Cropland management practices have traditionally focused on maximising the production of food, feed and fibre. However, croplands also provide valuable regulating ecosystem services, including carbon (C) storage in soil and biomass. Consequently, management impacts the extents to which croplands act as sources or sinks of atmospheric carbon dioxide (CO2). And so, reliable information on cropland ecosystem C fluxes and yields are essential for policy-makers concerned with climate change mitigation and food security. Eddy-covariance (EC) flux towers can provide observations of net ecosystem exchanges (NEE) of CO2 within croplands, however the tower sites are temporally and spatially sparse. Process-based crop models simulate the key biophysical mechanisms within cropland ecosystems, including the management impacts, crop cultivar, soil and climate on crop C dynamics. The models are therefore a powerful tool for diagnosing and forecasting C fluxes and yield. However, crop model spatial upscaling is often limited by input data (including meteorological drivers and management), parameter uncertainty and model complexity. Earth observation (EO) sensors can provide regular estimates of crop condition over large extents. Therefore, EO data can be used within data assimilation (DA) schemes to parameterise and constrain models. Research presented in this thesis explores the key challenges associated with crop model upscaling. First, fine-scale (20-50 m) EO-derived data, from optical and radar sensors, is assimilated into the Soil-Plant-Atmosphere crop (SPAc) model. Assimilating all EO data enhanced the simulation of daily C exchanges at multiple European crop sites. However, the individually assimilation of radar EO data (as opposed to combined with optical data) resulted in larger improvements in the C fluxes simulation. Second, the impacts of reduced model complexity and driver resolution on crop photosynthesis estimates are investigated. The simplified Aggregated Canopy Model (ACM) – estimating daily photosynthesis using coarse-scale (daily) drivers – was calibrated using the detailed SPAc model, which simulates leaf to canopy processes at half-hourly time-steps. The calibrated ACM photosynthesis had a high agreement with SPAc and local EC estimates. Third, a model-data fusion framework was evaluated for multi-annual and regional-scale estimation of UK wheat yields. Aggregated model yield estimates were negatively biased when compared to official statistics. Coarse-scale (1 km) EO data was also used to constrain the model simulation of canopy development, which was successful in reducing the biases in the yield estimates. And fourth, EO spatial and temporal resolution requirements for crop growth monitoring at UK field-scales was investigated. Errors due to spatial resolution are quantified by sampling aggregated fine scale EO data on a per-field basis; whereas temporal resolution error analysis involved re-sampling model estimates to mimic the observational frequencies of current EO sensors and likely cloud cover. A minimum EO spatial resolution of around 165 m is required to resolve the field-scale detail. Monitoring crop growth using EO sensors with a 26-day temporal resolution results in a mean error of 5%; however, accounting for likely cloud cover increases this error to 63%.

LOBO, A. K. M. Photosynthesis regulation by sucrose metabolism under water deficit and source-sink alterations in sugarcane. 2016. 118 f. Tese (Doutorado em Bioquímica)-Universidade Federal do Ceará, Fortaleza, 2016. / Submitted by Anderson Silva Pereira (anderson.pereiraaa@gmail.com) on 2017-01-03T20:20:30Z No. of bitstreams: 1 2016_tese_akmlobo.pdf: 4541762 bytes, checksum: e3450d87e286b1a5213eb866921b6c94 (MD5) / Approved for entry into archive by Jairo Viana (jairo@ufc.br) on 2017-01-09T21:32:52Z (GMT) No. of bitstreams: 1 2016_tese_akmlobo.pdf: 4541762 bytes, checksum: e3450d87e286b1a5213eb866921b6c94 (MD5) / Made available in DSpace on 2017-01-09T21:32:52Z (GMT). No. of bitstreams: 1 2016_tese_akmlobo.pdf: 4541762 bytes, checksum: e3450d87e286b1a5213eb866921b6c94 (MD5) Previous issue date: 2016 / Water deficit stress is the major limiting factor for plant growth and development, constraining food production. In order to survive in such dry conditions, many biochemical and physiological changes must be triggered by plants. In general, the responses to drought are loss of water content, reductions of stomatal conductance and photosynthesis and increase of carbohydrates. Soluble sugars play a key role in plant metabolism, acting as substrates and modulators of enzyme activity in carbon-related pathways and controlling the expression of different genes related to carbon, lipid and nitrogen routs. However, the mechanisms involved with photosynthesis down-regulation by drought and sugars in C4 plants are not fully understood. The aim of this study was to investigate how drought and source-sink perturbation regulate photosynthesis in sugarcane plants. Therefore, two studies were conducted with sugarcane plants with four months old cultivated under greenhouse conditions. In the first study sugarcane plants (cv. IACSP94-2094) were subjected to water deficit for 5 days (WD) with concomitant spraying of 50 mM exogenous sucrose (WD + Suc). While in the second study source-sink relationship was perturbed in two sugarcane cultivars (cv. IACSP94-2094 and cv. IACSP95-5000) by imposing partial darkness, spraying 50 mM exogenous sucrose and their combination for 5 days. The negative effects of WD in the gas exchange and photochemical parameters were aggravated by exogenous sucrose. Photosynthesis reductions were related to both stomatal and biochemical limitations, but exogenous sucrose intensified metabolic restrictions mainly through down-regulation of Rubisco initial activity and PSII effective quantum efficiency in drought-stressed plants. In addition, Rubisco activation state was decreased by WD + Suc, indicating perhaps that the activity of this enzyme was reduced by tight-binding inhibitors, such as sugars phosphates. Sucrose metabolism enzymes and sugars amount were also differently altered by WD and WD + Suc in leaves, sheath and stalk in WD and WD +Suc plants. Interestingly, Sucrose/hexose ratio decreased in both leaf and sheath whereas it was increased in stalk, suggesting that sucrose and related sugars were intensely metabolized and transported in drought-stressed plants. In well-watered conditions, photosynthesis was inhibited by sucrose spraying in both genotypes, through decreases in maximum Rubisco carboxylation rate (Vcmax), initial slope of A-Ci curve (k), stomatal conductance (gs) and ATP production driven by electron transport (Jatp). The partial darkness and sucrose spraying combination did not change photosynthesis in both genotypes. Significant increases in Vcmax, gs and Jatp and marginal increases in k were noticed when combining partial darkness and sucrose spraying compared with sucrose spraying alone. Altogether, these results suggest that CO2 assimilation impairment is aggravated by exogenous sucrose in drought-stressed plants. This limitation was mainly related to biochemical restrictions, specially associated with Rubisco initial activity and PSII quantum efficiency. In contrast, in vitro PEPCase activity and amount were increased in sucrose-treated plants, suggesting that C4 cycle efficiency was reduced in vivo by C3 cycle inhibition under drought conditions. Moreover, sucrose amount was increased in the stalk, suggesting the feedback regulation from stalk to source leaves in drought-stressed plants. Our data also revealed that increases in sink strength due to partial darkness offset the inhibition of sugarcane photosynthesis caused by sucrose spraying, enhancing the knowledge on endogenous regulation of sugarcane photosynthesis through the source-sink relationship. / A deficiência hídrica é o principal fator limitante para o crescimento e desenvolvimento das culturas. Para sobreviver nessas condições adversas, várias modificações bioquímicas e fisiológicas são desencadeadas pelas plantas. Em geral, os efeitos da seca em plantas são diminuição do status hídrico, reduções da condutância estomática, fotossíntese e crescimentos e aumentos nos níveis de carboidratos. Os açúcares solúveis desempenham papéis chave no metabolismo das plantas, atuando como substratos e moduladores da atividade enzimática em vias relacionadas com o carbono. Além disso, os açúcares controlam a expressão de genes associados com as rotas do metabolismo do carbono, lipídios e nitrogênio. Entretanto, os mecanismos envolvidos com a regulação negativa da fotossíntese por deficiência hídrica e açúcares em plantas C4 não estão totalmente entendidos. O objetivo deste estudo foi investigar como a deficiência hídrica e perturbações na relação fonte-dreno regulam a fotossíntese em plantas de cana-de-açúcar. Dois estudos foram conduzidos com plantas de cana-de-açúcar com quatro meses de idade cultivadas sob condições de casa de vegetação. No primeiro estudo, plantas de cana-de-açúcar (cv. IACSP94-2094) foram submetidas a deficiência hídrica por 5 dias (WD) com subsequente aplicação de sacarose exógena 50 mM (WD + Suc). Enquanto que no segundo estudo a relação fonte-dreno foi perturbada em duas cultivares de cana-de-açúcar (cv. IACSP94-2094 and cv. IACSP95-5000) pela imposição parcial de sombreamento, aplicação de sacarose exógena 50 mM e por suas combinações por 5 dias. Os efeitos negativos de WD nos parâmetros de trocas gasosas e fotoquímicos foram agravados por sacarose exógena. As reduções na fotossíntese foram relacionadas com limitações estomáticas e bioquímicas, porém a sacarose exógena intensificou as restrições bioquímicas principalmente por reduções na atividade inicial de Rubisco e eficiência quântica do PSII em plantas sob seca. Além disso, o estado de ativação de Rubisco foi inibido por WD + Suc, sugerindo que a atividade inicial dessa enzima foi possivelmente reduzida por inibidores que se ligam fortemente em seu sitio ativo, tais como açúcares fosfato. As enzimas do metabolismo de sacarose e a concentração de açúcares foram modificados diferentemente por WD e WD + Suc em folhas, bainha e colmo. Interessantemente, a relação sacarose/hexose decresceu em folhas e bainha, enquanto que no colmo essa relação aumentou, sugerindo que sacarose e outros açúcares relacionados foram intensamente metabolizados e transportados. Em condições irrigadas a fotossíntese foi inibida pela aplicação de sacarose nos dois genótipos, através de decréscimos da taxa máxima de carboxilação de Rubisco (Vcmax), inclinação inicial da curva A-Ci (k), condutância estomática (gs) e produção de ATP direcionada pelo transporte de elétrons (Jatp). A combinação de sombreamento parcial e sacarose não alterou a fotossíntese em ambos os genótipos. Significantes aumentos em Vcmax, gs, Jatp e k foram observados quando sombreamento parcial e sacarose foram combinados em comparação com plantas tratadas apenas com sacarose. Em conclusão, esses resultados sugerem que o impedimento da assimilação de CO2 é agravada por adição de sacarose exógena em plantas sob estresse hídrico. Essa limitação foi relacionada principalmente com restrições bioquímicas, especialmente associadas com reduções na atividade inicial de Rubisco e eficiência quântica do FSII. Em contraste, a atividade in vivo e concentração de PEPCase foram aumentadas em plantas tratadas com sacarose e estresse hídrico, sugerindo que a eficiência do ciclo C4 foi reduzida in vivo por inibições do ciclo C3 sob condições de seca. Além disso, o conteúdo de sacarose aumentou no colmo, indicando uma regulação de feedback do colmo para as folhas em plantas sob seca. Nossos dados revelam ainda que aumentos na força do dreno devido ao sombreamento parcial aliviaram os efeitos inibitórios na fotossíntese de cana-de-açúcar causados pela aplicação de sacarose, aumentando o conhecimento na regulação endógena da fotossíntese de cana-de-açúcar através da relação fonte-dreno.

The vermicomposting of coffee grounds shows great promise for urban areas and university campuses. Several studies have examined using coffee grounds as a substrate for vermicomposting, however, little is known about its effect on plant growth, yield and quality. Therefore, two studies were conducted to assess these effects on greenhouse spinach and field grown bell peppers. Coffee vermicompost (VC) was utilized in a greenhouse spinach study over two spring growing seasons (2011 and 2012). Coffee VC was added to a 1:1:1 (peat, soil, sand) medium by volume (0, 25, 50 and 75%) in 4.5 L clay pots and seeded with `Bloomsdale Longstanding' spinach (Spinacia oleracea). Growth parameters evaluated were fresh leaf weight (FLW), leaf area (LA), spinach plant height, number of leaves, fresh leaf ascorbic acid (AA) and dry leaf nitrate (DLN). In 2011, the highest FLW, LA, number of leaves, AA and nitrate content were seen at the 75% VC application rate. Many parameters were correlated: The FLW and the VC application rate (r=0.41, P<0.0001); the AA content in leaves and the VC application rate (r=0.60, P<0.0001); and the AA and nitrate content in leaves (r=0.45, P=0.011). In 2012, the greatest FLW, height and number of leaves were observed at the 50% VC application rate; the highest nitrate content at 75% VC application rate, with no difference in AA content. A field study was conducted over three seasons (2009–-2011) to compare four treatments (coffee VC, dairy compost, standard fertility (SFT) and no treatment) for their effects on growth, yield, and AA content of bell peppers. Coffee VC and dairy compost were applied to beds at the rate of 22 t/ha and SFT was applied as 212 kg/ha 12:12:12 (N:P:K). Parameters evaluated were total number and weight of marketable and cull (unmarketable) pepper fruits, plant height, leaf chlorophyll index, and fresh fruit AA content. There were no differences detected for pepper fruit yields or AA content, however, plant height and chlorophyll index were greater for the VC and SFT treatments than for the compost and control treatments. These results indicated that coffee VC can improve the yield of greenhouse spinach and that the AA content does not decrease with higher VC application rates, even as nitrate content increases. Results of the bell pepper field study indicated that the coffee VC treatment produces similar growth, yield and AA content as SFT.

O presente trabalho se propõe a realizar uma análise histórica que nos possibilite interrogar quais são as condições de produção e reprodução de certos modelos (idéias e práticas) sobre o que se instituiu denominar a questão racial brasileira. O objetivo deste trabalho é mapear a proliferação de uma série de discursos em torno da construção de um projeto nacional e civilizatório que teve como eixo principal a produção de um discurso racializado, ou seja, discutir de que modo certos fatores permitiram engendrar a produção de uma estranheza eficaz a partir da criação do que convencionamos chamar de elemento negro, constituindo-se como o representante mais eficaz desse espaço social destinado a demarcar um lugar de estranhamento (o outro como perigoso, anormal, diferente etc.). Utilizando como recurso analítico principal os trabalhos de Raimundo Nina Rodrigues e da Escola Baiana de Antropologia, discutimos como esse saber acadêmico possibilitou a formulação de um modelo psicofísico de explicação sobre a degeneração da raça brasileira. Mais tarde este modelo seria substituído por uma estratégia mais englobante, o que pode ser verificado pela aplicação dos conceitos de cultura ou aculturação, e mesmo pelo emprego dos modernos conceitos psicanalíticos. O nosso propósito consistiu em analisar a produção de certas práticas sociais: a constituição de uma ciência médico-psicológica; a difusão de certas opiniões a respeito do elemento negro através da imprensa e da literatura; a constituição jurídica do cidadão negro em decorrência da implementação de uma discussão política e legislativa pré e pós-abolicionista que se produziram em torno da construção de um projeto nacional e civilizatório e que tiveram como eixo principal a produção do elemento negro como personagem principal desse novo enredo: uma ortodoxia da cor. Optamos por discutir o processo de formação do Brasil e do brasileiro em finais do século XIX e início do século XX (período compreendido entre as décadas de 1870 e 1930), a partir das rupturas provocadas pelo iminente processo abolicionista. / This work aims at performing a historical analysis towards questioning the production and reproduction conditions of certain models (ideas and practices) regarding what become to be known as the Brazilian race problem. The purpose of this work is the mapping of the proliferation of various discourses regarding the construction of a national and civilized project whose backbone was the formation of a racial speech. In other words, we want to discuss how certain factors contributed to the engineering of an \"efficient strangeness\" derived from the creation of the so-called black element. This element turned out to be the most efficient representative of the social space designated to determine a strangeness locale (the other as dangerous, abnormal, different, etc). We discussed, based primarily on the research works of Raimundo Nina Rodrigues and of the Escola Baiana de Antropologia, the means by which such an academic knowledge facilitated the formulation of an explicatory psychophysical model for the degeneration of the Brazilian race. This model was subsequently generalized, which can be verified by concepts of culture or \"acculturations\", as well as by modern psychoanalytical concepts. Our proposal was to analyze the production of certain social practices: the constitution of a medico-psychology science; the diffusion of certain opinions regarding the black element through the press and the literature; and the judicial constitution of the black citizen. These practices were the result of the implementation of pre- and post-abolitionist legislative and political discussions inspired by a national and civil project whose backbone was the production of the black element as the principal character of this new script: the orthodoxy of color. We opted to focus our discussion on the formation process of Brazil and of the Brazilian between the final decades of the 19th century and the first decades of the 20th century (between 1870 and 1930), characterized by the ruptures aggravated by the imminent abolitionist process.

abstract: The steady influx of Venezuelan immigrants to the United States has resulted in the creation of a close-knit community of these immigrants in the city of Doral, Florida, now nicknamed Doralzuela given the strong imprint Venezuelan have left in this city. This study aimed at gaining understanding on how the process of immigration and settlement in the context has affected Venezuelan immigrants’ identity, their perception and use of English and Spanish in daily interactions, and how, or if, their bonds with the home country has affected their incorporation to the host society. The study followed a qualitative design. Eight semi-structured interviews were conducted and analyzed following Riessman’s (2008) notion of dialogic narrative analysis. Six themes emerged from the data; (re)configuration of the self, the role of social networks, negotiating identity through language, issues of assimilation, transnational identity, and Doralzuela, the new Venezuela. These themes were discussed, and multiple and distinct views on each theme were identified. / Dissertation/Thesis / Masters Thesis English 2018

Ce travail porte sur la construction d'un modèle réduit en hydraulique fluviale avec une méthode de décomposition en polynôme du chaos. Ce modèle réduit remplace le modèle direct afin de réduire le coût de calcul lié aux méthodes ensemblistes en quantification d'incertitudes et assimilation de données. Le contexte de l'étude est la prévision des crues et la gestion de la ressource en eau. Ce manuscrit est composé de cinq parties, chacune divisée en chapitres. La première partie présente un état de l'art des travaux en quantification des incertitudes et en assimilation de données dans le domaine de l'hydraulique ainsi que les objectifs de la thèse. On présente le cadre de la prévision des crues, ses enjeux et les outils dont on dispose pour prévoir la dynamique des rivières. On présente notamment la future mission SWOT qui a pour but de mesurer les hauteurs d'eau dans les rivières avec un couverture globale à haute résolution. On précise notamment l'apport de ces mesures et leur complémentarité avec les mesures in-situ. La deuxième partie présente les équations de Saint-Venant, qui décrivent les écoulements dans les rivières, ainsi qu'une discrétisation numérique de ces équations, telle qu'implémentée dans le logiciel Mascaret-1D. Le dernier chapitre de cette partie propose des simplifications des équations de Saint-Venant. La troisième partie de ce manuscrit présente les méthodes de quantification et de réduction des incertitudes. On présente notamment le contexte probabiliste de la quantification d'incertitudes et d'analyse de sensibilité. On propose ensuite de réduire la dimension d'un problème stochastique quand on traite de champs aléatoires. Les méthodes de décomposition en polynômes du chaos sont ensuite présentées. Cette partie dédiée à la méthodologie s'achève par un chapitre consacré à l'assimilation de données ensemblistes et à l'utilisation des modèles réduits dans ce cadre. La quatrième partie de ce manuscrit est dédiée aux résultats. On commence par identifier les sources d'incertitudes en hydraulique que l'on s'attache à quantifier et réduire par la suite. Un article en cours de révision détaille la validation d'un modèle réduit pour les équations de Saint-Venant en régime stationnaire lorsque l'incertitude est majoritairement portée par les coefficients de frottement et le débit à l'amont. On montre que les moments statistiques, la densité de probabilité et la matrice de covariances spatiales pour la hauteur d'eau sont efficacement et précisément estimés à l'aide du modèle réduit dont la construction ne nécessite que quelques dizaines d'intégrations du modèle direct. On met à profit l'utilisation du modèle réduit pour réduire le coût de calcul du filtre de Kalman d'Ensemble dans le cadre d'un exercice d'assimilation de données synthétiques de type SWOT. On s'intéresse précisément à la représentation spatiale de la donnée telle que vue par SWOT: couverture globale du réseau, moyennage spatial entre les pixels observés. On montre notamment qu'à budget de calcul donné les résultats de l'analyse d'assimilation de données qui repose sur l'utilisation du modèle réduit sont meilleurs que ceux obtenus avec le filtre classique. On s'intéresse enfin à la construction du modèle réduit en régime instationnaire. On suppose ici que l'incertitude est liée aux coefficients de frottement. Il s'agit à présent de juger de la nécessité du recalcul des coefficients polynomiaux au fil du temps et des cycles d'assimilation de données. Pour ce travail seul des données in-situ ont été considérées. On suppose dans un deuxième temps que l'incertitude est portée par le débit en amont du réseau, qui est un vecteur temporel. On procède à une décomposition de type Karhunen-Loève pour réduire la taille de l'espace incertain aux trois premiers modes. Nous sommes ainsi en mesure de mener à bien un exercice d'assimilation de données. Pour finir, les conclusions et les perspectives de ce travail sont présentées en cinquième partie.

A presente tese define-se como uma reflexão acerca das políticas de identidade promovidas pelo estado colonial português e pela Frente de Libertação de Moçambique, com ênfase nos cem anos que antecederam a independência, proclamada em junho de 1975. Procurando uma perspectiva multidisciplinar, a análise é orientada por conceitos que procuram destacar fatores determinantes da concepção de dualismo inerente à situação colonial. A abordagem das várias estratégias culturais a que recorreu a metrópole para sustentar sua \"vocação\" imperial constitui um dado significativo do trabalho que procurou compreender algumas particularidades do projeto lusitano, com a preocupação de enquadrá-lo num processo mais amplo que não poderia desconsiderar os passos da História no ocidente. Partindo do estudo das duas concepções de assimilação e sua continuidade no luso-tropicalismo (e sua instrumentalização pelo Estado Novo português), a análise focaliza a gênese do nacionalismo e a nova dinâmica que a tática de guerrilha, implementada pela luta de libertação nacional, introduz no território de Moçambique. No que se refere à política de identidade nacional proposta pela FRELIMO, foi privilegiada pela pesquisa a dialética que ela estabelece com as sociedades tradicionais. / The present thesis deals with the identity policies promoted by the Portuguese colonial State and the Mozambique Liberation Front (FRELIMO), with emphasis on the last hundred years before independence, proclaimed in June 1975. Aiming at a multidisciplinary approach, the analysis is oriented by concepts that put in evidence determinant factors of the dualistic nature of colonial situation. The approach to the various cultural strategies used by Portugal to support its imperial \"vocation\" represents a significative part of this study. It tries to understand some details of the Portuguese project, by framing it within a wider perspective that could not neglect Western History. Starting from the study of the two conceptions of assimilation and its luso-tropicalistic development (the use of the theory carried out by the Portuguese \"New State\" regime) the analysis focus on the origins of nationalism as well as on the new dynamics introduced in the territory by the guerrilla tactics used during national liberation struggle. Concerning FRELIMO\'s national identity policy, this research privileges the dialectics it establishes with traditional societies of Mozambique.

Kimberlite hosts two populations of olivine that are distinguished on the basis of grain size and morphology; the populations are commonly described genetically as xenocrysts and phenocrysts. Recent studies of zoning patterns in kimberlitic olivine phenocrysts have cast doubt on the actual origins of the smaller olivine crystals. Here, we elucidate the nature and origins of the textural and chemical zonation that characterize both populations of olivine. Specifically, we show that both olivine-I and olivine-II feature chemically distinct overgrowths resulting from magmatic crystallization on pre-existing olivine xenocrysts. These results suggest that the total volume of olivine crystallized during transport is substantially lower (≤5%) than commonly assumed (e.g. ~25%), and that crystallization is dominantly heterogeneous. This reduces estimates of the Mg# in primitive kimberlite melt to more closely reconcile with measured phenocryst compositions. Several additional textures are observed in olivine, and include: sealed cracks, healed cracks, phases trapping in cracks, rounded grains, overgrowths and phase trapping in overgrowths. These features record processes that operate in kimberlite during ascent, and from these features we create a summary model for kimberlite ascent: • Olivine is incorporated into kimberlitic melts at great depths as peridotitic mantle xenoliths. • Shortly after the incorporation of these xenocrysts the tensile strength of the crystals within xenoliths is reached at a minimum of 20 km from its source. Disaggregation of mantle xenoliths producing xenocrysts is facilitated by expansion of the minerals within the xenoliths. • The void space produced by the failure of the crystals is filled with melt and crystals consisting of primary carbonate (high-Sr), chromite and spinel crystals. The carbonate later crystallizes to produce sealed fractures. • Subsequent decompression causes cracks that are smaller than the sealed cracks and are preserved as healed cracks that crosscut sealed cracks. • Mechanical rounding of the xenocrysts post-dates, and/or occurs contemporaneously with decompression events that cause cracking. • Saturation of olivine produces rounded overgrowths on large xenocrysts, euhedral overgrowths on smaller xenocrysts, and a volumetrically minor population of olivine phenocrysts. Olivine growth traps fluid, solid and melt inclusions. Calculations based on these relationships suggest that the melt saturates with olivine at a maximum depth of 20 km and a minimum depth of 7 km. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate

Reconstruire des champs géophysiques à partir d'observations bruitées et partielles est un problème classique bien étudié dans la littérature. L'assimilation de données est une méthode populaire pour aborder ce problème, et se fait par l'utilisation de techniques classiques, comme le filtrage de Kalman d’ensemble ou des filtres particulaires qui procèdent à une évaluation online du modèle physique afin de fournir une prévision de l'état. La performance de l'assimilation de données dépend alors fortement de du modèle physique. En revanche, la quantité de données d'observation et de simulation a augmenté rapidement au cours des dernières années. Cette thèse traite l'assimilation de données d'une manière data-driven et ce, sans avoir accès aux équations explicites du modèle. Nous avons développé et évalué l'assimilation des données par analogues (AnDA), qui combine la méthode des analogues et des méthodes de filtrage stochastiques (filtres Kalman, filtres à particules, chaînes de Markov cachées). Des applications aux modèles chaotiques simplifiés et à des études de cas de télédétection réelle (température de surface de lamer, anomalies du niveau de la mer), nous démontrons la pertinence d'AnDA pour l'interpolation de données manquantes des systèmes dynamiques non linéaires et à haute dimension à partir d'observations irrégulières et bruyantes.Motivé par l'essor du machine learning récemment, la dernière partie de cette thèse est consacrée à l'élaboration de modèles deep learning pour la détection et de tourbillons océaniques à partir de données de sources multiples et/ou multi temporelles (ex: SST-SSH), l'objectif général étant de surpasser les approches dites expertes. / Reconstructing geophysical fields from noisy and partial remote sensing observations is a classical problem well studied in the literature. Data assimilation is one class of popular methods to address this issue, and is done through the use of classical stochastic filtering techniques, such as ensemble Kalman or particle filters and smoothers. They proceed by an online evaluation of the physical modelin order to provide a forecast for the state. Therefore, the performanceof data assimilation heavily relies on the definition of the physical model. In contrast, the amount of observation and simulation data has grown very quickly in the last decades. This thesis focuses on performing data assimilation in a data-driven way and this without having access to explicit model equations. The main contribution of this thesis lies in developing and evaluating the Analog Data Assimilation(AnDA), which combines analog methods (nearest neighbors search) and stochastic filtering methods (Kalman filters, particle filters, Hidden Markov Models). Through applications to both simplified chaotic models and real ocean remote sensing case-studies (sea surface temperature, along-track sea level anomalies), we demonstrate the relevance of AnDA for missing data interpolation of nonlinear and high dimensional dynamical systems from irregularly-sampled and noisy observations. Driven by the rise of machine learning in the recent years, the last part of this thesis is dedicated to the development of deep learning models for the detection and tracking of ocean eddies from multi-source and/or multi-temporal data (e.g., SST-SSH), the general objective being to outperform expert-based approaches.

L'assimilation de données permet de construire un état initial du modèle (l'analyse) à partir de deux sources d'information : les observations disponibles et une prévision récente (l'ébauche). L'importance relative de chacune des sources dépend du modèle d'erreurs qui leur est attribué. Le modèle le plus commun en prévision numérique du temps (PNT) consiste à formuler l'erreur d'ébauche comme un incrément additif en amplitude et, avec une approche probabiliste, de spécifier sa fonction de densité de probabilité (PDF) par une Gaussienne spécifiée avec une moyenne nulle et une matrice de covariance B. Le premier problème abordé durant cette thèse est le manque de dépendance au flux des modélisations de B. Le deuxième est l'écartement à l'hypothèse Gaussienne notamment en présence d'erreurs de déplacement. La démarche suivie est d'utiliser des déformations spatiales dans un cadre ensembliste pour raffiner la modélisation des corrélations d'erreurs d'ébauche, et de corriger les erreurs de déplacement pour tenter de se rapprocher de l'hypothèse Gaussienne. La première partie du travail de thèse consiste à améliorer la modélisation de B, en prenant en compte objectivement l'anisotropie des corrélations d'erreurs d'ébauche à l'aide de déformations spatiales estimées statistiquement à partir d'un ensemble de prévisions générées avec un ensemble d'assimilation (EDA). Cette méthode statistique (ST) est appliquée à une simulation réelle du modèle global de Météo-France ARPEGE, et son efficacité est comparée à celle de la transformée géostrophique (GT) à l'aide de diagnostics d'anisotropie. Les résultats montrent que la ST est plus efficace que la GT avec une prise en compte plus systématique de l'anisotropie des corrélations spatiales. Dans une deuxième partie, une documentation de la non-Gaussianité (NG) des erreurs d'ébauche d'AROME, le modèle à échelle convective de Météo-France, est proposée. La NG des distributions d'erreurs d'ébauche, diagnostiquées à partir d'un EDA, est quantifiée par un indice uniquement basé sur leur asymétrie et leur aplatissement. La NG diagnostiquée a une forte dépendance spatiale et temporelle, avec des structures spatiales qui suivent les structures météorologiques. Le lien avec certains processus non-linéaires , liés par exemple à la présence de nuages, est notamment observé. Les diagnostics montrent également que le processus d'analyse réduit fortement la NG observée dans l'ébauche, et que la vorticité et la divergence sont les variables de contrôle de l'assimilation les moins Gaussiennes. Une des causes possibles de la NG étant la présence d'erreurs de déplacement, la troisième partie de ce travail de thèse s'attache à documenter les apports de ce modèle d'erreurs alternatif. Un algorithme d'estimation des erreurs de déplacements a tout d'abord été mis en place. Cet algorithme, qui s'inspire des méthodes d'estimation du flot optique, se base sur une approche variationnelle quasi-linéaire, quasi Gaussienne et multi-échelles. Le comportement de la méthode a été testé sur des cas idéalisés, puis sur des cas réels basés sur des prévisions AROME. Ceci permet d'envisager à l'avenir l'estimation statistique des principaux paramètres caractérisants ce type d' erreurs, ainsi que leur exploitation dans des algorithmes visant à corriger ces erreurs en amont de la correction d'amplitude. / Data assimilation provides initial conditions (the analysis) for the model from two sources of information: the available observations and a short term forecast (the background). The relative weights of both sources depend on the chosen error model. Usually, background errors are modeled as additive amplitude errors and, following a probabilistic approach, their probability density function (PDF) are modeled as Gaussians, specified with a zero mean and a covariance matrix B. For this PhD, the tackled issue are the lack of flow dependency of B and the non-Gaussianity (NG) frequently arising in numerical weather prediction (NWP), especially in case of displacement error. Using spatial deformation methods and ensemble of forecasts generated from an ensemble of data (EDA), the aim of this work is to refine the model of background error correlations, and to correct displacement errors as a possible way to get closer to the Gaussian hypothesis. A first part of this work is to improve the B modeling, by accounting objectively for the anisotropy of the background error correlations thanks to spatial deformation, which is statistically estimated from an EDA. This statistical method (ST) is applied to a real dataset of ARPEGE, the global model at Météo-France. The ST is compared to an other spatial warping method, the geostrophic transform (GT), and their efficiency are quantified by an anisotropy index. Results show a better efficiency of the ST than the GT, with a more systematic accounting of the anisotropy of the spatial background correlations. The second part of this work is a documentation of the NG for the background errors of AROME, the convective scale model at Météo-France. Distributions are diagnosed from an EDA, and their Gaussianity is quantified with a statistical index only based on two shape parameters, skewness and kurtosis. The diagnosed NG is spatially and temporally dependent, with large similarities between meteorological and NG structures. Links with nonlinear processes, linked for example to cloud covering, are observed. Diagnostics also show that the analysis step is largely decreasing the observed NG of the background, and that the most non-Gaussian variables in control space are vorticity and divergence. Since, a possible reason of the NG is the displacement errors, the third part of this PhD work aims at getting insight in the documentation of this alternative error model. First, a quasi-linear, quasi-Gaussian and multi-scale method, inspired from optical flow estimation methods, is derived. It has been tested on ideal case and on real cases based on AROME forecast fields. Future works may be, first, a statistical estimation of the main parameters characterizing this kind of error, and second, the use of this algorithm to correct displacement error upstream of the additive error correction.