Why did a vehicle stop? A methodology for detection and classification of stops in vehicle trajectories

ABSTRACT

Trajectory data mining is a lively research field in the domain of spatio-temporal data mining. Trajectory pattern mining comprises a set of specific pattern mining methods, which are applied as consecutive steps on a trajectory with the goal to extract and classify re-occurring spatio-temporal patterns. Despite the common nature and frequent usage of such methods by the GIScience community, a methodological approach is missing so far, especially when it comes to the use of machine learning-based classification methods. The current work closes this gap by proposing and evaluating a machine learning-based 3-steps trajectory data mining methodology using the detection and classification of stop points in vehicle trajectories as example. The work describes in detail the applied methodologies with respect to the three mining steps ‘stop detection’, ‘feature extraction’ and ‘classification in traffic-relevant and non-traffic-relevant stops’ and evaluates six machine learning-based classification algorithms using a real-world dataset of 15,498 vehicle trajectories with 5,899 detected stops (thereof 2,032 manually classified). Due to its exemplary nature, the presented methodology is suited to act as blueprint for similar trajectory data mining problems.     

Estimation of crack opening from a two‐dimensional continuum‐based finite element computation

Damage models are capable of representing crack initiation and mimicking crack propagation within a continuum framework. Thus, in principle, they do not describe crack openings. In durability analyses of concrete structures however, transfer properties are a key issue controlled by crack propagation and crack opening. We extend here a one‐dimensional approach for estimating a crack opening from a continuum‐based finite element calculation to two‐dimensional cases. The technique operates in the case of mode I cracking described in a continuum setting by a nonlocal isotropic damage model. We used the global tracking method to compute the idealized crack location as a post‐treatment procedure. The original one‐dimensional problem devised in Dufour et al. [4] is recovered as profiles of deformation orthogonal to the idealized crack direction are computed. An estimate of the crack opening and an error indicator are computed by comparing finite element deformation profiles and theoretical profiles corresponding to a displacement discontinuity. Two estimates have been considered: In the strong approach, the maxima of the profiles are assumed to be equal; in the weak approach, the integrals of each profile are set equal. Two‐dimensional numerical calculations show that the weak estimates perform better than do the strong ones. Error indicators, defined as the distance between the numerical and theoretical profiles, are less than a few percentages. In the case of a three‐point bending, test results are in good agreement with experimental data, with an error lower than 10% for widely opened crack (> 40µm). Copyright © 2011 John Wiley & Sons, Ltd.     

Landscape care paradoxes: Swedish landscape care arrangements in a European context

Contemporary European agriculture has a number of additional aims beside of food production, such as safeguarding environmental services and conservation values. Substantial efforts at official levels are aimed towards sustainable development but also towards maintaining values of what may be termed vanishing landscapes. Selected areas and landscape features are set aside for protection or restoration. Individual efforts of this type have a long history in Sweden, and the issue has recently received increased attention, primarily due to more ambitious government goals concerning biodiversity conservation and Sweden’s ratification of the European Landscape Convention. This has resulted in an increased scientific and official interest in vanishing values in the rural landscape, where parts of Eastern Europe, such as the Maramures district in Romania, have been used as model examples of land use regimes which in the past was common in Sweden. In this context, the dilemma of romanticizing peasants’ use of land is highlighted and discussed more than has hitherto been done. This paper sheds light on some paradoxes inherent in official policies in relation to land use practices concerning the management of rural landscapes in Sweden, and relates the Swedish situation to a contrasting example of landscape practice in Romania. We discuss the concept of landscape care in relation to the construction and perception of landscape values and valuable landscapes through the lenses of rural realities and official policies. When Swedish authorities engage in the promotion of landscape care, they tend to work with slices of land, specific predefined values and individual farmers, and they often disregard the need to treat the landscape as a socio-ecological complex dynamic in space and time. We discuss how environmental policy generally could be improved through the adoption of a more inclusive and flexible approach towards aiding the different aims inherent in multifunctional rural landscapes.     

Seabird aggregative patterns: A new tool for offshore wind energy risk assessment

The emerging development of offshore wind energy has raised public concern over its impact on seabird communities. There is a need for an adequate methodology to determine its potential impacts on seabirds. Environmental Impact Assessments (EIAs) are mostly relying on a succession of plain density maps without integrated interpretation of seabird spatio-temporal variability. Using Taylor’s power law coupled with mixed effect models, the spatio-temporal variability of species’ distributions can be synthesized in a measure of the aggregation levels of individuals over time and space. Applying the method to a seabird aerial survey in the Ebro Delta, NW Mediterranean Sea, we were able to make an explicit distinction between transitional and feeding areas to define and map the potential impacts of an offshore wind farm project. We use the Ebro Delta study case to discuss the advantages of potential impacts maps over density maps, as well as to illustrate how these potential impact maps can be applied to inform on concern levels, optimal EIA design and monitoring in the assessment of local offshore wind energy projects.     

A new patchwork simulation method with control of the local-mean histogram

We present a new stochastic simulation method that builds two-dimensional images by assembling together square image pieces called blocks. The blocks are taken from a reference image. Our method, called patchwork simulation method (PSM), enforces pattern continuity in the image. Moreover, PSM allows to control the image local-mean histogram. This histogram bin-frequencies can be set to user-defined target values that may differ from the reference image local-mean histogram. This flexibility enhances the PSM generality by enlarging the set of all possible simulations. The local-mean histogram control is achieved by adjusting suitably the transition probabilities that associate a new block to an existing neighborhood in the partly simulated image. For several types of synthetic images and one polymer blend image, we show that PSM reproduces faithfully the reference image visual appearance (i.e. patterns are correctly shaped) and that simulated images are statistically compatible with the target local-mean histogram. Moreover, we show that our method has the ability to produce simulations that respect conditional hard data as well as a target local-mean histogram.     

Detecting stratigraphical issues using direct radiocarbon dating from small-mammal remains

Frequently preserved in archaeological and palaeontological sites, the tiny size of small-mammal remains favours percolations into underlying layers along stratigraphic sequences. This is one of the various post-depositional processes that may affect the integrity of the original deposits and therefore the subsequent scientific interpretations. Recent developments in sample preparation offer the possibility of detecting intrusive episodes through the absolute dating of minute amounts of bone (down to 10 mg), meaning that isolated elements (such as mandibles in this case) are sufficient to obtain reliable radiocarbon dates if collagen is moderately to well preserved. The radiocarbon dates obtained here for small-mammal bones (recovered from pre-Bølling to recent deposits) and their comparison with previous dates obtained from other sources (large-mammal bones, charcoal, botanical samples, etc.), with different protocols and instruments, illustrate the potential of small-mammal dating to reveal (and eventually contribute a solution to) stratigraphical issues in different archaeological contexts.     

Distribution and chronology of Pleistocene permafrost features in France: Database and first results

Numerous periglacial features (polygons, nets, soil stripes, ice‐wedge pseudomorphs and sand‐wedge casts, involutions) have been recorded in France by examining bibliographical sources and aerial photographs. These data show that a large part of France was affected by permafrost during the Pleistocene and only the southern Aquitaine Basin and Languedoc seem to have been beyond its maximum extent. The first OSL ages obtained from the aeolian infill of wedge structures indicate that at least two phases of thermal contraction cracking occurred in southwestern France between ～25 and 36 ka. Chronostratigraphical data from loess in northern France indicate that these episodes correspond to the formation of ice‐wedge networks associated with tundra gleys. In the latter region, two additional permafrost episodes probably occurred during the Last Glacial, the older one corresponding to the end of Marine Isotope Stage (MIS) 4 around 60 ka and the more recent one to MIS 2 around 19–16 ka. Although stratigraphical data indicate that these episodes were relatively short (about one millennium), relict permafrost may have existed for longer periods in northern France.     

Early evaluation of coastal nutrient over-enrichment: new procedures and indicators

Recent studies have provided compelling evidence for an accelerated anthropogenic impact on coastal systems, resulting in intense inputs of materials and nutrients from the continent. This has led scientists and policymakers to encourage the implementation of monitoring programmes, which have resulted in the multiplicity of datasets. However surprisingly, only a few attempts have been made to couple observations with statistical and mathematical tools to detect, as soon as the data become available perturbations in coastal systems. Here, we propose new mathematical procedures to evaluate the state of a system, based on the building of relative reference state and indicators of nutrient over-enrichment. The techniques were tested in some French coastal systems using data from the programme SOMLIT. Applied to this dataset, the multivariate procedures rapidly identified and evaluated anthropogenic nutrient anomalies from the continent on three sites (Wimereux, Roscoff and Villefranche-sur-Mer) from 1997 onwards.     

Melt evolution in subarc mantle: evidence from heating experiments on spinel-hosted melt inclusions in peridotite xenoliths from the andesitic Avacha volcano (Kamchatka,Russia)

Glass-bearing inclusions hosted by Cr-spinel in harzburgite xenoliths from Avacha are grouped based on homogenization temperatures and daughter minerals into high-T (1,200°C; opx + cpx), intermediate (900–1,100°C; cpx ± amph), and low-T (900°C; amph) and are commonly accompanied by larger “melt pockets”. Unlike previous work on unheated inclusions and interstitial glass in xenoliths from Kamchatka, the homogenized glass compositions in this study are not affected by low-pressure melt fractionation during transport and cooling or by interaction with host magma. Primary melt compositions constrained for each inclusion type differ in major and trace element abundances and were formed by different events, but all are silica saturated, Ca-rich, and K-poor, with enrichments in LREE, Sr, Rb, and Ba and negative Nb anomalies. These melts are inferred to have been formed with participation of fluids produced by dehydration of slab materials. The high-T inclusions trapped liquids produced by ancient high-degree, fluid-induced melting in the mantle wedge. The low-T inclusions are related to percolation of low-T melts or hydrous fluids in arc mantle lithosphere. Melt pockets arise from localized heating and fluid-assisted melting induced by rising magmas shortly before the entrapment of the xenoliths. The “high-T” melt inclusions in Avacha xenoliths are unique in preserving evidence of ancient, high-T melting events in arc mantle, whereas the published data appear to characterize pre-eruption enrichment events.