitive impulsive sources are well suited to coda-wave interferometry to investigate time-dependent velocity structures. Relative velocity variations (dv/v) show an annual cycle with amplitude of ~ 0.2%. Local minima are observed in the late spring, and dv/v increases through the summer monsoon rainfall, possibly reflecting patchy saturation as rainfall infiltrates near the eastern edge of the Albuquerque basin. Importantly, the cumulative results demonstrate that forensic seismology can provide insight into long-term operation of facilities such as pulsed-power laboratories, and that their recurring signals may be valuable for studies of time-dependent structure

54 ENVIRONMENTAL SCIENCES↗

Standardized beamline particulate characterization analysis: Initial application to CEBAF and LCLS-II cryomodule components

Despite continuously evolving efforts to minimize and manage the occurrence of particulates in operational SRF accelerator systems, the presence of electron field emission from contaminating particulates on SRF surfaces with high surface electric fields remains a consistent challenge. Jefferson Lab has recently initiated a standardized particulate sampling and characterization practice in order to gain more specific and systematic knowledge of the particulates actually present. It is expected that patterns emerging from such sampling will strengthen source attribution and guide improvement efforts. Initial samples were gathered from a cryomodule and adjoining warm girders that were removed from the CEBAF tunnel for reprocessing. The collection and analysis techniques were also used to characterize particulates on the inside of LCLS-II string components. Samples are transferred to clean industry-standard forensic GSR carbon tape spindles and examined via automated cleanroom SEM scanning for particle localization trends and sizing. The particulates are subsequently analyzed with EDS for elemental composition. A catalogue of found particle types is being accumulated. The methods used and results obtained from these initial applications will be presented.

Reece, Charles↗

Automation of particulate characterization

Foreign particles residing on the field carrying surface of accelerator cavities are a known mechanism for field emission. Developing the methods and tools for collecting and characterizing particles found in an accelerator enables process development towards field emission free SRF cavities. Methods are presented for sampling assemblies, components, processes, and environmental conditions utilizing forensic techniques with specialized tooling. Sampling activities to date have produced an inventory of over 850 GSR spindles. Traditional SEM + EDS analysis of this volume of spindles is challenged by labor investment, spindle sampling methods, and the subsequent data pipeline which ultimately results in a statically inadequate dataset for any particulate distribution characterization. A complete systematic analysis of the spindles is enabled by third party software controlling SEM automation for EDS data acquisition. Details of spindle creation, collection equipment, component sampling, automating particle assessment, and data analysis used to characterize samples from beamline elements in CEBAF are presented.

Spradlin, Josh↗

Characterization of Oil Well Cement Performance during Early Hydration under Simulated Borehole Conditions

Summary Experiments on oil well cement (OWC) slurries were performed using the newly developed laboratory-scale wellbore simulation chamber (WSC). The WSC can simulate hydrostatic pressure reduction in the cemented annulus and possible gas migration under representative conditions. Forensic analysis shows that pressurized fluids can result in porous cement and gas channeling during cement slurry gelation. The effects of different factors on slurry pore pressure were also studied, including formation permeability, initial overburden pressure (OBP) representing the depth of interest, wellbore temperature, water/cement (w/c) ratio, cement composition, and the use of a calcium chloride (CaCl2)-based accelerator. By analyzing the temperature history of hydrating cement using degree of hydration, the evolution of cement hydration was characterized for slurry designs cured at different hydration rates. This provides the opportunity to parameterize the slurry designs and other important factors associated with wellbore conditions.

Engineering↗

Study of Chemical Changes in Uranium Oxyfluoride Particles Progress Report March - October 2009

Nuclear forensics relies on the analysis of certain sample characteristics to determine the origin and history of a nuclear material. In the specific case of uranium enrichment facilities, it is the release of trace amounts of uranium hexafluoride (UF{sub 6}) gas - used for the enrichment of uranium - that leaves a process-characteristic fingerprint. When UF{sub 6} gas interacts with atmospheric moisture, uranium oxyfluoride particles or particle agglomerates are formed with sizes ranging from several microns down to a few tens of nanometers. These particles are routinely collected by safeguards organizations, such as the International Atomic Energy Agency (IAEA), allowing them to verify whether a facility is compliant with its declarations. Spectrometric analysis of uranium particles from UF{sub 6} hydrolysis has revealed the presence of both particles that contain fluorine, and particles that do not. It is therefore assumed that uranium oxyfluoride is unstable, and decomposes to form uranium oxide. Understanding the rate of fluorine loss in uranium oxyfluoride particles, and the parameters that control it, may therefore contribute to placing boundaries on the particle's exposure time in the environment. Expressly for the purpose of this study, we prepared a set of uranium oxyfluoride particles at the Institute for Reference Materials and Measurements (EU-JRC-IRMM) from a static release of UF{sub 6} in a humid atmosphere. The majority of the samples was stored in controlled temperature, humidity and lighting conditions. Single particles were characterized by a suite of micro-analytical techniques, including NanoSIMS, micro-Raman spectrometry (MRS), scanning (SEM) and transmission (TEM) electron microscopy, energy-dispersive X-ray spectrometry (EDX) and focused ion beam (FIB). The small particle size was found to be the main analytical challenge. The relative amount of fluorine, as well as the particle chemical composition and morphology were determined at different stages in the ageing process, and immediately after preparation. This report summarizes our most recent findings for each of the analytical techniques listed above, and provides an outlook on what remains to be resolved. Additional spectroscopic and mass spectrometric measurements were carried out at Pacific Northwest National Laboratory, but are not included in this summary.

38 RADIATION CHEMISTRY, RADIOCHEMISTRY, AND NUCLEA↗

Fabrication of a 238Pu target

Precision neutron-induced reaction data are important for modeling the network of isotope production and destruction within a given diagnostic chain. This network modeling has many applications such as the design of advanced fuel cycle for reactors and the interpretation of radiochemical data related to the stockpile stewardship and nuclear forensics projects. Our current funded effort is to improve the neutron-induced reaction data on the short-lived actinides and the specific goal is to improve the neutron capture data on {sup 238}Pu with a half-life of 87.7 years. In this report, the fabrication of a {sup 238}Pu target for the proposed measurement using the DANCE array at LANL is described. The {sup 238}Pu target was fabricated from a sample enriched to 99.35%, acquired from ORNL. A total of 395 {micro}g was electroplated onto both sides of a 3 {micro}m thick Ti foil using a custom-made plating cell, shown in Fig 1. The target-material loaded Ti foil is sandwiched between two double-side aluminized mylar foils with a thickness of 1.4 {micro}m. The mylar foil is glued to a polyimide ring. This arrangement is shown partially in Fig. 2. The assembled target is then inserted into an aluminum container with a wall thickness of 0.76 mm, shown in Fig. 3. A derlin ring is used to keep the target assembly in place. The ends of this cylindrical container are vacuum-sealed by two covers with thin Kapton foils as windows for the beam entrance and exit. Shown in Fig. 4 is details of the arrangement. This target is used for phase I of the proposed measurement on {sup 238}Pu scheduled for Nov 2010 together with the DANCE array to address the safety issues raised by LANL. Shown in Fig. 5 is the preliminary results on the yield spectrum as a function of neutron incident energy with a gate on the total {gamma}-ray energy of equivalent Q value. Since no fission PPAC is employed, the distinction between the capture and fission events cannot be made, which is important for the higher neutron incident energy. However, it indicates that a cross section of less than one barn can be measured. The second phase of this experiment will be carried out in 2011 by assembling a PPAC with the {sup 238}Pu target to extend the measurement to higher neutron incident energies by distinguishing the capture from fission events. The fission cross section becomes dominant for neutron incident energies above 30 keV. This PPAC was developed in FY2010 under the NA22 funding and performed very well for the {sup 239}Pu and {sup 241}Pu measurements. A new {sup 238}Pu target will be fabricated for the phase II measurement using the same electroplating technique.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Rapid Automated Sample Preparation for Biological Assays

Our technology utilizes acoustic, thermal, and electric fields to separate out contaminants such as debris or pollen from environmental samples, lyse open cells, and extract the DNA from the lysate. The objective of the project is to optimize the system described for a forensic sample, and demonstrate its performance for integration with downstream assay platforms (e.g. MIT-LL's ANDE). We intend to increase the quantity of DNA recovered from the sample beyond the current {approx}80% achieved using solid phase extraction methods. Task 1: Develop and test an acoustic filter for cell extraction. Task 2: Develop and test lysis chip. Task 3: Develop and test DNA extraction chip. All chips have been fabricated based on the designs laid out in last month's report.

42 ENGINEERING↗

CBM.DIAGB.03.10.LLNL.007 Final Report

The purpose of this project was to construct a system for characterizing the threat potential of genomic sequences, specifically assembled draft genomes. New genomes are characterized by initially comparing them against already-sequenced genomes. If the new genome is determined to be from a high-threat species, detailed (forensic-level) characterization is done based on gene and SNP (Single Nucleotide Polymorphism) data comparisons with all other previously sequenced members of that high-threat species. New genomes are compared against a large set of known virulence and antibiotic-resistance genes and also compared against a large set of vectors that could be used for bacterial genetic engineering. Together, these analyses provide a comprehensive initial assessment of the most likely phylogenetic placement of a new genome, plus an assessment of the known-gene content and an indication of any possible bacterial genetic engineering utilizing vector-mediated techniques. This provides an initial threat potential summary based on high information content comparisons (e.g., thousands of genes, SNPs, and potential genetic engineering vectors) that can be used to guide subsequent operational response or more detailed laboratory characterization.

59 BASIC BIOLOGICAL SCIENCES↗

A simplified method for obtaining high-purity perchlorate from groundwater for isotope analyses.

Investigations into the occurrence and origin of perchlorate (ClO{sub 4}{sup -}) found in groundwater from across North America have been sparse until recent years, and there is mounting evidence that natural formation mechanisms are important. New opportunities for identifying groundwater perchlorate and its origin have arisen with the utilization of improved detection methods and sampling techniques. Additionally, application of the forensic potential of isotopic measurements has begun to elucidate sources, potential formation mechanisms and natural attenuation processes. Procedures developed appear to be amenable to enable high precision stable isotopic analyses, as well as lower precision AMS analyses of {sup 36}Cl. Immediate work is in analyzing perchlorate isotope standards and developing full analytical accuracy and uncertainty expectations. Field samples have also been collected, and will be analyzed when final qa/qc samples are deemed acceptable.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Report for Development of a Census Array and Evaluation of the Array to Detect Biothreat Agents and Environmental Samples for DHS

The objective of this project is to provide DHS a comprehensive evaluation of the current genomic technologies including genotyping, Taqman PCR, multiple locus variable tandem repeat analysis (MLVA), microarray and high-throughput DNA sequencing in the analysis of biothreat agents from complex environmental samples. This report focuses on the design, testing and results of samples on the Census Array. We designed a Census/Detection Array to detect all sequenced viruses (including phage), bacteria (eubacteria), and plasmids. Family-specific probes were selected for all sequenced viral and bacterial complete genomes, segments, and plasmids. Probes were designed to tolerate some sequence variation to enable detection of divergent species with homology to sequenced organisms, and to be unique relative to the human genome. A combination of 'detection' probes with high levels of conservation within a family plus 'census' probes targeting strain/isolate specific regions enabled detection and taxonomic classification from the level of family down to the strain. The array has wider coverage of bacterial and viral targets based on more recent sequence data and more probes per target than other microbial detection/discovery arrays in the literature. We tested the array with purified bacterial and viral DNA/RNA samples, artificial mixes of known bacterial/viral samples, spiked DNA against complex background including BW aerosol samples and soil samples, and environmental samples to evaluate the array's sensitivity and forensic capability. The data were analyzed using our novel maximum likelihood software. For most of the organisms tested, we have achieved at least species level discrimination.

59 BASIC BIOLOGICAL SCIENCES↗

DNA Extraction by Isotachophoresis in a Microfluidic Channel

Biological assays have many applications. For example, forensics personnel and medical professionals use these tests to diagnose diseases and track their progression or identify pathogens and the host response to them. One limitation of these tests, however, is that most of them target only one piece of the sample - such as bacterial DNA - and other components (e.g. host genomic DNA) get in the way, even though they may be useful for different tests. To address this problem, it would be useful to extract several different substances from a complex biological sample - such as blood - in an inexpensive and efficient manner. This summer, I worked with Maxim Shusteff at Lawrence Livermore National Lab on the Rapid Automated Sample Prep project. The goal of the project is to solve the aforementioned problem by creating a system that uses a series of different extraction methods to extract cells, bacteria, and DNA from a complex biological sample. Biological assays can then be run on purified output samples. In this device, an operator could input a complex sample such as blood or saliva, and would receive separate outputs of cells, bacteria, viruses, and DNA. I had the opportunity to work this summer with isotachophoresis (ITP), a technique that can be used to extract nucleic acids from a sample. This technique is intended to be the last stage of the purification device. Isotachophoresis separates particles based on different electrophoretic mobilities. This technique is convenient for out application because free solution DNA mobility is approximately equal for DNA longer than 300 base pairs in length. The sample of interest - in our case DNA - is fed into the chip with streams of leading electrolyte (LE) and trailing electrolyte (TE). When an electric field is applied, the species migrate based on their electrophoretic mobilities. Because the ions in the leading electrolyte have a high electrophoretic mobility, they race ahead of the slower sample and trailing electrolyte ions. Conversely, the trailing electrolyte ions have a slow electrophoretic mobility, so they lag behind the sample, thus trapping the species of interest between the LE and TE streams. In a typical isotachophoresis configuration, the electric field is applied in a direction parallel to the direction of flow. The species then form bands that stretch across the width of the channel. A major limitation of that approach is that only a finite amount of sample can be processed at once, and the sample must be processed in batches. For our purposes, a form of free-flow isotachophoresis is more convenient, where the DNA forms a band parallel to the edges of the channel. To achieve this, in our chip, the electric field is applied transversely. This creates a force perpendicular to the direction of flow, which causes the different ions to migrate across the flow direction. Because the mobility of the DNA is between the mobility of the leading and the trailing electrolyte, the DNA is focused in a tight band near the center of the channel. The stream of DNA can then be directed to a different output to produce a highly concentrated outlet stream without batch processing. One hurdle that must be overcome for successful ITP is isolating the electrochemical reactions that result from the application of high voltage for the actual process of isotachophoresis. The electrochemical reactions that occur around metal electrodes produce bubbles and pH changes that are detrimental to successful ITP. The design of the chips we use incorporates polyacrylamide gels to serve as electrodes along the central channel. For our design, the metal electrodes are located away from the chip, and high conductivity buffer streams carry the potential to the chip, functioning as a 'liquid electrode.' The stream then runs alongside a gel barrier. The gel electrode permits ion transfer while simultaneously isolating the separation chamber from any contaminants in the outer, 'liquid electrode' streams. The difference in potential from one side of the chip to the other creates an electric field. This field traverses the inner, separation channel, containing the leading electrolyte, the trailing electrolyte, and the sample of interest (DNA). To increase the ease of use of the chips, a newer chip design has been fabricated. This design has wire electrodes integrated on the chip, rather than elsewhere. To keep the pH changes and bubbling isolated from the separation channel, the chip contains deeper wells near the electrodes so that the flowing buffer can wash away any gases that form around the electrode. This design is significantly more compact because it eliminates the cumbersome electrode boxes. Eliminating the electrode boxes also decreases the required voltage, making the experiments safer. This happens because when the 'liquid electrode' streams travel through small diameter tubing, they lose much of their voltage due to the electrical resistance of the fluid in the tubing.

42 ENGINEERING↗

Hydrodynamic Modeling of Air Blast Propagation from the Humble Redwood Chemical High Explosive Detonations Using GEODYN

Two-dimensional axisymmetric hydrodynamic models were developed using GEODYN to simulate the propagation of air blasts resulting from a series of high explosive detonations conducted at Kirtland Air Force Base in August and September of 2007. Dubbed Humble Redwood I (HR-1), these near-surface chemical high explosive detonations consisted of seven shots of varying height or depth of burst. Each shot was simulated numerically using GEODYN. An adaptive mesh refinement scheme based on air pressure gradients was employed such that the mesh refinement tracked the advancing shock front where sharp discontinuities existed in the state variables, but allowed the mesh to sufficiently relax behind the shock front for runtime efficiency. Comparisons of overpressure, sound speed, and positive phase impulse from the GEODYN simulations were made to the recorded data taken from each HR-1 shot. Where the detonations occurred above ground or were shallowly buried (no deeper than 1 m), the GEODYN model was able to simulate the sound speeds, peak overpressures, and positive phase impulses to within approximately 1%, 23%, and 6%, respectively, of the actual recorded data, supporting the use of numerical simulation of the air blast as a forensic tool in determining the yield of an otherwise unknown explosion.

42 ENGINEERING↗

Annual Report on the Activities and Publications of the DHS-DNDO-NTNFC-Sponsored Post-doctoral Fellow at Los Alamos National Laboratory

This report is a summary update on the projects Magen Coleman is working on as a postdoctoral fellow funded by DHS. These research projects are designed to explore different radioanalytical facets of nuclear forensics, including destructive and non-destructive analysis, chronometry measurements and modeling, spectral analysis, and actinide separations. Method development and instrumental analysis are integral parts of these projects, along with strong radiochemical knowledge. The current projects discussed here include method development for 232U analysis, the assembly of a gamma mapping system, the testing of neutron bubble detectors for neutron spectrometry, and continued work on the chronometry of plutonium and uranium materials. This report documents the work that has been performed since April 2011.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

A Report on the Activities, Publications, and Pending Research of DHS/DOD Sponsored Post-doctoral Research Associate at Los Alamos National Laboratory

Since beginning at Los Alamos National Laboratory in February of 2012, I have been working as a DHS./DNDO Postdoctoral Research Associate under the mentorship of Lav Tandon and Khalil Spencer (NA-22 and mass spectrometry). The focus of my efforts, in addition to pursuing needed training and qualifications, has been the application of various instrumental approaches (e.g. Thermal Ionization Mass Spectrometry; TIMS) to a range of systems of interest in materials characterization and nuclear forensics. Research to be pursued in the coming months shall include the continued use of such approaches to advance current methods for: modified total evaporation, monitoring critical minor isotope systems, and chronometry. Each of the above points will be discussed.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Uncertainty Quantification of Hypothesis Testing for the Integrated Knowledge Engine

The Integrated Knowledge Engine (IKE) is a tool of Bayesian analysis, based on Bayesian Belief Networks or Bayesian networks for short. A Bayesian network is a graphical model (directed acyclic graph) that allows representing the probabilistic structure of many variables assuming a localized type of dependency called the Markov property. The Markov property in this instance makes any node or random variable to be independent of any non-descendant node given information about its parent. A direct consequence of this property is that it is relatively easy to incorporate new evidence and derive the appropriate consequences, which in general is not an easy or feasible task. Typically we use Bayesian networks as predictive models for a small subset of the variables, either the leave nodes or the root nodes. In IKE, since most applications deal with diagnostics, we are interested in predicting the likelihood of the root nodes given new observations on any of the children nodes. The root nodes represent the various possible outcomes of the analysis, and an important problem is to determine when we have gathered enough evidence to lean toward one of these particular outcomes. This document presents criteria to decide when the evidence gathered is sufficient to draw a particular conclusion or decide in favor of a particular outcome by quantifying the uncertainty in the conclusions that are drawn from the data. The material in this document is organized as follows: Section 2 presents briefly a forensics Bayesian network, and we explore evaluating the information provided by new evidence by looking first at the posterior distribution of the nodes of interest, and then at the corresponding posterior odds ratios. Section 3 presents a third alternative: Bayes Factors. In section 4 we finalize by showing the relation between the posterior odds ratios and Bayes factors and showing examples these cases, and in section 5 we conclude by providing clear guidelines of how to use these for the type of Bayesian networks used in IKE.

97 MATHEMATICS AND COMPUTING↗