Com a evolução da informática, diferentes meios de comunicação passaram a explorar a Web como um meio de divulgação de suas informações. Diferentes fontes de informações, diferentes estilos de escrita e a curiosidade nata do ser humano despertam o interesse de leitores por conhecer mais de um relato sobre um mesmo tema. Para que a leitura de diferentes relatos com conteúdo similar seja possível, leitores precisam procurar, ler e analisar informações fornecidas por diferentes fontes de informação. Essa atividade, além de exigir grande investimento de tempo, sobrecarrega cognitivamente usuários. Faz parte das pesquisas da área de Hipermídia investigar mecanismos que apóiem usuários no processo de identificação de informações em repositórios homogêneos, sejam eles disponibilizados na Web ou não. No contexto desta tese, repositórios com informações de conteúdo homogêneo são aqueles cujas informações tratam do mesmo assunto. Esta tese tem por objetivo investigar a especificação, a instanciação e a experimentação de um arcabouço para apoiar a tarefa de criação automática de ligações hipertexto entre repositórios homogêneos. O arcabouço proposto, denominado CARe (Criação Automática de Relacionamentos), é representado por um conjunto de classes que realizam a coleta de informações a serem relacionadas e que processam essas informações para a geração de índices. Esses índices são relacionados e utilizados na criação automática de ligações hipertexto entre a informação original. A definição do arcabouço se deu após uma fase de análise de domínio na qual foram identificados requisitos e construídos componentes de software. Nessa fase, vários protótipos também foram construídos de modo iterativo / With the evolution of the Internet, distinct communication media have focused on the Web as a channel of information publishing. An immediate consequence is an abundance of sources of information and writing styles in the Web. This effect, combining with the inherent curiosity of human beings, has led Web users to look for more than a single article about a same subject. To gain access to separate on a same subject, readers need to search, read and analyze information provided by different sources of information. Besides consuming a great amount of time, that activity imposes a cognitive overhead to users. Several hypermedia researches have investigated mechanisms for supporting users during the process of identifying information on homogeneous repositories, available or not on the Web. In this thesis, homogeneous repositories are those containing information that describes a same subject. This thesis aims at investigating the specification and the construction of a framework intended to support the task of automatic creation of hypertext links between homogeneous repositories. The framework proposed, called CARe (Automatic Creation of Relationships), is composed of a set of classes, methods and relationships that gather information to be related, and also process that information for generating an index. Those indexes are related and used in the automatic creation of hypertext links among distinct excerpts of original information. The framework was defined based on a phase of domain analysis in which requirements were identified and software components were built. In that same phase several prototypes were developed in an iterative prototyping

Introdução: O câncer de pulmão (CP) é o tipo de câncer que mais mata no mundo e o cigarro ainda é sua causa mais importante. Além disso, a alimentação tem sido associada ao CP, por ser fonte de vitaminas e aminoácidos que fazem parte do metabolismo do carbono (MC). O MC é considerado mecanismo chave na manutenção da integridade do DNA e na regulação da expressão gênica, que, dessa forma, deve estar relacionado à carcinogênese. A ativação da imunidade está associada ao envelhecimento em indivíduos saudáveis, assim como a uma série de patologias, incluindo o câncer. Objetivo: Estudar como o MC, a ativação da imunidade e o tabaco estão relacionados ao risco de CP em um estudo caso-controle aninhado à coorte do EPIC (European Prospective Investigation into Cancer and Nutrition). Métodos: Para avaliar se os níveis plasmáticos de cotinina são um bom biomarcador da exposição ao tabaco, foram utilizados modelos lineares generalizados. Para avaliar os efeitos do tabaco, do MC e da ativação da imunidade no risco de CP, foram aplicados modelos de equações estruturais (MEE) de duas maneiras diferentes (com e sem variáveis latentes). Resultados: Com base nas respostas aos questionários de qualidade de vida, com relação às questões sobre fumo ativo e passivo, a cotinina se mostrou um bom biomarcador de exposição recente ao tabaco (tanto o aumento da exposição passiva quanto ativa foram significativas, P<0,001 e P<0,001 respectivamente). Em um MEE com variáveis observadas, incluindo o MC e a via de ativação da imunidade, a metionina e o folato como causas proximais apresentaram uma forte e inversa associação com o risco de CP. O aumento em um desvio-padrão nos níveis séricos de metionina e de folato significou uma redução no risco de 19 por cento (P<0,01) e 12 por cento (P=0,03) respectivamente. Em um MEE com variáveis latentes (cada uma representando o conjunto de vitaminas e aminoácidos importantes para promover: metilação de DNA, síntese de núcletídeos e imune ativação), foram encontrados efeitos protetores diretos da metilação do DNA (P=0,018) e da imune ativação (P=0,037); por outro lado, a síntese de nucletídeos não apresentou efeito no risco do câncer (P=0,098). Nas duas abordagens de MEE o cigarro permaneceu como a causa de maior impacto. Conclusões: A cotinina mostrou-se um bom biomarcador da exposição ao tabaco (ativa e passivamente). Confirmou-se que a via de metilação é um fator de proteção contra o CP. A ativação da imunidade apresentou um efeito direto de proteção contra o CP no modelo com variáveis latentes, equanto que, a síntese de nucletídeos não apresentou relação com o CP. O tabaco continua sendo o fator de maior impacto no risco de CP / Background: Lung cancer (LC) continues to be the most common cancer death in the world. Tobacco exposure continues to be the most important cause. In addition, micronutrient intake has been linked to LC, because they are the main source of vitamins and amino acids involved in the one-carbon metabolism (OCM) which is considered key in maintaining DNA integrity, regulating gene expression, and may thus affect carcinogenesis. Immune activation is involved in the aging process in normal healthy individuals as well as in a number of pathologies, including cancer. Objectives: To investigate how OCM, immune activation and tobacco are related to LC incidence in a nested case-control study from the European Prospective Investigation into Cancer and Nutrition (EPIC) cohort. Methods: To validate plasma cotinine levels as a good biomarker for tobacco exposure, a generalized linear model was applied. To evaluate the effects of tobacco, OCM and immune activation in LC, structural equation models (SEM) were applied in two different ways. Results: Based on questions about smoking, passive smoking and number of cigarettes smoked, it was shown that cotinine is a good biomarker for tobacco exposure (passive and active exposure with significant relation, p<0.001 and P<0.001, respectively). In a SEM model with only observed variables, including OCM and immune activation, methionine and folate as proximal causes presented a strong and inverse relation with LC risk. An increase in one standard deviation of serum levels of methionine and folate meant a 19 per cent (P<0.01) and 12 per cent (P<0.01) reduction in LC risk, respectively. In a SEM including latent variables (each one including vitamins and amino acids important to promote DNA methylation, nucleotide synthesis and immune activity), a direct and protective effect for DNA methylation (p=0.018) and immune activation was found (p=0.037), whereas nucleotide synthesis did not present a significant total effect. In both approaches of SEM, tobacco exposure remains with the highest impact on LC risk. Conclusions: It was found that cotinine is a good biomarker of tobacco exposure (active and passive). It was confirmed that methylation protects against LC. Immune activation presented a direct protective effect in the latent model, while nucleotide synthesis was not confirmed to be related to LC risk. Tobacco effect remains as the factor with highest impact in lung cancer

Submitted by Maria Beatriz Vieira (mbeatriz.vieira@gmail.com) on 2017-04-25T16:35:31Z No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) dissertacao_thiago_strobel.pdf: 495296 bytes, checksum: 891da86ed6f9b4154bb03577981bbf35 (MD5) / Approved for entry into archive by Aline Batista (alinehb.ufpel@gmail.com) on 2017-04-25T17:25:13Z (GMT) No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) dissertacao_thiago_strobel.pdf: 495296 bytes, checksum: 891da86ed6f9b4154bb03577981bbf35 (MD5) / Approved for entry into archive by Aline Batista (alinehb.ufpel@gmail.com) on 2017-04-25T17:26:11Z (GMT) No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) dissertacao_thiago_strobel.pdf: 495296 bytes, checksum: 891da86ed6f9b4154bb03577981bbf35 (MD5) / Made available in DSpace on 2017-04-25T17:26:21Z (GMT). No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) dissertacao_thiago_strobel.pdf: 495296 bytes, checksum: 891da86ed6f9b4154bb03577981bbf35 (MD5) Previous issue date: 2015-12-08 / Sem bolsa / O objetivo do trabalho foi avaliar a qualidade física e fisiológica de sementes de soja colhidas com diferentes sistemas de trilha após a colheita e o armazenamento. A cultivar utilizada foi Syn 1059 RR (V Top), semeada sob sistema de semeadura direta na palha no dia 25/10/2014 com densidade de semeadura de 240.000 sementes por hectare e espaçamento de 0,5 m. Os tratamentos foram constituídos de diferentes sistemas de trilha, onde avaliou-se a) colhedora com sistema de trilha axial; b) colhedora com sistema de trilha axial equipada com plataforma “draper”; c) colhedora com sistema de trilha radial; d) controle - debulha manual. Foram analisados a porcentagem de “bandinhas”, a danificação mecânica pelo testes de hipoclorito, germinação, envelhecimento acelerado e emergência em campo. Os maiores valores de danos mecânicos, determinados pelo teste de hipoclorito, foram constatados para o sistema de trilha radial, seguido pelo sistema de trilha axial e axial + “draper” e os menores valores para sementes colhidas manualmente. A porcentagem de bandinha e a danificação mecânica foram maiores para o sistema de trilha radial, enquanto, sementes colhidas com este sistema atingiram menores valores de germinação, envelhecimento acelerado e emergência em campo. A germinação, envelhecimento acelerado e emergência em campo atingiram os menores valores em sementes colhidas com sistema de trilha radial, no entanto, após o armazenamento, estes resultados mantiveram-se apenas para o teste de emergência em campo. Sementes colhidas com sistemas de trilha axial e axial equipado com plataforma “draper” e trilha manual não apresentaram diferença para germinação, resultados de envelhecimento acelerado e emergência em campo. A qualidade física e fisiológica é superior quando empregadas as colhedoras com sistema axial e axial + “draper”, sem diferir de sementes colhidas manualmente. / The aim of the work was to evaluate the physical and physiological quality of soybean seeds harvested with different track system after harvest and storage. The cultivar used was Syn 1059 RR (V Top), sown under no-tillage system in straw in day 25/10/2014 with sowing rate of 240,000 seeds per hectare spacing of 0.5 m. The treatments were different track system, where was evaluated: a) harvesters with axial track system; b) harvesters with axial track system equipped with "draper" platform; c) harvesters with radial track system; d) Control - manual threshing. Were analyzed the percentage of "splits", mechanical damaging by the hypochlorite test, germination, accelerated aging and field emergence. The highest values of mechanical damage, determinated by the hypochlorite test, were seen to radial track system, followed by axial and axial track system + "draper" and the lowest values for seeds harvested manually. The percentage of splits and mechanical damage were higher for the radial track system, being that, the seeds harvested with this system reached the minor germination values, accelerated aging and field emergence. The germination, accelerated aging and field emergence reached the lowest values in seed harvested radial track system, however, after storage, these results have remained only for the field emergence test. Seeds harvested with axial and axial track systems equipped with platform "draper" and manual track also did not differ for germination, results of accelerated aging and field emergence. The physical and physiological seed quality is superior when employed harvesters with axial and axial system + "draper", without differ from seeds manually harvested.

Recomendações para rastreamento de Tuberculoses Latente (TBL) em pacientes que receberão tratamento com antagonistas do TNF-alfa (anti- TNF) permanecem controversas para regiões endêmicas Objetivo: Esse estudo buscou demonstrar a eficácia em longo prazo do rastreamento e tratamento da TBL em pacientes portadores de Artrite Reumatoide (AR) recebendo anti-TNF. Métodos: 202 pacientes com AR, antes do início do anti-TNF, foram rastreados para TBL por meio do teste tuberculínico (TT), Radiografia de tórax (RX) e história de prévia de exposição à tuberculose (EXP). Todos os pacientes foram seguidos com intervalos de um a três meses. Resultados: 85 pacientes (42%) foram tratados com um único agente anti-TNF e 117 pacientes (58%) mudaram de anti-TNF uma ou duas vezes. O rastreamento para TBL foi positivo em 66 pacientes, 44 apresentaram TT positivo, 23 apresentavam história de exposição (EXP), e 14, alterações radiográficas (RX). EXP isoladamente foi responsável por 14 diagnósticos em pacientes TT negativos. Pacientes portadores de TBL receberam tratamento com Isoniazida (300 mg/dia por seis meses) e nenhum deles desenvolveu TB. Durante os seguimentos, o TT foi repetido em 51 pacientes. A conversão foi observada em cinco: três foram diagnosticados com TBL e dois com TB ativa (14 e 36 meses após receber terapia anti-TNF), sugerindo nova exposição a TB. Conclusão: O rastreamento e tratamento da TBL antes do início da terapia com anti-TNF é efetiva em regiões endêmicas, e reforça a relevância da história de contato com TB para o diagnóstico da TBL em pacientes com AR / Recommendations for screening of latent tuberculosis infection (LTBI) in patients eligible for anti-TNF agents remain unclear in endemic regions. Objective: This study aimed to evaluate the long-term efficacy of LTBI screening/treatment in patients with rheumatoid arthritis (RA) receiving TNF blockers. Design: 202 RA patients were screened for LTBI prior to receiving anti-TNF treatment, by means of tuberculin skin test (TST), chest radiography (X-Ray), and history of tuberculosis exposure (EXP). All subjects were regularly followed at 1- to 3-month intervals. Results: Eighty-five patients (42%) were treated with a single anti-TNF agent, and 117 patients (58%) switched anti-TNF agents once or twice. LTBI screening was positive in 66 patients, 44 presented positive TST, 23 had a history of EXP, and 14, abnormal X-Ray. Exposure alone accounted for LTBI diagnosis in 14 patients with negative TSTs. LTBI patients were treated with Isoniazid (300 mg/day during six months) and none developed TB. During follow up, TST was repeated in 51 patients. Conversion was observed in five: three were diagnosed with LTBI and two with active TB (14 and 36 months after receiving anti-TNF therapy, suggesting new TB exposure). Conclusion: LTBI screening and treatment prior to anti-TNF treatment is effective in endemic areas and reinforces the relevance of contact history for diagnosing LTBI in RA patients

Les travaux de cette thèse présentent le développement d'algorithmes de classification de documents d'une part, ou d'analyse de réseaux complexes d'autre part, en s'appuyant sur la prétopologie, une théorie qui modélise le concept de proximité. Le premier travail développe un cadre pour la classification de documents en combinant une approche de topicmodeling et la prétopologie. Notre contribution propose d'utiliser des distributions de sujets extraites à partir d'un traitement topic-modeling comme entrées pour des méthodes de classification. Dans cette approche, nous avons étudié deux aspects : déterminer une distance adaptée entre documents en étudiant la pertinence des mesures probabilistes et des mesures vectorielles, et effet réaliser des regroupements selon plusieurs critères en utilisant une pseudo-distance définie à partir de la prétopologie. Le deuxième travail introduit un cadre général de modélisation des Réseaux Complexes en développant une reformulation de la prétopologie stochastique, il propose également un modèle prétopologique de cascade d'informations comme modèle général de diffusion. De plus, nous avons proposé un modèle agent, Textual-ABM, pour analyser des réseaux complexes dynamiques associés à des informations textuelles en utilisant un modèle auteur-sujet et nous avons introduit le Textual-Homo-IC, un modèle de cascade indépendant de la ressemblance, dans lequel l'homophilie est fondée sur du contenu textuel obtenu par un topic-model. / The work of this thesis presents the development of algorithms for document classification on the one hand, or complex network analysis on the other hand, based on pretopology, a theory that models the concept of proximity. The first work develops a framework for document clustering by combining Topic Modeling and Pretopology. Our contribution proposes using topic distributions extracted from topic modeling treatment as input for classification methods. In this approach, we investigated two aspects: determine an appropriate distance between documents by studying the relevance of Probabilistic-Based and Vector-Based Measurements and effect groupings according to several criteria using a pseudo-distance defined from pretopology. The second work introduces a general framework for modeling Complex Networks by developing a reformulation of stochastic pretopology and proposes Pretopology Cascade Model as a general model for information diffusion. In addition, we proposed an agent-based model, Textual-ABM, to analyze complex dynamic networks associated with textual information using author-topic model and introduced Textual-Homo-IC, an independent cascade model of the resemblance, in which homophily is measured based on textual content obtained by utilizing Topic Modeling.

Les alliages à mémoire de forme (AMF) sont des matériaux qui possèdent des propriétés mécaniques étonnantes : super élasticité, mémoire de forme proprement dite, grande capacité d'amortissement. Ces différentes propriétés thermomécaniques existent en fait généralement pour un même AMF, mais pour des températures d'utilisation différentes. La composition chimique d'un AMF est un paramètre clé de son comportement macroscopique. L'objectif de ce mémoire est l'étude des mécanismes microstructuraux qui sont à l'origine des propriétés thermomécaniques des AMF. Celles-ci prennent leur source dans un phénomène physique de changement de phase solide-solide nommé " transformation martensitique ". La phase mère (austénite) possède une structure cristalline cubique centrée pour tous les AMF connus. La transformation martensitique a pour effet de rompre cette symétrie. Elle donne naissance à de nouvelles structures cristallines qui sont fonctions principalement de la composition chimique de l'AMF. Le présent travail traite de la compréhension de l'organisation spatiale des phases austénite et (variantes de) martensite existant dans les AMF : les microstructures martensitiques. Dans une première partie de la thèse, une évolution de la notion de compatibilité cristallographique d'une microstructure est proposée. Une microstructure peut être non-compatible à contrainte nulle et pour autant exister dans le matériau sans créer " trop " d'irréversibilités mécaniques. Ceci permet de conserver la réversibilité de la transformation de phase et au final d'assurer la mémoire de forme. Le qualifcatif de microstructure " presque-compatible " est proposé. La seconde partie de la thèse se propose d'observer expérimentalement des microstructures martensitiques obtenues sous chargement mécanique dans un monocristal d'AMF en Cu-Al-Be. Deux techniques de mesure de champs sont couplées dans ce but : la thermographie infrarouge, qui délivre des champs de températures, et la méthode de la grille qui permet d'accéder aux champs de déformations. Des microstructures martensitiques sont révélées et analysées.

<p>I ulykkessammenheng går to av de vanligste undersøkelsesmetodene ut på henholdsvis å kartlegge det juridiske ansvaret for ulykken og å klarlegge årsaksforholdene. I media får vi stort sett bare høre om førstnevnte i form av oppslag om at ulykken skyldes menneskelig svikt. Årsaksforholdene tar gjerne lengre tid å kartlegge og har heller ikke elementer av helt eller syndebukk i seg, og er således mindre interessante å skrive om. I forebyggende sammenheng er det likevel gjerne her det er mest å hente. I erkjennelsen av at mennesket er og blir feilbarlig, bør man nettopp gripe tak i disse bakenforliggende årsakene, gjerne kalt latente feil, som også kan medvirke til å avverge at et tilløp får lov å utvikle seg til en ulykke. I denne oppgaven blir det gjort rede for bakenforliggende årsaker til en alvorlig samferdselsulykke på Vestlandet i 1999, nemlig hurtigbåten MS Sleipners forlis. Det blir også vist hvordan ulykken kunne vært unngått, eller i det minste at skadeomfanget kunne vært vesentlig redusert, dersom det hadde vært større fokus på å avdekke slike latente forhold. For å få offentligheten til å forstå viktigheten av slik årsakskartlegging, og ikke minst at ansvarlige instanser følger opp funnene i ettertid, er det viktig at media griper fatt i problemstillingen og setter den på dagsorden.</p> / <p>Two of the most common forms of investigations after accidents are the search for the judicial aspects of guilt on the one hand and the investigation to find the causes of the accident on the other hand. Media mainly tell us the result of the first one, namely that “the cause of the accident was human error”. Usually it takes more time to establish an understanding of the underlying causes. This investigation does not point out a hero, neither a scapegoat. Therefore the real causes are less interesting to the media. Still, when it comes to accident prevention, this is where one should make an effort. By recognizing the fact that humans are eligible to fail, it is here one should try to take preventive measures. By recognizing these latent conditions one may succeed to avoid potentially dangerous situations from developing to serious accidents. In this paper the underlying causes related to a serious accident that struck public transportation in Western Norway will be presented. The accident is the shipwreck of the high speed craft MS Sleipner in 1999. It will also be shown how this accident could have been prevented, or at least how the extent of the damage could have been substantially reduced, if only latent conditions had been paid a little more attention to. To make the general public understand the importance of uncovering such latent conditions, not to say make sure that they who are responsible take action to remove these conditions, it is vital that the media put these questions on the agenda.</p>

Wir versuchen herauszufinden, ob das subjektive Maß der Cloze-Vorhersagbarkeit mit der Kombination objektiver Maße (semantische und n-gram-Maße) geschätzt werden kann, die auf den statistischen Eigenschaften von Textkorpora beruhen. Die semantischen Maße werden entweder durch Abfragen von Internet-Suchmaschinen oder durch die Anwendung der Latent Semantic Analysis gebildet, während die n-gram-Wortmaße allein auf den Ergebnissen von Internet-Suchmaschinen basieren. Weiterhin untersuchen wir die Rolle der Cloze-Vorhersagbarkeit in SWIFT, einem Modell der Blickkontrolle, und wägen ab, ob andere Parameter den der Vorhersagbarkeit ersetzen können. Unsere Ergebnisse legen nahe, dass ein computationales Modell, welches Vorhersagbarkeitswerte berechnet, nicht nur Maße beachten muss, die die Relatiertheit eines Wortes zum Kontext darstellen; das Vorhandensein eines Maßes bezüglich der Nicht-Relatiertheit ist von ebenso großer Bedeutung. Obwohl hier jedoch nur Relatiertheits-Maße zur Verfügung stehen, sollte SWIFT ebensogute Ergebnisse liefern, wenn wir Cloze-Vorhersagbarkeit mit unseren Maßen ersetzen. / We try to determine whether it is possible to approximate the subjective Cloze predictability measure with two types of objective measures, semantic and word n-gram measures, based on the statistical properties of text corpora. The semantic measures are constructed either by querying Internet search engines or by applying Latent Semantic Analysis, while the word n-gram measures solely depend on the results of Internet search engines. We also analyse the role of Cloze predictability in the SWIFT eye movement model, and evaluate whether other parameters might be able to take the place of predictability. Our results suggest that a computational model that generates predictability values not only needs to use measures that can determine the relatedness of a word to its context; the presence of measures that assert unrelatedness is just as important. In spite of the fact, however, that we only have similarity measures, we predict that SWIFT should perform just as well when we replace Cloze predictability with our measures.

I ulykkessammenheng går to av de vanligste undersøkelsesmetodene ut på henholdsvis å kartlegge det juridiske ansvaret for ulykken og å klarlegge årsaksforholdene. I media får vi stort sett bare høre om førstnevnte i form av oppslag om at ulykken skyldes menneskelig svikt. Årsaksforholdene tar gjerne lengre tid å kartlegge og har heller ikke elementer av helt eller syndebukk i seg, og er således mindre interessante å skrive om. I forebyggende sammenheng er det likevel gjerne her det er mest å hente. I erkjennelsen av at mennesket er og blir feilbarlig, bør man nettopp gripe tak i disse bakenforliggende årsakene, gjerne kalt latente feil, som også kan medvirke til å avverge at et tilløp får lov å utvikle seg til en ulykke. I denne oppgaven blir det gjort rede for bakenforliggende årsaker til en alvorlig samferdselsulykke på Vestlandet i 1999, nemlig hurtigbåten MS Sleipners forlis. Det blir også vist hvordan ulykken kunne vært unngått, eller i det minste at skadeomfanget kunne vært vesentlig redusert, dersom det hadde vært større fokus på å avdekke slike latente forhold. For å få offentligheten til å forstå viktigheten av slik årsakskartlegging, og ikke minst at ansvarlige instanser følger opp funnene i ettertid, er det viktig at media griper fatt i problemstillingen og setter den på dagsorden. / Two of the most common forms of investigations after accidents are the search for the judicial aspects of guilt on the one hand and the investigation to find the causes of the accident on the other hand. Media mainly tell us the result of the first one, namely that “the cause of the accident was human error”. Usually it takes more time to establish an understanding of the underlying causes. This investigation does not point out a hero, neither a scapegoat. Therefore the real causes are less interesting to the media. Still, when it comes to accident prevention, this is where one should make an effort. By recognizing the fact that humans are eligible to fail, it is here one should try to take preventive measures. By recognizing these latent conditions one may succeed to avoid potentially dangerous situations from developing to serious accidents. In this paper the underlying causes related to a serious accident that struck public transportation in Western Norway will be presented. The accident is the shipwreck of the high speed craft MS Sleipner in 1999. It will also be shown how this accident could have been prevented, or at least how the extent of the damage could have been substantially reduced, if only latent conditions had been paid a little more attention to. To make the general public understand the importance of uncovering such latent conditions, not to say make sure that they who are responsible take action to remove these conditions, it is vital that the media put these questions on the agenda.

La circulation atmosphérique en Afrique de l'Ouest est caractérisée par des vents de sud-ouest (mousson) pendant la saison humide et par des vents de nord-est (harmattan) pendant la saison sèche. Cette alternance des saisons est due aux variations de pression liée à l'état des surfaces (rugosité, albédo, végétation) en réaction au forçage solaire. Ces mêmes états de surface génèrent une variabilité de flux turbulents de surface et des circulations secondaires qui rendent complexes les analyses des mesures effectuées sur place en vue de documenter les interactions surface-atmosphère. La modélisation fine échelle (LES) couramment utilisée dans l'étude de la couche limite atmosphérique est requise pour pouvoir palier à ces difficultés en raison de sa capacité à prendre en compte les flux turbulents en 3D et sur topographie complexe. Notre site d'étude est le bassin versant d'ARA située au Nord du Bénin dans un contexte Soudanien avec des propriétés de surface variables. Une analyse climatique est effectuée sur la base des observations de radiosondage, de radar UHF et de stations au sol afin d'extraire des données composites représentatives des saisons sèche et humide. Ces données composites ont servi par la suite à forcer le modèle Méso-NH dans sa version LES. Pour pouvoir caractériser les échelles de longueur des flux turbulents de surface relatives aux saisons sèche et humide, les données standard de forçage de surface de Méso-NH que sont le relief GTOPO30 (1km de résolution) et la végétation ECOCLIMAP (1km de résolution) ont été respectivement remplacer par le SRTM (90m de résolution) et les données de SPOT/HRV (20m de résolution) reéchantillonné à 90m de résolution. A l'aide d'outils statistiques comme la variographie 2D et le suivi Lagrangien, il ressort que la variabilité spatiale de la chaleur sensible H est gouvernée par le couple vent-relief tandis que celle de la chaleur latente E est difficile à mettre en lien sur végétation hétérogène (SPOT/HRV) en saison sèche. En saison humide, la variabilité spatiale du champ H dépend du vent tandis que celle du champ E dépend de la végétation. Cette étude révèle dans tous les cas que les échelles caractéristiques de ces deux champs diffèrent dans les mêmes conditions de forçage de surface et atmosphérique.