Les moyennes et hautes latitudes de l'hémisphère nord sont caractérisées par une forte variabilité climatique en hiver, incluant l'occurrence d'évènements extrêmes tels que les vagues de froid ou les tempêtes, et présentent une faible prévisibilité aux échéances mensuelle à saisonnière dans les systèmes opérationnels. Un nombre croissant d'études montre qu'au-delà du couplage océan-atmosphère, le couplage troposphère-stratosphère contribue également à la variabilité climatique à ces échelles de temps. Cette thèse vise à mieux comprendre l'influence de la stratosphère sur la variabilité climatique hivernale à nos latitudes, et à quantifier sa contribution potentielle à la prévisibilité climatique saisonnière en comparaison de la contribution océanique. Dans un premier temps, un état des lieux des connaissances sur le couplage troposphère-stratosphère est dressé et la variabilité inter-annuelle du vortex stratosphérique polaire est revisitée par le biais d'analyses composites sur la base des réanalyses atmosphériques du CEPMMT. Ensuite, les principaux outils de cette thèse sont présentés et validés, à savoir le modèle ARPEGE-Climat et la technique de « nudging » permettant de relaxer (guider) le modèle vers les réanalyses. Comme beaucoup de modèles, les versions 4 et 5 d'ARPEGE-Climat en configuration T63L31 simulent un vortex stratosphérique polaire nettement décalé vers le sud, ce qui peut avoir des conséquences négatives sur la variabilité simulée via la modification des interactions ondes-écoulement moyen. Si la faible résolution verticale dans la stratosphère est souvent mise en avant pour expliquer le manque de prévisibilité dans les modèles, nos travaux sur la version 5 d'ARPEGE-Climat montrent que l'augmentation de la résolution verticale et l'élévation du toit du modèle à 0.1 hPa ne suffisent pas pour obtenir un climat plus réaliste, que ce soit en termes d'état moyen, de variabilité ou de prévisibilité à l'échelle saisonnière. C'est pourquoi, tout au long de cette thèse, la technique de la relaxation de la stratosphère vers les réanalyses issues du CEPMMT a été exploitée afin de montrer, de manière idéalisée, sa forte influence sur la variabilité climatique hivernale aux extratropiques de l'hémisphère nord, par rapport au seul forçage par les températures de surface de la mer observées. L'étude des hivers 1976-1977 et 2009-2010 via la réalisation de simulations d'ensemble avec et sans nudging a permis de confirmer la contribution de la stratosphère à la phase négative de la NAO et aux fortes anomalies négatives de température observées sur l'Europe du nord. La généralisation des ensembles à la période 1958-2007 (avec initialisation au 1er Novembre) confirme l'impact positif du nudging extratropical mais montre un effet très limité du nudging équatorial qu'il conviendrait d'évaluer de manière plus précise en augmentant la taille des ensembles. Ainsi, si elle confirme l'importance de la stratosphère pour la prévision saisonnière hivernale à nos latitudes, cette thèse ouvre de nombreuses perspectives concernant les mécanismes qui sous-tendent le couplage troposphère-stratosphère et l'intérêt d'une prévision statistico-dynamique consistant à relaxer le modèle ARPEGE-Climat vers une stratosphère prévue de manière statistique. / In the Northern Extratropics, winter climate shows a large inter-annual variability compared to other regions and seasons, with the occurrence of extreme weather events such as cold spells, heavy snowfall and wind storms. Unfortunately, current dynamical seasonal forecasting systems still show low predictability in the northern mid-latitudes. Besides ocean-atmosphere coupling, there is growing observational and numerical evidence that troposphere-stratosphere coupling also contributes to climate variability on a wide range of scales. The aim of this thesis is to evaluate this additional forcing by focusing on the stratospheric polar vortex influence on the wintertime climate variability in the northern mid-latitudes at inter-annual and intra-seasonal timescales. We first make a synthesis of the knowledge about troposphere-stratosphere coupling. The inter-annual variability of the stratospheric polar vortex is assessed using composite analysis of atmospheric ECMWF reanalysis. Then, the main tools used during this thesis are described, namely the ARPEGE-Climat model and the nudging of the stratosphere towards the ECMWF reanalysis. Like many other models, ARPEGE-Climat has a polar jet which is too weak and displaced southward compared to reanalysis data, regardless his version, which could have negative consequences on the wintertime variability. Although the poorly resolved stratosphere is often suggested to explain the lack of previsibility in the models, our results show that, in the ARPEGE-Climat V5 model, the improved vertical resolution is not sufficient to simulate a more realistic climate variability and predictability. For these reasons, throughout this thesis, the stratospheric relaxation towards the reanalysis data from the ECMWF is used in order to compare his relative contribution to the tropospheric extra-tropical climate variability, compared to the sea surface temperature forcing. Case studies have been carried out for the 1976-1977 and 2009-2010 winters though ensemble simulations with free or nudged stratosphere. Results confirm the relative contribution of the stratosphere to the negative phase of the NAO and the temperature anomalies over the northern Europe. Similar initialized ensemble simulations over the 1958-2007 period confirm the positive impact of the extra-tropical nudging, but show a limited influence of the equatorial nudging, that needs to be analyzed more precisely with a larger ensemble. In conclusion, this thesis shows the importance of a realistic stratosphere in the extra-tropical seasonal forecast in winter, but a lot of questions remains opened, like mechanisms related to the troposphere-stratosphere coupling, and the interest of a statistico-dynamical forecast including a relaxation towards a statistical stratosphere

This dissertation consists of two essays on predictability of asset prices. "Benchmarking problems and long horizon abnormal returns" and, "Low R-square in the cross section of expected returns". Long run abnormal returns following Initial Public Offerings (IPOs), Seasoned Equity Offers (SEO) and other firm level events are well documented in the finance literature. These findings are difficult to reconcile in an efficient markets world. I examine the seriousness of potential benchmarking errors on the measurement of abnormal returns. I find that the simpler, more parsimonious models perform better in practice and finds that excess performance is not predictable regardless of the APM. Thus, the long run underperformance following SEOs found in the literature is consistent with market efficiency because excess performance itself is not predictable. In the other essay, "Low R-square in the cross section of expected returns", I examine the “low R-square” phenomenon observed in the literature. CAPM predicts exact linear relationship between return and betas (SML). This means that estimated time series betas for firms should be related with firms' future returns. However, the estimated betas have almost no relationship with future returns. The cross-sectional R2 are surprising low (3% average) while time series R2 are higher (around 30 % average). He develops a simple asset pricing model that explains this phenomenon. Even in a perfect world where there are no errors in the benchmark measurement or estimation of the price of market risk the difference in R-squares can be quite large due to the difference in variance between the "market" and average returns. I document that market variance exceeds the variance of average returns, with few exceptions, for the last 74 years.

The primary aim of this thesis was to identify the mechanism under-pinning the word-predictability effect, while a secondary aim was to investigate whether words are processed in serial or parallel. In five experiments, adults’ eye-movements were monitored as they read sentences for comprehension on a computer screen. In Experiments 1 and 2, a critical target-word that was either of high- or low-frequency and either predictable or unpredictable was embedded in experimental sentences. The nature of the preview of the target word was manipulated such that it was either identical to the target or was misspelled (the misspelling was more severe in Experiment 2). Predictability effects were apparent in the identical preview condition in both experiments, whilst they were only apparent in the misspelled condition of Experiment 1. This outcome is compatible with early Guessing Game type models of reading which propose that readers predictions about up-coming words using contextual parafoveal information. When taken together, the results of Experiments 1 and 2 also suggested that frequency and predictability exert additive effects on fixation durations.In Experiment 3, four levels of word-predictability were employed. The function relating word-predictability and word-processing time was strictly monotonic: word-processing time decreased as predictability increased. This outcome was consistent with a word-prediction account of predictability in which there is no penalty for incorrect guessing. Experiment 3 also showed that processing time on the pre-target word increased as the predictability of the up-coming increased. This outcome replicated an effect obtained by Kliegl, Nuthmann and Engbert (2006) who claim that it arises as a result of memory retrieval processes cued by prior sentence context Experiment 4 replicated the manipulation in Experiment 3 but included additional condition in which the preview of the target word was masked while in parafoveal vision, using a pixel scrambling technique. The target-predictability effect was again a graded one, and did not depend upon the availability of initial information, providing evidence against the word-prediction theory. Additionally, there was no pre-target predictability effect in the unmasked condition. There was a pre-target effect in an direction in the masked condition, although this appeared to be a consequence of the mask. Experiment 5 replicated Experiment 4, but replaced the masked condition with a non-predictable but semantically related word, and the results showed no pre-target effects at all. It was concluded that inverted pre-predictability effects are more likely to be related to higher-level sentential processing.

The finding of reversals in weekly returns has been attributed to a combination of microstructure issues and overreaction to information. I provide new evidence eliminating overreaction as a source of reversal. I show that well-known weekly contrarian profits are followed by a long run of momentum profits. In fact, these profits are strong enough to produce a significant momentum effect over the full year following portfolio formation. Thus, the market does not appear to view extreme weekly returns as excessive, as implied by an overreaction story. To the contrary, this return continuation is consistent with underreaction to the news driving extreme weekly returns. This is supported by cross-sectional tests in which I find this week's news is positively related to next week's returns. The evidence presented here is consistent with growing evidence that underreaction to firm-specific information is a pervasive feature of price formation. Therefore, if any short-run contrarian profits can be realized, they are better viewed as compensation for providing liquidity than as a reward for arbitrage.

This dissertation examines stock return predictability with aggregate credit conditions. The aggregate credit conditions are empirically measured by credit standards (Standards) derived from the Federal Reserve Board's Senior Loan Officer Opinion Survey on Bank Lending Practices. Using Standards, this study investigates whether the aggregate credit conditions predict the expected returns and volatility of the stock market. The first essay, "Credit Conditions and Expected Stock Returns," analyzes the predictability of U.S. aggregate stock returns using a measure of credit conditions, Standards. The analysis reveals that Standards is a strong predictor of stock returns at a business cycle frequency, especially in the post-1990 data period. Empirically the essay demonstrates that a tightening of Standards predicts lower future stock returns. Standards performs well both in-sample and out-of-sample and is robust to a host of consistency checks including a small sample analysis. The second essay, "Credit Conditions and Stock Return Volatility," examines the role played by credit conditions in predicting aggregate stock market return volatility. The essay employs a measure of credit conditions, Standards in the stock return volatility prediction. Using the level and the log of realized volatility as the estimator of the stock return volatility, this study finds that Standards is a strong predictor of U.S. stock return volatility. Overall, the forecasting power of Standards is strongest during tightening credit periods.

The finding of reversals in weekly returns has been attributed to a combination of microstructure issues and overreaction to information. I provide new evidence eliminating overreaction as a source of reversal. I show that well-known weekly contrarian profits are followed by a long run of momentum profits. In fact, these profits are strong enough to produce a significant momentum effect over the full year following portfolio formation. Thus, the market does not appear to view extreme weekly returns as excessive, as implied by an overreaction story. To the contrary, this return continuation is consistent with underreaction to the news driving extreme weekly returns. This is supported by cross-sectional tests in which I find this week's news is positively related to next week's returns. The evidence presented here is consistent with growing evidence that underreaction to firm-specific information is a pervasive feature of price formation. Therefore, if any short-run contrarian profits can be realized, they are better viewed as compensation for providing liquidity than as a reward for arbitrage.

Recently, there has been an explosion of studies focusing on animal “personality”, defined as consistent individual differences in behaviour. Researchers within this field commonly partition phenotypic variation in behavioural traits at the among- and within-individual levels, because significant among-individual variation implies that a behaviour is repeatable and thus can be subject to selection. It is well known that individuals can not only differ in their mean-level behavioural expression, but also in how their phenotype varies over time, known as temporal plasticity. However, another important yet largely overlooked aspect of behavioural variation is that some individuals are consistently more “unpredictable” than others. This level of variance corresponds to (unexplained) variation among successive measurements on a given individual. Personality and behavioural plasticity have both been studied in many different contexts, however research regarding the possible correlations between these two traits and unpredictability is lacking. Here, I repeatedly assayed locomotor activity in wild white-footed mice (Peromyscus leucopus) exposed to a 10-min novel environment test, to determine whether their behavioural reaction norms exhibited consistent among-individual differences in intercept (personality), slope (plasticity), and variance (unpredictability). Using a doubly hierarchical generalized linear mixed model, I found that both the intercept and slope of the temporal reaction norms were significantly repeatable (Rint= 0.39, Rslp= 0.23) and negatively correlated (r=-0.51). Moreover, unpredictability was found to be significantly repeatable (Rsd= 0.20), positively correlated with intercept (r=0.92), and negatively correlated with slope (r=-0.51). To the extent that these correlations reflect the underlying quantitative genetic architecture of behaviour, my results suggest the presence of constraints on the evolution of behaviour across multiple levels of variation.

Eddies are one of the key features of the Red Sea circulation. They are not only crucial for energy conversion among dynamics at different scales, but also for materials transport across the basin. This thesis focuses on studying the characteristics of Red Sea eddies, including their temporal and spatial properties, their energy budget, the mechanisms of their evolution, and their predictability. Remote sensing data, in-situ observations, the oceanic general circulation model, and data assimilation techniques were employed in this thesis. The eddies in the Red Sea were first identified using altimeter data by applying an improved winding-angle method, based on which the statistical properties of those eddies were derived. The results suggested that eddies occur more frequently in the central basin of the Red Sea and exhibit a significant seasonal variation. The mechanisms of the eddies’ evolution, particularly the eddy kinetic energy budget, were then investigated based on the outputs of a long-term eddy resolving numerical model configured for the Red Sea with realistic forcing. Examination of the energy budget revealed that the eddies acquire the vast majority of kinetic energy through conversion of eddy available potential energy via baroclinic instability, which is intensified during winter. The possible factors modulating the behavior of the several observed eddies in the Red Sea were then revealed by conducting a sensitivity analysis using the adjoint model. These eddies were found to exhibit different sensitivities to external forcings, suggesting different mechanisms for their evolution. This is the first known adjoint sensitivity study on specific eddy events in the Red Sea and was hitherto not previously appreciated. The last chapter examines the predictability of Red Sea eddies using an ensemble-based forecasting and assimilation system. The forecast sea surface height was used to evaluate the overall performance of the short-term eddy predictability. Different ensemble sampling schemes were implemented, and the investigation among different schemes is followed by a discussion of performance and challenges based on the results of a case study. The thesis not only enhances understanding of the Red Sea dynamics, but also deepens knowledge of the physical-biological and air-sea interactions within the basin. Further, it is a stepping stone to building a robust regional operational system with refined forecasting skills.

京都大学 / 0048 / 新制・論文博士 / 博士(理学) / 乙第12830号 / 論理博第1541号 / 新制||理||1578(附属図書館) / 31368 / (主査)教授 余田 成男, 教授 石川 裕彦, 教授 向川 均 / 学位規則第4条第2項該当 / Doctor of Science / Kyoto University / DGAM

京都大学 / 新制・課程博士 / 博士(理学) / 甲第24124号 / 理博第4852号 / 新制||理||1694(附属図書館) / 京都大学大学院理学研究科地球惑星科学専攻 / (主査)教授 竹見 哲也, 教授 榎本 剛, 准教授 重 尚一 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DFAM