Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
The Ukrainian shield hosts two Palaeoproterozoic anorthosite-mangerite-charnockite-granite (AMCG) complexes (the Korosten and Korsun-Novomyrhorod complexes) that intruded Palaeoproterozoic continental crust in north-western and central parts of the shield, respectively. We report results of U-Pb zircon and baddeleyite dating of 16 samples from the Korosten plutonic complex (KPC), and 6 samples from the Korsun-Novomyrhorod plutonic complex (KNPC). Fifteen zircon samples from both complexes were also analysed for Hf isotopes. These new, together with previously published data indicate that the formation of the KPC started at c. 1815 Ma and continued until 1743 Ma with two main phases of magma emplacement at 1800–1780 and 1770–1758 Ma. Each of the main phases of magmatic activity included both basic and silicic members. The emplacement history of the KNPC is different from that of the KPC. The vast majority of the KNPC basic and silicic rocks were emplaced between c. 1757 and 1750 Ma; the youngest stages of the complex are represented by monzonites and syenites that were formed between 1748 and 1744 Ma. Both Ukrainian AMCG complexes are closely associated in space and time with mantle-derived mafic and ultramafic dykes. The Hf isotope ratios in the zircons indicate a predominantly crustal source for the initial melts with some input of juvenile Hf from mantle-derived tholeiite melts.The preferred model for the formation of the Ukrainian AMCG complexes involves the emplacement of large volumes of hot mantle-derived tholeiitic magma into the lower crust. This resulted in partial melting of mafic lower-crustal material, mixing of lower crustal and tholeiitic melts, and formation of ferromonzodioritic magmas. Further fractional crystallization of the ferromonzodioritic melts produced the spectrum of basic rocks in the AMCG complexes. Emplacement of the ferromonzodioritic and tholeiitic melts into the middle crust and their partial crystallization caused abundant melting of the ambient crust and formation of the large volumes of granitic rocks present in the complexes.
ISSN: 0301-9268

hal-01622715
https://hal.archives-ouvertes.fr/hal-01622715
DOI : 10.1016/j.precamres.2017.02.009

Résumé : En 1999, le risque écologique encouru par les écosystèmes côtiers soumis à la prolifération de la crépidule, Crepidula fornicata, a été pleinement reconnu par la communauté scientifique nationale et internationale. Au niveau national, le Programme National Environnement Côtier (PNEC) et le programme national LITEAU ont tous deux identifié des actions thématiques prioritaires sur la dynamique de prolifération de la crépidule et les interactions crépidule/bivalves exploités. Les actions du LEMAR (UMR 6539 CNRS) sur cette thématique s’inscrivent dans deux projets : 1 - La crépidule : identifier les mécanismes de sa prolifération et caractériser ses effets sur le milieu pour envisager sa gestion. » (LITEAU) 2 - Recrutement, milieu physique et interactions d'espèces en environnement côtier : cas des mollusques » (PNEC ART2) En rade de Brest, l’accent est mis sur l’interaction des cycles biologiques de la crépidule et de la coquille Saint-Jacques. Les objectifs poursuivis en rade de Brest s’articulent en 4 points, dans le cadre de LITEAU : 1- Dynamique de colonisation de la crépidule en rade de Brest : acquisition des paramètres de dynamique de population (caractéristiques du cycle reproducteur, déterminisme du recrutement, dispersion, croissance, sensibilité aux efflorescences toxiques...) ; 2- Impact de cette prolifération sur l’évolution de la nature sédimentaire des zones colonisées et l’enrichissement en matière organique (par le biais de l’activité alimentaire et de la production de biodépôts) ; 3- Modifications engendrées sur les peuplements benthiques dans leur globalité (sites ateliers). La strate à crépidules, au sens biosédimentaire et écologique, verra sa définition affinée et sa dynamique analysée en vue d'obtenir un modèle prévisionnel des évolutions possibles en matière de biodiversité ; 4- Rôle de la crépidule dans le contrôle de la production primaire phytoplanctonique (par le biais des activités de filtration et de production de biodépôts).

Droits : info:eu-repo/semantics/openAccess
http://archimer.ifremer.fr/doc/00032/14292/11570.pdf
http://archimer.ifremer.fr/doc/00032/14292/

Éditeur(s) : HAL CCSD Public Library of Science
Résumé : International audience
ISSN: 1935-2727

hal-01447691
https://hal.archives-ouvertes.fr/hal-01447691
DOI : 10.1371/journal.pntd.0005254
PUBMED : 28081120

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
We measure throw distributions for graben-bounding normal faults from two areas on Mars to investigate fault growth, displacement-length (D-max-L) scaling, and extensional strain using a complementary suite of techniques. Faults in the northern plains are inferred to be restricted at 2-3 km depth, as shown by a transition from linear scaling, with D-max-L ratios of similar to 1 x 10(-3), to nonlinear scaling for faults >50 km long. On the Alba Patera volcano, faults conform to linear Dmax-L scaling, with a Dmax-L ratio of similar to 6 x 10(-3), consistent with more deeply penetrating faults that are not restricted at depth. These grabens accommodate larger extensional strains (similar to 0.84%) than the faults in the northern plains (similar to 0.23%), with a temporal change from regionally distributed to localized deformation and associated increases in D-max-L ratio extensional strain, and perhaps down-dip fault height. The results suggest that both spatial and temporal variations in extensional strain and displacement-length scaling relations, along with fault restriction, are recorded by Martian fault populations. (C) 2009 Elsevier Ltd. All rights reserved.
ISSN: 0191-8141

hal-00420888
https://hal.archives-ouvertes.fr/hal-00420888
DOI : 10.1016/j.jsg.2009.03.016

Éditeur(s) : HAL CCSD Landes Bioscience
Résumé : International audience
In 2008 we published the first set of guidelines for standardizing research in autophagy. Since then, research on this topic has continued to accelerate, and many new scientists have entered the field. Our knowledge base and relevant new technologies have also been expanding. Accordingly, it is important to update these guidelines for monitoring autophagy in different organisms. Various reviews have described the range of assays that have been used for this purpose. Nevertheless, there continues to be confusion regarding acceptable methods to measure autophagy, especially in multicellular eukaryotes. A key point that needs to be emphasized is that there is a difference between measurements that monitor the numbers or volume of autophagic elements (e.g., autophagosomes or autolysosomes) at any stage of the autophagic process vs. those that measure flux through the autophagy pathway (i.e., the complete process); thus, a block in macroautophagy that results in autophagosome accumulation needs to be differentiated from stimuli that result in increased autophagic activity, defined as increased autophagy induction coupled with increased delivery to, and degradation within, lysosomes (in most higher eukaryotes and some protists such as Dictyostelium) or the vacuole (in plants and fungi). In other words, it is especially important that investigators new to the field understand that the appearance of more autophagosomes does not necessarily equate with more autophagy. In fact, in many cases, autophagosomes accumulate because of a block in trafficking to lysosomes without a concomitant change in autophagosome biogenesis, whereas an increase in autolysosomes may reflect a reduction in degradative activity. Here, we present a set of guidelines for the selection and interpretation of methods for use by investigators who aim to examine macroautophagy and related processes, as well as for reviewers who need to provide realistic and reasonable critiques of papers that are focused on these processes. These guidelines are not meant to be a formulaic set of rules, because the appropriate assays depend in part on the question being asked and the system being used. In addition, we emphasize that no individual assay is guaranteed to be the most appropriate one in every situation, and we strongly recommend the use of multiple assays to monitor autophagy. In these guidelines, we consider these various methods of assessing autophagy and what information can, or cannot, be obtained from them. Finally, by discussing the merits and limits of particular autophagy assays, we hope to encourage technical innovation in the field.
Autophagy

hal-00735751
https://hal.archives-ouvertes.fr/hal-00735751
PUBMED : 22966490

Éditeur(s) : HAL CCSD Nature Publishing Group
Résumé : International audience
Forest ecosystems have integral roles in climate stability, biodiversity and economic development. Soil stewardship is essential for sustainable forest management. Organic matter (OM) removal and soil compaction are key disturbances associated with forest harvesting, but their impacts on forest ecosystems are not well understood. Because microbiological processes regulate soil ecology and biogeochemistry, microbial community structure might serve as indicator of forest ecosystem status, revealing changes in nutrient and energy flow patterns before they have irreversible effects on long-term soil productivity. We applied massively parallel pyrosequencing of over 4.6 million ribosomal marker sequences to assess the impact of OM removal and soil compaction on bacterial and fungal communities in a field experiment replicated at six forest sites in British Columbia, Canada. More than a decade after harvesting, diversity and structure of soil bacterial and fungal communities remained significantly altered by harvesting disturbances, with individual taxonomic groups responding differentially to varied levels of the disturbances. Plant symbionts, like ectomycorrhizal fungi, and saprobic taxa, such as ascomycetes and actinomycetes, were among the most sensitive to harvesting disturbances. Given their significant ecological roles in forest development, the fate of these taxa might be critical for sustainability of forest ecosystems. Although abundant bacterial populations were ubiquitous, abundant fungal populations often revealed a patchy distribution, consistent with their higher sensitivity to the examined soil disturbances. These results establish a comprehensive inventory of bacterial and fungal community composition in northern coniferous forests and demonstrate the long-term response of their structure to key disturbances associated with forest harvesting. The ISME Journal (2012) 6, 2199-2218; doi:10.1038/ismej.2012.84; published online 2 August 2012
ISSN: 1751-7362

hal-01546179
https://hal.archives-ouvertes.fr/hal-01546179
DOI : 10.1038/ismej.2012.84

Éditeur(s) : Pergamon-elsevier Science Ltd
Résumé : Society's needs for a network of in situ ocean observing systems cross many areas of earth and marine science. Here we review the science themes that benefit from data supplied from ocean observatories. Understanding from existing studies is fragmented to the extent that it lacks the coherent long-term monitoring needed to address questions at the scales essential to understand climate change and improve geo-hazard early warning. Data sets from the deep sea are particularly rare with long-term data available from only a few locations worldwide. These science areas have impacts on societal health and well-being and our awareness of ocean function in a shifting climate. Substantial efforts are underway to realise a network of open-ocean observatories around European Seas that will operate over multiple decades. Some systems are already collecting high-resolution data from surface, water column, seafloor, and sub-seafloor sensors linked to shore by satellite or cable connection in real or near-real time, along with samples and other data collected in a delayed mode. We expect that such observatories will contribute to answering major ocean science questions including: How can monitoring of factors such as seismic activity, pore fluid chemistry and pressure, and gas hydrate stability improve seismic, slope failure, and tsunami warning? What aspects of physical oceanography, biogeochemical cycling, and ecosystems will be most sensitive to climatic and anthropogenic change? What are natural versus anthropogenic changes? Most fundamentally, how are marine processes that occur at differing scales related? The development of ocean observatories provides a substantial opportunity for ocean science to evolve in Europe. Here we also describe some basic attributes of network design. Observatory networks provide the means to coordinate and integrate the collection of standardised data capable of bridging measurement scales across a dispersed area in European Seas adding needed certainty to estimates of future oceanic conditions. Observatory data can be analysed along with other data such as those from satellites, drifting floats, autonomous underwater vehicles, model analysis, and the known distribution and abundances of marine fauna in order to address some of the questions posed above. Standardised methods for information management are also becoming established to ensure better accessibility and traceability of these data sets and ultimately to increase their use for societal benefit. The connection of ocean observatory effort into larger frameworks including the Global Earth Observation System of Systems (GEOSS) and the Global Monitoring of Environment and Security (GMES) is integral to its success. It is in a greater integrated framework that the full potential of the component systems will be realised. Crown Copyright (C) 2011 Published by Elsevier Ltd. All rights reserved.
Progress In Oceanography (0079-6611) (Pergamon-elsevier Science Ltd), 2011-10 , Vol. 91 , N. 1 , P. 1-33

Droits : 2011 Published by Elsevier Ltd. All rights reserved.
http://archimer.ifremer.fr/doc/00044/15557/15923.pdf
DOI:10.1016/j.pocean.2011.05.001
http://archimer.ifremer.fr/doc/00044/15557/

Éditeur(s) : Elsevier
Résumé : The distribution and speciation of mercury (Hg) in air, rain, and surface waters from the artificial tropical lake of Petit-Saut in French Guiana were investigated during the 2003/04 period. In the air, total gaseous mercury (TGM) at the dam station averaged 12 +/- 2 pmol m(-3) of which > 98% was gaseous elemental mercury (GEM). GEM distribution depicted a day-night cycling with high concentrations (up to 15 pmol m(-3)) at dawn and low concentrations (down to 5 pmol m(-3)) at nightfall. Reactive gaseous mercury (RGM) represented < 1% of the GEM with a mean concentration of 4 3 fmol m(-3). Diel RGM variations were negatively related to GEM. In the rain, the sum of all Hg species in the unfiltered (HgTUNF) averaged 16 +/- 112 pmol L-1. Temporal distribution of HgTUNF exhibited a pattern of high concentrations during the late dry seasons (up to 57.5 pmol L-1) and low concentrations (down to 2.7 pmol L-1) in the course of the wet seasons. Unfiltered reactive (HgRUNF), dissolved gaseous (DGM) and monomethyl (MMHgUNF.) Hg constituted 20, 5 and 5% of HgTUNF, respectively. All measured Hg species were positively related and displayed negative relationships with the pH of the rain. In the reservoir surface waters, dissolved total mercury (HgTD) averaged 3.4 +/- 1.2 pmol L-1 of which 10% consisted of DGM. DGM showed a trend of high concentrations during the dry seasons (480 +/- 270 finol L-1 and lower (230 +/- 130 finol L-1) in the course of the wet seasons. Diel variations included diurnal photo-induced DGM production (of about 60 firiol L-1 h(-1)) coupled to minute to hour oxidation/ reduction cycles (of > 100 firiol L-1 amplitude). Finally, calculated atmospheric Hg inputs to the Petit-Saut reservoir represented 14 mot yr(-1) whereas DGM evasion reached 23 mol yr(-1). Apportionment among forms of Hg deposition indicated that up to 75% of the total Hg invasive flux follows the rainfall pathway. (C) 2007 Elsevier B.V. All rights reserved.
Science of The Total Environment (0048-9697) (Elsevier), 2007-10 , Vol. 385 , N. 1-3 , P. 132-145

Droits : 2007 Elsevier B.V. All rights reserved
http://archimer.ifremer.fr/doc/2007/publication-3329.pdf
DOI:10.1016/j.scitotenv.2007.06.013
http://archimer.ifremer.fr/doc/00000/3329/

Éditeur(s) : Amer Geophysical Union
Résumé : Bolide impacts and large-scale volcanic eruptions have been proposed as possible causes of the massive extinction of life that has occurred episodically in Earth's history. In spite of the catastrophic disruption of the local environment that accompanies bolide impacts and volcanic eruptions, it has been difficult to explain why these events sometimes lead to global extinction of species. We propose that, in some cases, the missing link may be provided by hypercanes, runaway hurricanes that are capable of injecting massive amounts of water and aerosols into the middle and upper stratosphere, where they may have profound effects on atmospheric chemistry and radiative transfer. Hypercanes are theorized to occur when the sea surface temperature exceeds a critical threshold, which may occur when sea water is locally heated by bolide impact, shallow-sea volcanism, or possibly, by overturning of superheated brine pools formed by underwater volcanic activity. Simulations using a convection-resolving nonhydrostatic, axisymmetric numerical model show that hypercanes can indeed develop when the sea surface temperature is high, and that they inject large amounts of mass into the stratosphere.
Journal Of Geophysical Research-atmospheres (2169-897X) (Amer Geophysical Union), 1995-07 , Vol. 100 , N. D7 , P. 13755-13765

Droits : 1995 by the American Geophysical Union.
http://archimer.ifremer.fr/doc/00258/36966/35611.pdf
DOI:10.1029/95JD01368
http://archimer.ifremer.fr/doc/00258/36966/

Éditeur(s) : HAL CCSD Nature Publishing Group
Résumé : International audience
ISSN: 1752-0894

hal-01667978
https://hal.archives-ouvertes.fr/hal-01667978
DOI : 10.1038/s41561-017-0028-x

Éditeur(s) : American Geophysical Union
Résumé : We use seismic reflection and rock sample data to propose that the first-order physiography of New Caledonia Trough and Norfolk Ridge formed in Eocene and Oligocene time and was associated with the onset of subduction and back-arc spreading at the Australia-Pacific plate boundary. Our tectonic model involves an initial Cretaceous rift that is strongly modified by Cenozoic subduction initiation. Hence, we are able to explain (1) complex sedimentary basins of inferred Mesozoic age; (2) a prominent unconformity and onlap surface of middle Eocene to early Miocene age at the base of flat-lying sediments beneath the axis of New Caledonia Trough; (3) gently dipping, variable thickness, and locally deformed Late Cretaceous strata along the margins of the trough; (4) platform morphology and unconformities on either side of the trough that indicate a phase of late Eocene to early Miocene uplift to near sea level, followed by rapid Oligocene and Miocene subsidence of similar to 1100-1800 m; and (5) seismic reflection facies tied to boreholes that suggest absolute tectonic subsidence at the southern end of New Caledonia Trough by 1800-2200 m since Eocene time. The Cenozoic part of the model involves delamination and subduction initiation followed by rapid foundering and rollback of the slab. This created a deep (>2 km) enclosed oceanic trough, similar to 2000 km long and 200-300 km across, in Eocene and Oligocene time as the lower crust detached, with simultaneous uplift and local land development along basin flanks. Disruption of Late Cretaceous and Paleogene strata was minimal during this Cenozoic phase and involved only subtle tilting and local reverse faulting or folding. Basin formation was possible through the action of at least one detachment fault that allowed the lower crust to either be subducted into the mantle or exhumed eastward into Norfolk Basin. We suggest that delamination of the lithosphere, with possible mixing of the lower crust back into the mantle, is more widespread than previously thought and may be commonly associated with subduction initiation, such as Cenozoic events in the Mediterranean and western Pacific.
Tectonics (0278-7407) (American Geophysical Union), 2010-03 , Vol. 29 , N. TC2004 , P. 1-16

Droits : Copyright 2010 by the American Geophysical Union.
http://archimer.ifremer.fr/doc/00001/11271/9283.pdf
DOI:10.1029/2009TC002476
http://archimer.ifremer.fr/doc/00001/11271/

Éditeur(s) : HAL CCSD Ecological Society of America
Résumé : Line transect surveys are widely used in neotropical rainforests to estimate the population abundance of medium- and large-sized vertebrates. The use of indices such as Encounter Rate has been criticized because the probability of animal detection may fluctuate due to the heterogeneity of environmental conditions among sites. In addition, the morphological and behavioral characteristics (biological traits) of species affect their detectability. In this study, we compared the extent to which environmental conditions and species' biological traits bias abundance estimates in terra firme rainforests in French Guiana. The selected environmental conditions included both physical conditions and forest structure covariates, while the selected biological traits included the morphological and behavioral characteristics of species. We used the distance sampling method to model the detection probability as an explicit function of environmental conditions and biological traits and implemented a model selection process to determine the relative importance of each group of covariates. Biological traits contributed to the variability of animal detectability more than environmental conditions, which had only a marginal effect. Detectability was best for large animals with uniform or disruptive markings that live in groups in the canopy top. Detectability was worst for small, solitary, terrestrial animals with mottled markings. In the terra firme rainforests which represent ~80% of the Amazonia and Guianas regions, our findings support the use of relative indices such as the encounter rate to compare population abundance between sites in species-specific studies. Even though terra firme rainforests may appear similar between regions of Amazonia and the Guianas, comparability must be ensured, especially in forests disturbed by human activity. The detection probability can be used as an indicator of species' vulnerability to hunting and, thus, to the risk of local extinction. Only a few biological trait covariates are required to correctly estimate the detectability of the majority of medium- and large-sized vertebrates. Thus, a biological trait model could be useful in predicting the detection probabilities of rare, uncommon or localized species for which few data are available to fit the detection function.
ISSN: 1051-0761

hal-01555267
https://hal.archives-ouvertes.fr/hal-01555267
DOI : 10.1002/eap.1549
PRODINRA : 396614
PUBMED : 28419598

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
The Tapajós Domain in the Amazonian Craton comprises hundreds of gold occurrences mainly hosted by Paleoproterozoic granitic rocks, whose geotectonic evolution and deformation style are poorly understood. The Tocantinzinho granite hosts a large amount of gold, forming the largest gold deposit known in the Tapajós Domain. The Tocantinzinho area is formed by plutonic rocks cut by subvolcanic rocks shallowly emplaced and constrained by NW-SE strike-slip faults, probably in a transtensive site. The magmatism started with the emplacement of huge granodiorite that presently formed the basement country rocks (2007–1997Ma), it was followed by the Tocantinzinho granite magmatism (1989–1979 Ma), and the coeval to late andesite emplacement. Petrological and textural evidence suggest the complex and synchronous occurrence of tectonics and few magmatic events expressed by the emplacement of successive granite, aplite, pegmatite, andesite, and dacite. The magmatic event finished with the cooling of this set of rocks below c. 320 °C, at c. 1950 Ma. The cooling rates vary from c. 3.6–14.7°C/Ma, with an average of 7.5°C/Ma, indicating no important vertical exhumations. The elongated geometry of the Tocantinzinho granite and related rocks, features of solid-state deformation within granites and syntectonic deformation of some andesite dikes imply the existence of wrench type tectonic control. The I-type signature of these rocks, their high-K calc-alkaline and metaluminous to peraluminous affinities combined with the tectonic style indicate a post-collisional tectonic setting. Based on geochemical and structural constrains, the rocks from the Tocantinzinho area are interpreted as a part of the Creporizão Suite and probably represent the first expressions of this syn-to late-tectonic magmatism in the region. A continuous and progressive process of magmatic emplacement controlled by tectonics will be proposed and discussed in conclusion.
ISSN: 0895-9811

hal-01526936
https://hal.archives-ouvertes.fr/hal-01526936
DOI : 10.1016/j.jsames.2017.02.003

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Voting rules relying on k-nearest neighbors (k-NN) are an effective tool in countless many machine learning techniques. Thanks to its simplicity, k-NN classification is very attractive to practitioners, as it enables very good performances in several practical applications. However, it suffers from various drawbacks, like sensitivity to "noisy" instances and poor generalization properties when dealing with sparse high-dimensional data. In this paper, we tackle the k-NN classification problem at its core by providing a novel k-NN boosting approach. Namely, we propose a supervised learning algorithm, called Universal Nearest Neighbors (UNN), that induces a leveraged k-NN rule by globally minimizing a surrogate risk upper bounding the empirical misclassification rate over training data. Interestingly, this surrogate risk can be arbitrary chosen from a class of Bregman loss functions, including the familiar exponential, logistic and squared losses. Furthermore, we show that UNN allows to efficiently filter a dataset of instances by keeping only a small fraction of data. Experimental results on the synthetic Ripley's dataset show that such a filtering strategy is able to reject "noisy" examples, and yields a classification error close to the optimal Bayes error. Experiments on standard UCI datasets show significant improvements over the current state of the art.
ISSN: 0925-2312

hal-00664462
https://hal.inria.fr/hal-00664462

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Cnidarians living in symbiosis with photosynthetic dinoflagellates (commonly named zooxanthellae) are exposed to high concentrations of reactive oxygen species (ROS) upon illumination. To quench ROS production, both the cnidarian host and zooxanthellae express a full suite of antioxidant enzymes. Studying antioxidative balance is therefore crucial to understanding how symbiotic cnidarians cope with ROS production. We characterized glutathione peroxidases (GPx) in the symbiotic cnidarian Anemonia viridis by analysis of their isoform diversity, their activity distribution in the three cellular compartments (ectoderm, endoderm and zooxanthellae) and their involvement in the response to thermal stress. We identified a GPx repertoire through a phylogenetic analysis showing 7 GPx transcripts belonging to the A. viridis host and 4 GPx transcripts strongly related to Symbiodinium sp. The biochemical approach, used for the first time with a cnidarian species, allowed the identification of GPx activity in the three cellular compartments and in the animal mitochondrial fraction, and revealed a high GPx electrophoretic diversity. The symbiotic lifestyle of zooxanthellae requires more GPx activity and diversity than that of free-living species. Heat stress induced no modification of GPx activities. We highlight a high GPx diversity in A. viridis tissues by genomic and biochemical approaches. GPx activities represent an overall constitutive enzymatic pattern inherent to symbiotic lifestyle adaptation. This work allows the characterization of the GPx family in a symbiotic cnidarian and establishes a foundation for future studies of GPx in symbiotic cnidarians.
ISSN: 0300-9084

hal-01404424
http://hal.upmc.fr/hal-01404424
http://hal.upmc.fr/hal-01404424/document
http://hal.upmc.fr/hal-01404424/file/Pey_Characterization_of.pdf
DOI : 10.1016/j.biochi.2016.10.016

Éditeur(s) : HAL CCSD Wiley
Résumé : International audience
Within the framework of the international field campaign COPS (Convective and Orographically induced Precipitation Study), a large suite of state-of-the-art meteorological instrumentation was operated, partially combined for the first time. This includes networks of in situ and remote-sensing systems such as the Global Positioning System as well as a synergy of multi-wavelength passive and active remote-sensing instruments such as advanced radar and lidar systems. The COPS field phase was performed from 01 June to 31 August 2007 in a low-mountain area in southwestern Germany/eastern France covering the Vosges mountains, the Rhine valley and the Black Forest mountains. The collected dataset covers the entire evolution of convective precipitation events in complex terrain from their initiation, to their development and mature phase until their decay. Eighteen Intensive Observations Periods with 37 operation days and eight additional Special Observations Periods were performed, providing a comprehensive dataset covering different forcing conditions. In this article, an overview of the COPS scientific strategy, the field phase, and its first accomplishments is given. Highlights of the campaign are illustrated with several measurement examples. It is demonstrated that COPS research provides new insight into key processes leading to convection initiation and to the modification of precipitation by orography, in the improvement of quantitative precipitation forecasting by the assimilation of new observations, and in the performance of ensembles of convection-permitting models in complex terrain.
ISSN: 0035-9009

hal-00565513
https://hal.archives-ouvertes.fr/hal-00565513
DOI : 10.1002/qj.752

Éditeur(s) : HAL CCSD IEEE
Résumé : International audience
Recent works display that large scale image classification problems rule out computationally demanding methods. On such problems, simple approaches like k-NN are affordable contenders, with still room space for statistical improvements under the algorithmic constraints. A recent work showed how to leverage k-NN to yield a formal boosting algorithm. This method, however, has numerical issues that make it not suited for large scale problems. We propose here an Adaptive Newton-Raphson scheme to leverage k-NN, N³, which does not suffer these issues. We show that it is a boosting algorithm, with several key algorithmic and statistical properties. In particular, it may be sufficient to boost a subsample to reach desired bounds for the loss at hand in the boosting framework. Experiments are provided on the SUN, and Caltech databases. They confirm that boosting a subsample -- sometimes containing few examples only -- is sufficient to reach the convergence regime of N³. Under such conditions, N³ challenges the accuracy of contenders with lower computational cost and lower memory requirement.
IEEE International Workshop on Machine Learning for Signal Processing
Southampton, United Kingdom

hal-00959125
https://hal.archives-ouvertes.fr/hal-00959125
https://hal.archives-ouvertes.fr/hal-00959125/document
https://hal.archives-ouvertes.fr/hal-00959125/file/BNNB_mlsp2013.pdf

Éditeur(s) : HAL CCSD
Résumé : International audience
The k-nearest neighbors (k-NN) classification rule is still an essential tool for computer vision applications, such as scene recognition. However, k-NN still features some major drawbacks, which mainly reside in the uniform voting among the nearest prototypes in the feature space. In this paper, we propose a new method that is able to learn the "relevance" of prototypes, thus classifying test data using a weighted k-NN rule. In particular, our algorithm, called Multi-class Leveraged k-nearest neighbor (MLNN), learns the prototype weights in a boosting framework, by minimizing a surrogate exponential risk over training data. We propose two main contributions for improving computational speed and accuracy. On the one hand, we implement learning in an inherently multiclass way, thus providing significant computation time reduction over one-versus-all approaches. Furthermore, the leveraging weights enable effective data selection, thus reducing the cost of k-NN search at classification time. On the other hand, we propose a kernel generalization of our approach to take into account real-valued similarities between data in the feature space, thus enabling more accurate estimation of the local class density. We tested MLNN on three datasets of natural images. Results show that MLNN significantly outperforms classic k-NN and weighted k-NN voting. Furthermore, using an adaptive Gaussian kernel provides significant performance improvement. Finally, the best results are obtained when using MLNN with an appropriate learned metric distance.
Proceedings of the 10th Asian Conference on Computer Vision, ACCV 2010, November 8-12, 2010, Queenstown, New Zealand
Queenstown, New Zealand

hal-00664606
https://hal.inria.fr/hal-00664606
https://hal.inria.fr/hal-00664606/document
https://hal.inria.fr/hal-00664606/file/mlnn_accvfinal.pdf

Éditeur(s) : HAL CCSD
Résumé : 15 pages
A Stochastic Differential Equation appearing in the statistical theory of turbulence is extended in random environment by assuming that its two parameters are switched by an unobserved continuoustime Markov chain whose states represent the states of the environment. A Dirichlet process is placed as a prior on the space of the sample paths of this chain, leading to a hierarchical Dirichlet model whose estimation is done both on simulated data and on real data of wind speed measured at the entrance of a mangrove ecosystem.
https://hal.archives-ouvertes.fr/hal-00455050

hal-00455050
https://hal.archives-ouvertes.fr/hal-00455050
https://hal.archives-ouvertes.fr/hal-00455050/document
https://hal.archives-ouvertes.fr/hal-00455050/file/CSDA06_Srik_Hal.pdf

Éditeur(s) : HAL CCSD Ecological Society of America
Résumé : Herbivory is viewed as a major driver of plant evolution and the mostimportant energy pathway from plants to higher trophic levels. Therefore, understandingpatterns of herbivory on plants remains a key focus in evolution and ecology. Theevolutionary impacts of leaf herbivory include altering plant fitness, local adaptation, theevolution of defenses, and the diversification of plants as well as natural enemies. Leafherbivory also impacts ecological processes such as plant productivity, communitycomposition, and ecosystem nutrient cycling. Understanding the impact of herbivory onthese ecological and evolutionary processes requires species-specific, as opposed tocommunity-level, measures of herbivory. In addition, species-specific data enables the use ofmodern comparative methods to account for phylogenetic non-independence. Althoughhundreds of studies have measured natural rates of leaf consumption, we are unaware of anyaccessible compilation of these data. We created such a data set to provide the raw dataneeded to test general hypotheses relating to plant–herbivore interactions and to test theinfluence of biotic and abiotic factors on herbivory rates across large spatial scales. A largerepository will make this endeavor more efficient and robust. In total, we compiled 2641population-level measures for either annual or daily rates of leaf herbivory across 1145 speciesof vascular plants collected from 189 studies. All damage measures represent naturaloccurrences of herbivory that span numerous angiosperm, gymnosperm, and fern species. Toenable researchers to explore the causes of variation in herbivory and how these mightinteract, we added information about the study sites including: geolocation, climateclassification, habitat descriptions (e.g., seashore, grassland, forest, agricultural fields), andplant trait information concerning growth form and duration (e.g., annual vs. perennial). Wealso included extensive details of the methodology used to measure leaf damage, includingseasons and months of sampling, age of leaves, and the method used to estimate percentagearea missing. We anticipate that these data will make it possible to test important hypothesesin the plant–herbivore literature, including the plant apparency hypothesis, the latitudinalherbivorydefense hypothesis, the resource availability hypothesis, and the macroevolutionaryescalation of defense hypothesis.
ISSN: 0012-9658

hal-01601627
https://hal.archives-ouvertes.fr/hal-01601627
DOI : 10.1890/13-1741.1
PRODINRA : 397136
PUBMED : 24870043