Anomalous Observations Research Articles

Environmental effects on the interferometric repeat-pass coherence of forests

The variability of C-band one-day tandem coherence measurements of forest is analyzed both statistically and with the aid of a soil-vegetation-atmosphere transfer model coupled to models for dielectric, backscattering coefficient, and coherence. Coherence depends strongly on both moisture and wind conditions. The dependence between coherence and moisture fluctuations in the soil and canopy is simulated using a simplified coherence model neglecting canopy coherence; trends are similar to those observed in measured coherence, but the range of variability observed in the satellite data is underestimated. Improved consistency with the data is achieved by using empirical estimates of canopy coherence derived from mature forest, but still with underestimated variability. The canopy coherence values cannot be predicted from the environmental model, and anomalous observations prevent the formulation of a reliable empirical predictor based on weather data; this could potentially be improved if higher time resolution wind data were made available. The study also suggests that canopy dielectric can vary rapidly in daylight, and that current models for dielectric may be inadequate to capture the effect of this behavior on coherence.

Towards an Information-Rich Process Development. Part I: Interfacing Experimentation with Qualitative/Semiquantitative Modelling

In today's competitive environment, the low-tonnage chemicals industry (fine and effect chemicals) is under increasing pressure to deliver more products to the market quicker and more cost effectively. Process development is a critical phase in bringing a fine chemical product from discovery to the market and requires the provision of an efficient framework for the capture of process knowledge and understanding. This is traditionally done by somewhat indirect meanseither through building understanding in individuals or the construction of a mathematical or physical model of the process. Such approaches suffer from two defects: first they do not collect understanding in a form that is readily shared by all the technologists involved, and second they do not collect understanding early enough in design. For example, the data required for the detailed modelling of a process is often only available well after the process configuration is fixed and properties can be measured under relevant conditions. Considering a liquid biphase reaction system, this paper demonstrates the interaction of targeted experimentation and qualitative/semiquantitative reaction evaluation to allow capture, testing, and sharing of understanding much earlier in design. The resultant qualitative predictions are tested against experimental observations to identify aspects where the qualitative model needs refinement. Recommendations for using more rigorous process analysis tools to explain these so-called anomalous observations are thus made.

Data mining methods for anomaly detection KDD-2005 workshop report

For many applications, data mining systems are required to detect anomalous (abnormal, unmodeled, or unexpected) observations. This has so far proven to be a difficult challenge because anomalies are usually considered to be "non-normal" observations, where "normality" is typically defined by very complex concepts. Because of these and other reasons, there are no standard and principled approaches for anomaly detection, yet, and the data mining processes that have led to successful solutions include most of the times ad-hoc (algorithmic, design, and implementation) decisions that incorporate prior or commonsense knowledge about the tasks that are addressed.Consequently, we considered that it would be beneficial for both researchers and practitioners interested in anomaly detection and data mining, to organize workshop that would bring together people interested in this topic. We considered that the International Conference on Knowledge Discovery and Data Mining would be a good venue for such a workshop because of the diversity of interests, backgrounds, and problems that motivate people to attend the conference.This paper describes the workshop on "Data Mining Methods for Anomaly Detection" - a one day event held in conjunction with KDD-2005 in Chicago, on August 21, 2005.

Reply to "Comment on 'Lisbon 1755: A Case of Triggered Onshore Rupture?' by Susana P. Vilanova, Catarina F. Nunes, and Joao F. B. D. Fonseca," by L. Matias, A. Ribeiro, M. A. Baptista, N. Zitellini, J. Cabral, P. Terrinha, P. Teves-Costa, and J. M. Miranda

In Vilanova et al. (2003), referred to in the sequence as “the article,” we suggested that several observations pertaining to the historical accounts of the 1755 Lisbon earthquake are better explained assuming the occurrence of a triggered onshore rupture near Lisbon, a few minutes after the main earthquake. Such observations include the distribution of intensities and the high values in and around Lisbon, the occurrence of an early tsunami-like wave inside the Tagus lagoon prior to the arrival of the external tsunami, a report of ground deformation near Lisbon, cartographic evidence of changes in the course of the Tagus river, and the distribution of destructive aftershocks. The proposed location of the triggered rupture is the well-established seismogenic Lower Tagus Valley (ltv) fault (Cabral, 1995). There is reasonable consensus that the Lisbon earthquake was composed of three shocks in quick succession (Reid, 1914). The main innovation introduced in the article is the thesis that the location of one of the subevents was close to Lisbon, at a significant distance from the first shock. The distance between the classical Gorringe Bank epicenter (Machado, 1966) and the ltv fault is about 350 km, but this value can be significantly reduced by taking into account the finite length of the offshore rupture or by adopting a source closer to the coast as proposed in recent literature (Zitellini et al. , 1999). In view of the progress in the understanding of fault interaction through dynamic or static stress changes, we believe that our hypothesis is plausible, besides it is effective as a way of explaining the observations. L. Matias and colleagues, next referred to as “the comment,” claim that our conclusion is speculative and unsubstantiated, arguing that each anomalous observation reported in the article can be explained in a way that is different from …

Fractally Invariant Distributions and an Application in Geochemical Exploration

Fractal modelling has been applied extensively as a means of characterizing the spatial distribution of geological phenomena that display self-similarity at differing scales of measurement. A fractal distribution exists where the number of objects exhibiting values larger than a specified magnitude displays a power-law dependence on that magnitude, and where this relationship is scale-invariant. This paper shows that a number of distributions, including power-function, Pareto, log-normal and Zipf, display fractal properties under certain conditions and that this may be used as the mathematical basis for developing fractal models for data exhibiting such distributions. Population limits, derived from fractal modelling using a summation method, are compared with those derived from more conventional probability plot modelling of stream sediment geochemical data from north-eastern New South Wales. Despite some degree of subjectivity in determining the number of populations to use in the models, both the fractal and probability plot modelling have assisted in isolating anomalous observations in the geochemical data related to the occurrence of mineralisation or lithological differences between sub-catchments. Thresholds for the main background populations determined by the fractal model are similar to those established using probability plot modelling, however the summation method displays less capacity to separate out anomalous populations, especially where such populations display extensive overlap. This suggests, in the geochemical data example provided, that subtle differences in the population parameters may not significantly alter the fractal dimension.

Spatiotemporal evolution of a transient slip event on the San Andreas fault near Parkfield, California

In 1993 several baselines of the two‐color electronic distance meter (EDM) network at Parkfield, California, deviated from their long‐term rates, coincident with anomalous observations from nearby strain meters and a creep meter, as well as an increase in microseismicity. Between October 1992 and December 1994, three M ∼ 4.5 earthquakes occurred beneath Middle Mountain, near the hypocenter of the 1934 and 1966 Parkfield M6 events. We analyzed the two‐color EDM data using a Kalman‐filtering based technique to image the spatiotemporal evolution of slip on the fault at Parkfield between the mid‐1980s and 2003. This method accounts for localized random walk motion of the geodetic monuments and a prominent seasonal signal that affects many baselines. We find that a slip rate increase occurred between January 1993 and July 1996 on the upper 8 km of the fault near Middle Mountain. The peak estimated slip rate during this time was 49 mm/yr, which exceeds the long‐term geologic rate of ∼35 mm/yr. The slip rate evolution appears episodic, with an initial modest increase after the M4.3 earthquake and a much larger jump following the shallower M4.7 event in December 1994. This temporal correlation between inferred slip and seismicity suggests that the moderate earthquakes triggered the aseismic fault slip. The EDM data cannot resolve whether transient slip propagated across the nucleation zone of the 1934 and 1966 M6 Parkfield earthquakes. However, transient slip and its associated stress release in the hypocentral area of previous Parkfield events is consistent with the nucleation of the 2004 M6 Parkfield earthquake elsewhere on the fault.

Sensitivity of Test for Overdispersion in Poisson Regression

Overdispersion or extra-Poisson variation is very common for count data. This phenomenon arises when the variability of the counts greatly exceeds the mean under the Poisson assumption, resulting in substantial bias for the parameter estimates. To detect whether count data are overdispersed in the Poisson regression setting, various tests have been proposed and among them, the score tests derived by Dean (1992) are popular and easy to implement. However, such tests can be sensitive to anomalous or extreme observations. In this paper, diagnostic measures are proposed for assessing the sensitivity of Dean's score test for overdispersion in Poisson regression. Applications to the well-known fabric faults and Ames salmonella assay data sets illustrate the usefulness of the diagnostics in analyzing overdispersed count data.

Observation of selective charge separation following strong-field single ionization

In this paper, we report on an apparently anomalous observation in the dissociation channels of singly ionized nitric oxide (NO) molecules. We observe that the NO molecules, when oriented perpendicular to the laser polarization, appear to preferentially dissociate into the ${\mathrm{N}}^{+}+\mathrm{O}$ channel rather than the $\mathrm{N}+{\mathrm{O}}^{+}$ channel, despite the fact that the latter channel is more energetically favorable. A careful examination of the unique electronic structure of the NO molecule suggests that this observation is most likely due to the influence of the detailed electronic structure of NO on its strong-field ionization dynamics, which leads $\mathrm{N}{\mathrm{O}}^{+}$ created in strong fields to dissociate differently from that in weak fields. This study reinforces our understanding of the fundamental differences between ionization dynamics in strong fields versus weak fields.

A Survey of Outlier Detection Methodologies

Outlier detection has been used for centuries to detect and, where appropriate, remove anomalous observations from data. Outliers arise due to mechanical faults, changes in system behaviour, fraudulent behaviour, human error, instrument error or simply through natural deviations in populations. Their detection can identify system faults and fraud before they escalate with potentially catastrophic consequences. It can identify errors and remove their contaminating effect on the data set and as such to purify the data for processing. The original outlier detection methods were arbitrary but now, principled and systematic techniques are used, drawn from the full gamut of Computer Science and Statistics. In this paper, we introduce a survey of contemporary techniques for outlier detection. We identify their respective motivations and distinguish their advantages and disadvantages in a comparative review.

A Survey of Outlier Detection Methodologies

Outlier detection has been used for centuries to detect and, where appropriate, remove anomalous observations from data. Outliers arise due to mechanical faults, changes in system behaviour, fraudu...

Abductive inference in defeasible reasoning: a model for research programmes

Abduction is regarded as the inference process that goes from observations to explanations within a more general context or theoretical framework. There are good characterizations of abduction of surprising observations in monotonic theories. Also, in normal logic programs there are a tight relation among SLDNF and the abduction of negative literals. However, a problem that faces abduction is the explanation of anomalous observations, i.e., observations that are contradictory with respect to the current theory. For this reason, in this work we will consider the problem of embedding abduction of surprising and anomalous observations in defeasible (nonmonotonic) theories. We discuss some issues arising the pragmatic acceptance of abductive inferences in defeasible theories, and how to accommodate anomalous observations and characterize all the possible outcomes that a defeasible theory may face when confronted with new evidence. We explore the use of partial structures approach as a semantic foundation for our system. Finally, we discuss an application of our system as a formal device for representing the methodology of scientific research programmes. In this representation, a programme is regarded as a defeasible theory that draws predictions. When confronted with surprising or anomalous observations, the programme protects itself by means of heuristic procedures, which are represented in our system as abductive inference procedures.

Non‐parametric statistical methods for multivariate calibration model selection and comparison

AbstractModel selection is an important issue when constructing multivariate calibration models using methods based on latent variables (e.g. partial least squares regression and principal component regression). It is important to select an appropriate number of latent variables to build an accurate and precise calibration model. Inclusion of too few latent variables can result in a model that is inaccurate over the complete space of interest. Inclusion of too many latent variables can result in a model that produces noisy predictions through incorporation of low‐order latent variables that have little or no predictive value. Commonly used metrics for selecting the number of latent variables are based on the predicted error sum of squares (PRESS) obtained via cross‐validation. In this paper a new approach for selecting the number of latent variables is proposed. In this new approach the prediction errors of individual observations (obtained from cross‐validation) are compared across models incorporating varying numbers of latent variables. Based on these comparisons, non‐parametric statistical methods are used to select the simplest model (least number of latent variables) that provides prediction quality that is indistinguishable from that provided by more complex models. Unlike methods based on PRESS, this new approach is robust to the effects of anomalous observations. More generally, the same approach can be used to compare the performance of any models that are applied to the same data set where reference values are available. The proposed methodology is illustrated with an industrial example involving the prediction of gasoline octane numbers from near‐infrared spectra. Published in 2004 by John Wiley & Sons, Ltd.

Altimeter estimates of anomalous transports into the northern California Current during 2000–2002

Surface transports into the California Current are calculated from TOPEX/POSEIDON altimeter surface height slopes during the 9.5 year period from October 1992–May 2002. These quantify the anomalous onshore and southward displacements of the water column during the 2000–2002 period, which had been hypothesized by others based on anomalous water properties observed in mid‐2002 off Oregon. Anomalous eastward transport into the boundary between 52°–54°N occurred first (mid‐2000 to mid‐2001) followed by eastward transport into the boundary between 50°–52°N in 2001. Equatorward transports during 2001 and 2002 resulted in approximately 800 km of anomalous southward displacements during the 1.5 years prior to the anomalous observations off Oregon. This sequence suggests that both onshore and southward displacement anomalies off British Columbia and the Pacific Northwest contributed to the subarctic characteristics of the water observed off Oregon.

Characterization and disequilibrium cases of radioelements in some Syrian hot waters

The radioactive and physical characteristics of Syrian hot waters were studied. The concentration of uranium, its daughters and the 234U/238U-activity ratios in the hot waters were measured by gamma and alpha ray spectroscopy. The 234U/238U-activity ratios showed that uranium in Syrian hot waters are in an equilibrium state. Some anomalous observations of these ratios were explained by former leaching of the uranium from Cretaceous phosphates by underground water. Also, the results showed that the temperature of the studied hot waters was independent of the contained radioelements. The temperature of hot waters is explained by the passing of these waters near the hot spot or by the relationship between the geological deposit of the aquifer rocks and the waters rather than the contents of the radioelements.

Letter: Cosmological Anomalies and Exotic Smoothness Structures

It seems to be generally accepted that apparently anomalous cosmological observations, such as accelerating expansion, etc., necessarily are inconsistent with standard general relativity and standard matter sources. Following the suggestions of Sladkowski, we point out that in addition to exotic theories and exotic matter there is another possibility. We refer to exotic differential structures on ℝ4 which could be the source of the observed anomalies without changing the Einstein equations or introducing strange forms of matter.

Leader fields due to a lightning channel with an inclined upper section

It is shown theoretically the dependence of the leader field on the angle of orientation with respect to a perfectly conducting horizontal plane of the uppermost part of the lightning leader. The results were able to discern significantly large differences in the static field variation of leader channels with the same length but a certain channel segment is oriented at different angles. The outcome of our calculations consistently explains the scatter of the total leader field observed in previous studies. Otherwise, one has to assume unrealistic charge source heights or unreasonable charge densities to calculate matching values for some of the total leader fields and total leader field to return stroke filed ratios, labeled as anomalous observations in the literature. In some cases, irrespective of the charge source height and the charge density, one cannot find a suitable fit with the straight channel model.

A fast method for robust principal components with applications to chemometrics

When faced with high-dimensional data, one often uses principal component analysis (PCA) for dimension reduction. Classical PCA constructs a set of uncorrelated variables, which correspond to eigenvectors of the sample covariance matrix. However, it is well-known that this covariance matrix is strongly affected by anomalous observations. It is therefore necessary to apply robust methods that are resistant to possible outliers. Li and Chen [J. Am. Stat. Assoc. 80 (1985) 759] proposed a solution based on projection pursuit (PP). The idea is to search for the direction in which the projected observations have the largest robust scale. In subsequent steps, each new direction is constrained to be orthogonal to all previous directions. This method is very well suited for high-dimensional data, even when the number of variables p is higher than the number of observations n. However, the algorithm of Li and Chen has a high computational cost. In the references [C. Croux, A. Ruiz-Gazen, in COMPSTAT: Proceedings in Computational Statistics 1996, Physica-Verlag, Heidelberg, 1996, pp. 211–217; C. Croux and A. Ruiz-Gazen, High Breakdown Estimators for Principal Components: the Projection-Pursuit Approach Revisited, 2000, submitted for publication.], a computationally much more attractive method is presented, but in high dimensions (large p) it has a numerical accuracy problem and still consumes much computation time. In this paper, we construct a faster two-step algorithm that is more stable numerically. The new algorithm is illustrated on a data set with four dimensions and on two chemometrical data sets with 1200 and 600 dimensions.

Characterization of dichromat and trichromat observers using a PC-based anomaloscope

A system is presented for characterizing anomalous colour observers according to the type of colour vision deficiency that they suffer. The system is based on the simulation of a Pickford–Nicholson type anomaloscope using a cathode ray tube and a personal computer. By means of this method, it has been possible to determine the type of deficiency of the red–green chromatic channel presented by various anomalous observers.

Active chlorine in the remote marine boundary layer: Modeling anomalous measurements of δ13C in methane

Measurements of δ13C in methane in the marine boundary layer (MBL) of the extratropical Southern Hemisphere imply a kinetic isotope fractionation much larger than would be expected if the hydroxyl radical were the only tropospheric methane sink. We use a simple chemical box model to show that the assumption of a MBL active chlorine (Cl•) sink can explain these anomalous observations provided there is a seasonal cycle in the Cl• sink with a summer‐winter concentration difference ∼ 6 · 10³ cm−3. The required summer maximum and yearly mean Cl• concentrations are plausible, and imply a global Cl• sink strength for methane of < 5 Tg yr−1. Choice of a Cl• sink seasonal cycle linked to the nonsinusoidal dimethyl sulfide seasonal cycle gives the observed fractionation with a smaller yearly mean Cl• concentration than equivalent sinusoidal Cl• cycles.

INTERPLAY OF ANTIFERROMAGNETISM AND SUPERCONDUCTIVITY IN RARE EARTH NICKEL BOROCARBIDES

A review of some of the anomalous properties observed in rare earth nickel borocarbide superconductors is presented. A model proposed earlier for heavy fermion superconductors is adopted to the rare earth nickel borocarbides. Even though the model is solved in the mean field approximation, it provides qualitative explanation to many of the anomalous experimental observations.