Martian clouds and dust play an important part of the radiative transfer and energy balance budget. To assist in fully understanding the impact of clouds and dust, the complete diurnal cycle needs to be characterized. One of the best methods to track diurnal variations on Mars is by measuring optical depth. The spatial and temporal trends of optical depth give insight into the dust and water cycles of the Martian atmosphere. Until now, spacecraft could only obtain optical depth during the day. In this thesis, nighttime images from the Mars Exploration Rover Spirit are used to calculate nighttime optical depth using photometric methods to capture star flux. Bright stars in well-known constellations are used in this analysis. The observed flux was compared to the expected flux to give nighttime optical depth values. The observed nighttime optical depth was consistently similar to the daytime optical depth values on both an individual image and sol-averaged basis. Recommendations are made going forward to use the Mars Science Laboratory Curiosity for conducting an optimal nighttime optical depth campaign to fully characterize the diurnal dust and water cycles of Mars. The Curiosity rover is well suited for nighttime imaging and can potentially provide valuable insight into the nighttime dust and cloud trends.

A lack of quality data on high clouds has led to inadequate representations within global weather and climate models. Recent advances in spaceborne measurements of the Earth’s atmosphere have provided complementary information on the interior of these clouds. This study demonstrate how an array of space-borne measurements can be used and combined, by close co-located comparisons in space and time, to form a more complete representation of high cloud processes and properties. High clouds are found in the upper atmosphere, where sub-zero temperatures frequently result in the formation of cloud particles that are composed of ice. Weather and climate models characterise the bulk properties of these ice particles to describe the current state of the cloud-sky atmosphere. By directly comparing measurements with simulations undertaken at the same place and time, this study demonstrates how improvements can be made to the representation of cloud properties. The results from this study will assist in the design of future cloud missions to provide a better quality input. These improvements will also help improve weather predictions and lower the uncertainty in cloud feedback response to increasing atmospheric temperature. Most clouds are difficult to monitor by more than one instrument due to continuous changes in: large-scale and sub-cloud scale circulation features, microphysical properties and processes and characteristic chemical signatures. This study undertakes co-located comparisons of high cloud data with a cloud ice dataset reported from the Microwave Limb Sounder (MLS) instrument onboard the Aura satellite that forms part of the A-train constellation. Data from the MLS science team include vertical profiles of temperature, ice water content (IWC) and the mixing ratios of several trace gases. Their vertical resolutions are 3 to 6 km. Initial investigations explore the link between cloud-top properties and the longwave radiation budget, developing methods for estimating cloud top heights using; longwave radiative fluxes, and IWC profiles. Synergistic trios of direct and indirect high cloud measurements were used to validate detections from the MLS by direct comparisons with two different A-train instruments; the NASA Moderate-resolution Imaging Spectroradiometer (MODIS) and the Clouds and the Earth’s Radiant Energy System (CERES) onboard on the Aqua satellite. This finding focuses later studies on two high cloud scene types that are well detected by the MLS; deep convective plumes that form from moist ascent, and their adjacent outflows that emanate outwards several hundred kilometres. The second part of the thesis identifies and characterises two different high cloud scenes in the tropics. Direct observational data is used to refine calculations of the climate sensitivity to upper tropospheric humidity and high cloud in different conditions. The data reveals several discernible features of convective outflows are identified using a large sample of MLS data. The key finding, facilitated by the use of co-location, reveals that deep convective plumes exert a large longwave warming effect on the local climate of 52 ± 28Wm−2, with their adjacent outflows presenting a more modest warming of 33 ± 20Wm−2.

Giant molecular clouds (GMCs), out of which stars form, are complex, dynamic systems, which both influence and are shaped by the process of star formation. In this dissertation, I examine three different facets of the dynamical motions within GMCs. Collapse modes in different dimensional objects. Molecular clouds contain lower dimensional substructures, such as filaments and sheets. The collapse properties of finite filaments and sheets differ from those of spherical objects as well as infinite sheets and filaments. I examine the importance of local collapse modes of small central perturbations, relative to global collapse modes, in different dimensional objects to elucidate whether strong perturbations are required for molecular clouds to fragment to form stars. I also calculate the dependence of the global collapse timescale upon the aspect ratio of sheets and filaments. I find that lower dimensional objects are more readily fragmented, and that for a constant density, lower dimensional objects and clouds with larger aspect ratios collapse more slowly. An edge-driven collapse mode also exists in sheets and filaments and is most important in elongated filaments. The failure to consider the geometry of a gas cloud is shown to lead to an overestimation of the star formation rate by up to an order of magnitude. Molecular tracers of turbulent energy dissipation. Molecular clouds contain supersonic turbulence that simulations predict will decay rapidly via shocks. I use shock models to predict which species emit the majority of the turbulent energy dissipated in shocks and find that carbon monoxide, CO, is primarily responsible for radiating away this energy. By combining these shock models with estimates for the turbulent energy dissipation rate of molecular clouds, I predict the expected shock spectra of CO from molecular clouds. I compare the results of these shock models to predictions for the emission from the unshocked gas in GMCs and show that mid-to-high rotational transitions of CO (e.g., J = 8 to 7), should be dominated by shocked gas emission and should trace the turbulent energy being dissipated in molecular clouds. Orion-Eridanus superbubble. The nearby Orion star forming region has created a large bubble of hot plasma in the local interstellar medium referred to as the Orion-Eridanus superbubble. This bubble is unusual in that it is highly elongated, is believed to be oriented roughly parallel to the galactic plane, and contains bright filamentary features on the Eridanus side. I fit models for a wind driven bubble in an exponential atmosphere to the Orion-Eridanus superbubble and show that the elongation of the bubble cannot be explained by such a model in which the scale height of the galactic disk is the typical value of 150 pc. Either a much smaller scale height must be adopted or some additional physics must be added to the model. I also show that the Eridanus filaments cannot be equilibrium objects ionized by the Orion star forming region. / Graduate / 0606 / andyrpon@gmail.com

This thesis work deals with the most actual topic in Computer Vision field which is scene understanding and this using matching of 3D feature point images. The objective is to make use of Saab’s latest breakthrough in extraction of 3D feature points, to identify the best alignment of at least two 3D feature point images. The thesis gives a theoretical overview of the latest algorithms used for feature detection, description and matching. The work continues with a brief description of the simultaneous localization and mapping (SLAM) technique, ending with a case study on evaluation of the newly developed software solution for SLAM, called slam6d. Slam6d is a tool that registers point clouds into a common coordinate system. It does an automatic high-accurate registration of the laser scans. In the case study the use of slam6d is extended in registering 3D feature point images extracted from a stereo camera and the results of registration are analyzed. In the case study we start with registration of one single 3D feature point image captured from stationary image sensor continuing with registration of multiple images following a trail. Finally the conclusion from the case study results is that slam6d can register non-laser scan extracted feature point images with high-accuracy in case of single image but it introduces some overlapping results in the case of multiple images following a trail.

The Magellanic Clouds (MCs) consist of a pair of galaxies, the Large Magellanic Cloud (LMC) and the Small Magellanic Cloud (SMC), which are located at a distance of 50 kpc and 60 kpc, with stellar masses of 1010 M and 109 M , respectively. Morphologically they are categorized as irregular type galaxies. The MCs are gas rich and metal poor (Z=0.008 for LMC, and 0.004 for SMC) as compared to the Milky Way (MW), and have active star-forming regions. Their proximity and location at high galactic latitude enable us to resolve their individual populations as well as detect faint stellar populations. It is well known that the MCs are interacting with each other, as well as with the MW. The interaction is supported by the presence of the Magellanic Bridge and the Magellanic Stream. The evolved stellar populations in the MCs help us to understand their evolution and interaction process. The MCs host both Population I as well as Population II stars. This extended range of star formation is a valuable source of information to understand the formation and evolution of galaxies in general, and the MCs in particular. Evolved stellar popu-lation means the stars that have evolved o the main sequence and the giants, such as red giants (RGs), red clump stars, and asymptotic giant branch stars. There is a dominant population of evolved stars present in the MCs, in star clusters as well as in the eld. The aim of the thesis is to study the evolved stellar populations for one of the component of the MCs, the LMC. The study is primarily divided into two parts. (1) Study of sparse star clusters in the LMC: To increase our understanding of sparse star clusters in the LMC, with well estimated parameters, using deep Washington photometric data for 45 LMC clusters. (2) To estimate a metallicity map of LMC: In order to understand the metallicity variation across the galaxy. This is done by creating a high spatial resolution metallicity map of the LMC, using red giant branch (RGB) stars, with the help of photometric data and calibrated using spectroscopic studies of RGs in eld and star clusters. The introduction to the thesis study along with the aim are described in Chapter 1 of the thesis. The three sets of photometric data used for this study are described in Chapter 2. The data sets are: CT1 Washington photometric data for 45 star clusters within the LMC, the VI photometric data from the Optical Gravitational Lensing Experiment Phase-III survey (OGLE III), and the Magellanic Cloud Photometric Survey (MCPS). Study of sparse star clusters in the LMC: A systematic study is per-formed to analyse the 45 cluster candidates, to estimate their parameters (radius, reddening, and age) using the main-sequence turn-o (MSTO), as well as the evolved portion of the colour{magnitude diagram (CMD). The basic parameters were estimated for 33 genuine clusters, whereas the other 12 cluster candidates have been classi ed as possible clusters/asterisms. The study of 33 star clusters are presented in Chapter 3. These clus-ters are categorized as genuine star clusters based on their strong density enhancement and cluster features with respect to their surrounding eld regions. Out of the 33 clusters, 23 are identi ed as single clusters and 10 are found to be members of double clusters. Detailed discussions of all the individual clusters are presented. The estimated parameters for the single and double clusters are listed in two di erent tables. About 50% of the clusters are in the age range 100{300 Myr, the rest of them being older or younger. Comparison with previous age estimates shows some agreement as well as some deviation. The remaining 12 clusters which could not be categorized as genuine star clusters are studied in Chapter 4. These clusters have poor (/suspi-cious) density enhancement and cluster features when compared to their surrounding elds. It is important to study such cluster candidates, as these objects probe the lower limit of the cluster mass function. Detailed discussion on these individual objects are presented and their estimated parameters are tabulated in this chapter. A detailed discussion based on the study of all the 45 inconspicuous clusters is presented in this chapter, including the estimated sizes (radii 2{10 pc), reddening with respect to eld, and location in the LMC. The mass limit estimated for genuine clusters is found to be 1000 M , whereas for possible clusters/asterisms it is few 100 M , using synthetic CMDs. The study of sparse clusters enlarged the number of objects con rmed as genuine star clusters (33) and estimated their fundamental parameters. The study emphasizes that the sizes and masses of the studied sample are found to be similar to that of open clusters in the MW. Thus, this study adds to the lower end of cluster mass distribution in the LMC, suggesting that the LMC, apart from hosting rich clusters, also has formed small, less massive open clusters in the 100{300 Myr age range. The 12 cases of possible clusters/asterisms are worthy of attention, in the sense that they can throw light on the survival time of such objects in the LMC. Photometric metallicity map of the LMC using RGB stars: A metallic-ity map of the LMC is estimated using OGLE III and MCPS photometric data. This is a rst of its kind map of metallicity up to a radius of 4{5 de-grees, derived using photometric data and calibrated using spectroscopic data of RGB stars. The RGB is identi ed in the V, (V I) CMDs of small areal subregions of varying sizes in both data sets. The slope of the RGB is used as an indicator of the average metallicity of a subregion, and this RGB slope is calibrated to metallicity using spectroscopic data for eld and cluster RGs in selected subregions. The metallicity map estimated using OGLE III photometric data is presented in Chapter 5. A method to identify the RGB of small subre-gions within the LMC and estimate its slope by using a consistent and automated method was developed. The technique is robust and indepen-dent of reddening and extinction. The details of calibrating the RGB slopes to metallicities, using previous spectroscopic results of RGs in eld and star clusters are presented. The OGLE III metallicity maps are pre sented, based on four cut-o criteria to separate regions with good ts. The OGLE III map has substantial coverage of the bar, the eastern and western LMC, but does not cover the northern and southern regions. The OGLE III metallicity map shows the bar region to be metal rich whereas the eastern and western regions to be relatively metal poor. The mean metallicity is estimated for three di erent regions within the LMC. For the complete LMC the mean [Fe/H] is = 0.39 dex ( [Fe/H] = 0.10); for the bar region it is = 0.35 dex ( [Fe/H] = 0.9); and for the outer LMC it is = 0.46 dex ( [Fe/H] = 0.11). The metallicity histogram for these di erent regions are also estimated. A radial metallicity gradient is estimated in the de-projected plane of the LMC. The metallicity gradient is seen to remain almost constant in the bar region (till a radius of 2.5 kpc) and has a shallow gradient of 0.066 0.006 dex kpc 1 beyond that till 4 kpc. In Chapter 6 the metallicity map based on MCPS photometric data is estimated. The MCPS data covers more of the northern and south-ern LMC (less of eastern and western regions) and is important to be analysed in order to reveal the metallicity trend of the overall disk. The systematic di erences between the lter systems of MCPS and OGLE III are corrected, and the MCPS slopes are then calibrated using the OGLE III slope{metallicity relation. The MCPS metallicity maps are presented, based on four cut-o criteria to separate regions with good ts. The bar region is found to be metal rich as was found using OGLE III data, whereas the northern and southern regions are marginally metal poor. The mean metallicity estimated for the complete LMC is = 0.37 dex ( [Fe/H] = 0.12); and for the outer LMC it is = 0.41 dex ( [Fe/H] = 0.11). The metallicity histogram for these di erent regions are estimated and compared with the OGLE III distribution. The metallicity range of the complete LMC is found to be almost similar for both data sets. The metallicity distribution within the bar has a narrow range as found using both data sets. The slight di erence between mean metallicity of outer LMC for the two data sets is attributed to their coverage. We suggest that the northern and southern regions of the LMC could be marginally more metal rich than the eastern and western regions. The metallicity gradient of the LMC disk, estimated from MCPS data is found to be shallow 0.049 0.002 dex kpc 1 till about 4 kpc. We also constructed a metallicity map of outliers using both OGLE III and MCPS data, and identi ed subregions where the mean metallic-ity di ers from the surrounding areas. We suggest further spectroscopic studies in order to assess their physical significance. The detailed conclusion of the thesis and future work are presented in Chapter 7. From the study of sparse star clusters in the LMC, it is concluded that LMC has open cluster like star cluster systems. It is important to include them to understand the cluster formation history (CFH) and their survival time scale. Presently, our understanding of the CFH is dominated by rich clusters. The bar of the LMC is found to be the most metal rich region, and the LMC metallicity gradient though shallow, resembles the gradient seen in spiral galaxies. The gradient is also similar to that found in our Galaxy. The higher metallicity in the bar region might indicate an active bar in the past.

Star and stellar cluster formation in spiral galaxies is one of the biggest questions of astrophysics. In this thesis, I study how star formation, and the formation of stellar clusters, proceeds using SPH simulations. These simulations model a region of 400 pc and 107 solar masses. Star formation is modelled through the use of sink particles which represent small groups of stars. Star formation occurs in high density regions, created by galactic spiral arm passage. The spiral shock compresses the gas and generates high density regions. Once these regions attain sufficiently high density, self-gravity becomes dominant and drives collapse and star formation. The regions fragment hierarchically, forming local small groups of stars. These fall together to form clusters, which grow through subsequent mergers and large scale gas infall. As the individual star formation occurs over large distances before forming a stellar cluster, this process can result in significant age spreads of 1-2 Myrs. One protocluster is found to fail to merge due to the large scale tidal forces from the nearby regions, and instead expands forming a dispersed population of young stars such as an OB association.

241 f. / Submitted by PPGE PPGE (pgedu@ufba.br) on 2013-09-18T16:18:26Z No. of bitstreams: 1 TESE TODA 6 abril Virgínia 13 MAIO 13.pdf: 2417557 bytes, checksum: 861d741b496474717e052ea3d3e6c85b (MD5) / Rejected by Maria Auxiliadora Lopes(silopes@ufba.br), reason: Prezado, Favor renomear o arquivo. Auxiliadora on 2013-09-18T16:51:19Z (GMT) / Submitted by PPGE PPGE (pgedu@ufba.br) on 2013-09-23T13:12:56Z No. of bitstreams: 1 SANDRA REGINA ROSA.pdf: 2417557 bytes, checksum: 861d741b496474717e052ea3d3e6c85b (MD5) / Approved for entry into archive by Maria Auxiliadora Lopes(silopes@ufba.br) on 2013-09-23T16:43:43Z (GMT) No. of bitstreams: 1 SANDRA REGINA ROSA.pdf: 2417557 bytes, checksum: 861d741b496474717e052ea3d3e6c85b (MD5) / Made available in DSpace on 2013-09-23T16:43:43Z (GMT). No. of bitstreams: 1 SANDRA REGINA ROSA.pdf: 2417557 bytes, checksum: 861d741b496474717e052ea3d3e6c85b (MD5) Previous issue date: 2013 / A Audiodescrição, ou AD é um recurso que visa tornar acessível ao público deficiente visual conteúdos imagéticos produzidos no âmbito educacional e cultural. No Brasil, as ADs são realizadas ainda experimentalmente, a partir da experiência do convívio com deficientes visuais ou de um modelo fundamentado nas normas britânica (ITC, 2000), espanhola (UNE, 2005) e americana (ADC, 2008). Ao seguir tais modelos, as ADs apontam para um padrão internacional, o qual prioriza a objetividade, a clareza e a fidelidade à obra transcrita. Este estudo tem como objetivo analisar o alcance de duas versões de AD realizadas para o filme português Atrás das Nuvens (2007) de Jorge Queiroga, fixando o olhar na poética produzida pela Linguagem Cinematográfica – LC nesta obra. Também se dispõe a discutir a questão da objetividade, expressividade e poética passadas nas versões abordadas. Para fundamentar a tese, é feito um estudo de caso com base na pesquisa qualitativa, alicerçada na análise de um trecho desse filme e aportada por entrevista semiestruturada com um grupo de pessoas deficientes visuais. Teoricamente, a discussão foi mediada, principalmente, nos trabalhos de Diniz (2007); Franco (2010); Neves (2011); Gomes (2004); Deleuze (2005) e Minayo (2001). Identificou-se que a AD não pode ser realizada apenas como um serviço de tradução de forma mecânica, identificando imagens, no intuito de favorecer ao espectador deficiente visual a captação apenas de forma instantânea. Os resultados demonstraram ser possível realizar a AD a partir da força embutida na poética da LC e transmiti-la de forma expressiva, criativa e poética. / Salvador

Esta tese tem por objetivo propor uma metodologia para recuperação de perfis verticais de temperatura na atmosfera com nuvens a partir de medidas de radiância feitas por satélite, usando redes neurais artificiais. Perfis verticais de temperatura são importantes condições iniciais para modelos de previsão de tempo, e são usualmente obtidos a partir de medidas de radiâncias feitas por satélites na faixa do infravermelho. No entanto, quando estas medidas são feitas na presença de nuvens, não é possível, com as técnicas atuais, efetuar a recuperação deste perfil. É uma perda significativa de informação, pois, em média, 20% dos pixels das imagens acusam presença de nuvens. Nesta tese, este problema é resolvido como um problema inverso em dois passos: o primeiro passo consiste na determinação da radiância que atinge a base da nuvem a partir da radiância medida pelos satélites; o segundo passo consiste na determinação do perfil vertical de temperaturas a partir da informação de radiância fornecida pelo primeiro passo. São apresentadas reconstruções do perfil de temperatura para quatro casos testes. Os resultados obtidos mostram que a metodologia adotada produz resultados satisfatórios e tem grande potencial de uso, permitindo incorporar informações sobre uma região mais ampla do globo e, consequentemente, melhorar os modelos de previsão do tempo. / This thesis presents a methodology for retrieving vertical temperature profiles in the atmosphere with clouds from radiance measurements made by satellite, using artificial neural networks. Vertical temperature profiles are important initial conditions for numerical weather prediction models, and are usually obtained from measurements of radiance using infrared channels. Though, when these measurements are performed in the atmosphere with clouds, it is not possible to retrieve the temperature profile with current techniques. It is a significant loss of information, since on average 20% of the pixels of the images have clouds. In this thesis, this problem is solved as a two-step inverse problem: the first step is an inverse problem of boundary condition estimation, where the radiance reaching the cloud basis is determined from radiance measured by satellite; the second step consists in determining the vertical temperature profile from the boundary condition estimated in the first step. Reconstructions of temperature profile are presented for four test cases. The results show that the proposed methodology produces satisfactory results and has great potential for use, allowing to incorporate information from a wider area of the planet and thus to improve numerical weather prediction models.

L'existence de molécules interstellaires soulève une question, ces molécules sont-elles les mêmes molécules que nous voyons dans le système Solaire aujourd'hui ? C'est une question toujours ouverte qui implique des conséquences profondes. Il est possible d’éclaircir cette question en étant capables de retracer l'héritage d'un groupe de molécules chimiquement liées, ce que nous appelons un réservoir. Le meilleur outil pour retracer l'héritage des réservoirs sont les rapports isotopiques. L'élément qui montre les plus grandes variations du rapport isotopique dans le système Solaire est l'azote. Ces variations indiquent que le rapport isotopique de l'azote est sensible aux conditions physiques de la formation des étoiles.L'objectif principal de cette thèse est d'identifier les réservoirs d'azote à différents étapes de la formation des étoiles et des planètes. La première étape de cette entreprise était d'identifier le rapport isotopique de la masse principal d'azote du milieu interstellaire local aujourd'hui.Cela a été déterminé égale à 323 ± 30 à partir du rapport CN/C 15 N mesuré dans le disque protoplanétaire autour de TW Hya. Parallèlement à cela, nous avons également mesuré un rapport HCN/HC 15 N=128 ± 36 dans le disque protoplanétaire autour de MWC 480. Ces rapports isotopiques très distinctes mesurées sur les disques protoplanétaires sont une indication claire de la présence d'au moins deux réservoirs d'azote dans les disques protoplanétaires. La façon dont ces réservoirs se séparent est cependant inconnue. Cela pourrait peut-être se produire en raison de réactions de fractionnement chimique ayant lieu dans les cœurs prestellaires. Nous avions donc comme objectif d'obtenir une mesure précise et directe du rapport isotopique de l'azote des molécules d'HCN dans le cœur prestellaire L1498.Pour obtenir cette mesure, l'obstacle le plus important à surmonter était due aux anomalies hyperfines des molécules d'HCN. Ces anomalies hyperfines sont induites par le chevauchement des composants hyperfins. Ceci sont particulièrement sensibles à la densité de colonne d'HCN, mais aussi au champ de vitesses et aux largeurs de raies. Ainsi les anomalies hyperfines sont un outil de mesure de l'abondance d'HCN permettant aussi de sonder la cinématique des cœurs prestellaires.Pour reproduire avec précision les anomalies hyperfines, et ainsi mesurer des densités de colonne précises d'HCN, nous avions besoin d'explorer un espace de paramètres dégénéré de 15 dimensions. Pour minimiser les dégénérescences nous avons obtenu un profil de densité basé sur des cartes du continuum de L1498. Ceci permettant de réduire à 12 dimensions l'espace des paramètres. L'exploration de cet espace de paramètre a été fait grâce à l'utilisation d'un méthode de minimisation MCMC. Grâce à cette exploration, nous avons obtenu HCN/HC 15 N = 338 ± 28 et HCN/H 13 CN = 45 ± 3. Les incertitudes sur ces valeurs sont limités par les erreurs de calibration et sont dé-terminés de manière non arbitraire par le méthode MCMC. Les implications de ces résultats sont discutées dans le chapitre de conclusion,où nous présentons également quelques perspectives sur l'avenir. / The existence of interstellar molecules raises the question, are thesemolecules the same molecules we see on the Solar system today? Thisis still an open question with far reaching consequences. Some lightmay be shed on this issue if we are able to trace the heritage of agroup of chemically linked molecules, a so-called reservoir. The besttool to trace the heritage of reservoirs are isotopic ratios. The elementthat shows the largest isotopic ratio variations in the Solar system isnitrogen. For this is an indication that the isotopic ratio of nitrogen issensitive to the physical conditions during star formation.The main objective of this thesis is to identify the reservoirs of ni-trogen at different stages of star and planet formation. The first stepin this endeavour was to identify the isotopic ratio of the bulk of ni-trogen in the local ISM today. This was determined to be 323 ± 30from the CN/C 15 N ratio in the protoplanetary disk around TW Hya.Along with it we also measured the HCN/HC 15 N= 128 ± 36 in theprotoplanetary disk around MWC 480. This very distinct nitrogen iso-topic ratios on protoplanetary disks are a clear indication that thereare at least two reservoirs of nitrogen in protoplanetary disks. Howthese reservoirs get separated is however unknown. This could pos-sibly happen due to chemical fractionation reactions taking place inprestellar cores. We therefore aimed to obtain an accurate direct mea-surement of the nitrogen isotopic ratio of HCN in the prestellar coreL1498.To obtain this measurement the most important hurdle to overcomewere the hyperfine anomalies of HCN. These hyperfine anomaliesarise due to the overlap of hyperfine components. They are especiallysensitive to the column density of HCN, but also to the velocity fieldand line widths. Thus hyperfine anomalies are a tool to measure theabundance of HCN and to probe the kinematics of prestellar cores.To accurately reproduce the hyperfine anomalies, and thus mea-sure accurate column densities for HCN, we needed to explore adegenerate parameter space of 15 dimensions. To minimise the de-generacies we have derived a density profile based on continuummaps of L1498. This reduced the parameter space to 12 dimensions.The exploration of this parameter space was done through the useof a MCMC minimisation method. Through this exploration we ob-tained HCN/HC 15 N = 338 ± 28 and HCN/H 13 CN = 45 ± 3. Theuncertainties on these values are calibration limited and determinednon-arbitrarily by the MCMC method. Implications of these resultsare discussed in the concluding chapter, where we also present somefuture perspectives.

Esta tese tem por objetivo propor uma metodologia para recuperação de perfis verticais de temperatura na atmosfera com nuvens a partir de medidas de radiância feitas por satélite, usando redes neurais artificiais. Perfis verticais de temperatura são importantes condições iniciais para modelos de previsão de tempo, e são usualmente obtidos a partir de medidas de radiâncias feitas por satélites na faixa do infravermelho. No entanto, quando estas medidas são feitas na presença de nuvens, não é possível, com as técnicas atuais, efetuar a recuperação deste perfil. É uma perda significativa de informação, pois, em média, 20% dos pixels das imagens acusam presença de nuvens. Nesta tese, este problema é resolvido como um problema inverso em dois passos: o primeiro passo consiste na determinação da radiância que atinge a base da nuvem a partir da radiância medida pelos satélites; o segundo passo consiste na determinação do perfil vertical de temperaturas a partir da informação de radiância fornecida pelo primeiro passo. São apresentadas reconstruções do perfil de temperatura para quatro casos testes. Os resultados obtidos mostram que a metodologia adotada produz resultados satisfatórios e tem grande potencial de uso, permitindo incorporar informações sobre uma região mais ampla do globo e, consequentemente, melhorar os modelos de previsão do tempo. / This thesis presents a methodology for retrieving vertical temperature profiles in the atmosphere with clouds from radiance measurements made by satellite, using artificial neural networks. Vertical temperature profiles are important initial conditions for numerical weather prediction models, and are usually obtained from measurements of radiance using infrared channels. Though, when these measurements are performed in the atmosphere with clouds, it is not possible to retrieve the temperature profile with current techniques. It is a significant loss of information, since on average 20% of the pixels of the images have clouds. In this thesis, this problem is solved as a two-step inverse problem: the first step is an inverse problem of boundary condition estimation, where the radiance reaching the cloud basis is determined from radiance measured by satellite; the second step consists in determining the vertical temperature profile from the boundary condition estimated in the first step. Reconstructions of temperature profile are presented for four test cases. The results show that the proposed methodology produces satisfactory results and has great potential for use, allowing to incorporate information from a wider area of the planet and thus to improve numerical weather prediction models.