The first part of this thesis is an assessment of the ability of global climate models to reproduce observed features of the leading Empirical Orthogonal Function (EOF) mode of North Pacific sea surface temperature (SST) anomalies known as the Pacific Decadal Oscillation (PDO). The simulations from 13 global climate models I am analyzing were performed under phase 3 of the coupled model intercomparison project (CMIP3). In particular, I am investigating whether these climate models capture tropical influences on the PDO, and the influences of the PDO on North American surface temperature and precipitation. My results are that 1) the models as group produce a realistic pattern of the PDO. The simulated variance of the PDO index is overestimated by roughly 30%. 2) The tropical influence on North Pacific SSTs is biased systematically in these models. The simulated response to El Niño-Southern Oscillation (ENSO) forcing is delayed compared to the observed response. This tendency is consistent with model biases toward deeper oceanic mixed layers in winter and spring and weaker air-sea feedbacks in the winter half-year. Model biases in mixed layer depths and air-sea feedbacks are also associated with a model mean ENSO-related signal in the North Pacific whose amplitude is overestimated by roughly 30%. Finally, model power spectra of the PDO signal and its ENSO-forced component are “redder” than observed due to errors originating in the tropics and extratropics. 3) The models are quite successful at capturing the influence of both the tropical Pacific related and the extratropical part of the PDO on North American surface temperature. 4) The models capture some of the influence of the PDO on North American precipitation mainly due to its tropical Pacific related part. In the second part of this thesis, I investigate the ability of one such coupled ocean- atmosphere climate model, carefully initialized with observations, to dynamically predict the future evolution of the PDO on seasonal to decadal time scales. I am using forecasts produced by the Canadian climate data assimilation and prediction system employing the Canadian climate model CanCM3 for seasonal (CHFP2) and CanCM4 for decadal (DHFP1) predictions. The skill of this system in predicting the future evolution of the PDO index is then inferred from a set of historical “forecasts” called hindcasts. In this manner, hindcasts are issued over the past 30 years (seasonal), or over the past 50 years (decadal) when they can be verified against the observed historical evolution of the PDO index. I find that 1) CHFP2 is successful at predicting the PDO at the seasonal time scale measured by mean-square skill score and correlation skill. Weather “noise” unpredictable at the seasonal time scale generated by substantial North Pacific storm track activity that coincides with a shallow oceanic mixed layer in May and June appear to pose a prediction barrier for the PDO. PDO skill therefore depends on the start season of the forecast. PDO skill also varies as a function of the target month. Variations in North Pacific storminess appear to impact PDO skill by means of a lagged response of the ocean mixed layer to weather “noise”. In CHFP2, times of increasing North Pacific storm track activity are followed by times of reduced PDO skill, while the North Pacific midwinter suppression of storm track activity with decreasing storminess is followed by a substantial recovery in PDO skill. 2) This system is capable of forecasting the leading 14 EOF modes of North Pacific SST departures, that explain roughly three quarters of the total SST variance. CHFP2 is less successful at predicting North Pacific SSTs, i.e., the combination of all the EOF modes, at the seasonal time scale. 3) Besides the skill in Pacific SST, CHFP2 skillfully predicts indices that measure the atmospheric circulation regime over the North Pacific and North America such as the Pacific/North American pattern (PNA) (skillful for three out of four start seasons) and the North Pacific index (NPI) (skillful for all four start seasons). 4) CHFP2 is successful at forecasting part of the influence of Pacific SST on North American climate at the seasonal time scale. Measured by 12-month average anomaly correlation skill, in this system the PDO is a better predictor for North American precipitation (skillful for all four start seasons) than temperature (skillful for one out of four start seasons). In CHFP2, ENSO is a better predictor for North American temperature (skillful for all four start seasons) than the PDO. Both ENSO and the PDO are, however, good predictors for North American precipitation (skillful for all four start seasons). Finally, DHFP1 is less successful at forecasting the PDO at the decadal time scale. Ten-year forecasts of the PDO index exhibit significantly positive correlation skill exclusively in the first year of the forecast. When the correlation skill of the predicted index averaged over lead years is considered, the PDO skill in this system stays significantly positive during the first three years of the decadal forecast. In other words, this climate data assimilation and prediction system is expected to skillfully predict the future three year averaged evolution of the PDO index, but not the evolution of the index in each year individually. / Graduate

There is much confidence in the global temperature change and its attribution to human activities. Global climate models have attained unprecedented complexity in representing the climate system and its response to external forcings. However, climate prediction remains a serious challenge and carries large uncertainty, particularly when the scale of interest becomes small. With the increasing interest in regional impact studies for decision-making, one of the urgent tasks is to make a systematic, quantitative evaluation of the expected skill of climate models over a range of spatiotemporal scales. The first part of this dissertation was devoted to this task, with focus on the predictive skill in the linear trend of surface air temperature. By evaluating the hindcasts for the last 120 year period in the form of deterministic and probabilistic predictions, it was found that the hindcasts can reproduce broad-scale changes in the surface air temperature, showing reliable skill at spatial scales larger than or equal to a few thousand kilometers (30° x 30°) and at temporal scales of 30 years or longer. However, their skill remains limited at smaller spatiotemporal scales, where we saw no significant improvement over climatology or a random guess. Over longer temporal scales, the feedbacks from the carbon cycle to atmospheric CO₂ concentration become important. Therefore the rest of the dissertation attempts to find key processes in the climate-carbon cycle feedback using one of the leading land-climate models, the National Center for Atmospheric Research Community Land Model. Evaluation of site-level simulations using field observations from the Amazon forest revealed that the current formulation for drought-related mortality, which lacks the ecophysiological link between short- and long-term drought stress, prevent the model from simulating realistic forest response. Global simulations showed that such dynamics of vegetation strongly influences the control of the nitrogen cycle on vegetation productivity, which then alters the sensitivity of the terrestrial biosphere to surface air temperature. This implies that if the state of the terrestrial biosphere is inconsistent with the simulated climate, either biased or prescribed, then its feedback to anthropogenic forcing could be also inconsistent.

Evidence for a human contribution to climatic changes during the past century is accumulating rapidly. Given the strength of the evidence, it seems natural to ask whether forcing projections can be used to forecast climate change. A Bayesian method for post-processing forced climate model simulations that produces probabilistic hindcasts of inter-decadal temperature changes on large spatial scales is proposed. Hindcasts produced for the last two decades of the 20th century are shown to be skillful. The suggestion that skillful decadal forecasts can be produced on large regional scales by exploiting the response to anthropogenic forcing provides additional evidence that anthropogenic change in the composition of the atmosphere has influenced our climate. In the absence of large negative volcanic forcing on the climate system (which cannot presently be forecast), the global mean temperature for the decade 2000-2009 is predicted to lie above the 1970-1999 normal with probability 0.94. The global mean temperature anomaly for this decade relative to 1970-1999 is predicted to be 0.35C (5-95% confidence range: 0.21C-0.48C). Reconstruction of temperature variability of the past centuries using climate proxy data can also provide important information on the role of anthropogenic forcing in the observed 20th century warming. A state-space model approach that allows incorporation of additional non-temperature information, such as the estimated response to external forcing, to reconstruct historical temperature is proposed. An advantage of this approach is that it permits simultaneous reconstruction and detection analysis as well as future projection. A difficulty in using this approach is that estimation of several unknown state-space model parameters is required. To take advantage of the data structure in the reconstruction problem, the existing parameter estimation approach is modified, resulting in two new estimation approaches. The competing estimation approaches are compared based on theoretical grounds and through simulation studies. The two new estimation approaches generally perform better than the existing approach. A number of studies have attempted to reconstruct hemispheric mean temperature for the past millennium from proxy climate indicators. Different statistical methods are used in these studies and it therefore seems natural to ask which method is more reliable. An empirical comparison between the different reconstruction methods is considered using both climate model data and real-world paleoclimate proxy data. The proposed state-space model approach and the RegEM method generally perform better than their competitors when reconstructing interannual variations in Northern Hemispheric mean surface air temperature. On the other hand, a variety of methods are seen to perform well when reconstructing decadal temperature variability. The similarity in performance provides evidence that the difference between many real-world reconstructions is more likely to be due to the choice of the proxy series, or the use of difference target seasons or latitudes, than to the choice of statistical method.

Potato (Solanum tuberosum L.), Solanaceae, is ranked fourth in food amount production, exceeded only by wheat, rice and corn. Brazil has potential in climate and soil for growing potatoes, but with a growing concern of society with possible changes in global and regional climate because of anthropogenic increase in greenhouse gases, this crop may be affected in the future. This study had the following objectives: to calibrate for Santa Maria and evaluate simulation models of potato tuber yield for Santa Maria (subtropical climate and São Joaquim (temperate climate), and evaluate the potato tuber yield in scenarios with increasing concentration of carbon dioxide and temperature in Santa Maria, RS, in different sowing dates, considering symmetric and asymmetric increase in minimum and maximum daily air temperature. We evaluated six simulation models of potato tuber yield and the statistics used to evaluate the performance of the models was the root mean square error (RQME). In order to evaluate the effect of climate change on potato tuber yield, one of the six evaluated models (SPITTERS, 1987) was used. Seven climate scenarios (0ºC, 1ºC, 2ºC, 3ºC, 4ºC, 5ºC to 6ºC) for the next 100 years were generated with the Weather Generator LARS-WG using as base series observed meteorological data from 1969 to 2003 in Santa Maria, RS. Potato tuber yield was simulated throughout the 100-year period of each climate scenario in several planting dates for the Spring and Fall growing seasons. Models with the original coefficients showed similar performance in the simulation of tuber yield in Santa Maria and São Joaquin. The models of Hartz; Moore, Johnson, and Pereira simulated better the potato tuber yield. After calibration, the Sands. model was the one that best predicted the potato tuber yield, followed by the models of Pereira and MacKerron; Waister. For the Spring growing season, a symmetrical increase in temperature of 4ºC and an asymmetrical increase of 5ºC offset the beneficial effect of increasing concentration of carbon dioxide. For the Fall growing season, the increase in temperature affected little the potato tuber yield. An anticipation of 14 days in the emergence date in the Spring growing season and a delay of seven days in emergence date in the Fall growing season decrease the negative impact of increased air temperature on the tuber yield of potato grown in Santa Maria, RS. / A batata (Solanum tuberosum L.), família Solanaceae, ocupa o quarto lugar em quantidade de produção de alimentos, sendo superada apenas pelo trigo, arroz e milho. O Brasil possui elevado potencial edafoclimático para o cultivo da batata, mas devido às possíveis mudanças no clima global e regional esta cultura poderá ser afetada no futuro. Esta tese teve como objetivos: calibrar para Santa Maria-RS e avaliar modelos de simulação da produtividade de tubérculos de batata em Santa Maria-RS (clima subtropical) e São Joaquim-SC (clima temperado) e avaliar a produtividade de tubérculos de batata em cenários com aumento da concentração de dióxido de carbono e da temperatura em Santa Maria, RS, em diferentes datas de plantio, considerando aumento simétrico e assimétrico na temperatura mínima e máxima diária do ar. Foram avaliados seis modelos de simulação da produtividade de tubérculos de batata e a estatística utilizada para avaliar o desempenho dos modelos foi a da raiz do quadrado médio do erro (RQME). Para verificar o efeito da mudança climática na produtividade de tubérculos de batata foi utilizado um dos modelos avaliados (SPITTERS, 1987). Sete cenários climáticos (0ºC, 1ºC, 2ºC, 3ºC, 4ºC, 5ºC e 6ºC) para os próximos 100 anos foram gerados com o aplicativo computacional Weather Generator LARS-WG usando-se como base a série de dados meteorológicos observados de 1969 a 2003 em Santa Maria, RS. A produtividade de tubérculos de batata foi simulada ao longo dos 100 anos de cada cenário climático em várias datas de plantio no cultivo de primavera e no cultivo de outono. Os modelos com os coeficientes originais apresentaram desempenho semelhante na simulação da produtividade de tubérculos de batata em Santa Maria e São Joaquim. Os modelos que melhor simulam a produtividade de tubérculos de batata são os modelos de Hartz; Moore, Johnson e Pereira. Após a calibração dos modelos por ajuste dos coeficientes, o modelo de Sands é o que melhor prediz a produtividade de tubérculos de batata, seguido pelos modelos de Pereira e MacKerron; Waister. Para os cultivos de primavera, um aumento simétrico na temperatura do ar a partir de 4ºC e assimétrico a partir de 5ºC (temperatura mínima 6ºC e temperatura máxima 4ºC) anulou o efeito benéfico do aumento da concentração de dióxido de carbono. Para os cultivos de outono, o aumento da temperatura do ar praticamente não afeta a produtividade de tubérculos de batata. A antecipação de 14 dias na data de emergência no cultivo de primavera e o atraso de sete dias na data de emergência no cultivo de outono diminui o impacto negativo do aumento da temperatura do ar na produtividade de tubérculos de batata em Santa Maria, RS.

Les épisodes de sécheresse et de canicule qui frappent épisodiquement les régions tempérées ont des conséquences préjudiciables sur les plans sanitaire, économique, social et écologique. Afin de pouvoir enclencher des stratégies de préparation et de prévention avec quelques semaines ou mois d'anticipation, les attentes sociétales en matière de prévision sont élevées, et ce d'autant plus que les projections climatiques font craindre la multiplication de ces épisodes au cours du 21ème siècle. Néanmoins, la saison d'été est la plus difficile à prévoir aux moyennes latitudes. Les sources connues de prévisibilité sont plus ténues qu'en hiver et les systèmes de prévision climatique actuels peinent à représenter correctement les mécanismes de téléconnexion associés. Un nombre croissant d'études a mis en évidence un lien statistique dans certaines régions entre l'humidité du sol au printemps et les températures et précipitations de l'été qui suit. Ce lien a été partiellement confirmé dans des modèles numériques de climat mais de nombreuses interrogations subsistent. L'objectif de cette thèse est donc de mieux comprendre le rôle joué par l'humidité du sol sur les caractéristiques et la prévisibilité du climat de l'été dans les régions tempérées. Grâce notamment au modèle couplé de circulation générale CNRM-CM, nous avons mis en œuvre des ensembles de simulations numériques qui nous ont permis d'évaluer le degré de persistance des anomalies d'humidité du sol printanière. En effet, une longue persistance est une condition nécessaire pour que ces anomalies influencent le climat à l'échelle de la saison, via le processus d'évapotranspiration de la surface. En imposant dans notre modèle des conditions initiales et aux limitées idéalisées d'humidité du sol, nous avons mis en évidence des régions du globe pour lesquelles l'état moyen et la variabilité des températures et des précipitations en été sont particulièrement sensibles à ces conditions. C'est notamment le cas sur une grande partie de l'Europe et de l'Amérique du nord, y compris à des latitudes élevées. Pour toutes ces régions, l'humidité du sol est une source prometteuse de prévisibilité potentielle du climat à l'horizon saisonnier, bien que de fortes incertitudes demeurent localement sur le degré de persistance de ses anomalies. Une expérience de prévisibilité effective coordonnée avec plusieurs systèmes de prévision montre qu'une initialisation réaliste de l'humidité du sol améliore la prévision de températures estivales principalement dans le sud-est de l'Europe. Dans d'autres régions, comme l'Europe du Nord, le désaccord des modèles provient de l'incertitude sur la persistance des anomalies d'humidité du sol. En revanche, sur les Grandes Plaines américaines, aucun modèle n'améliore ses prévisions qui restent donc très médiocres. La littérature ainsi que nos évaluations de sensibilité du climat à l'humidité du sol ont pourtant identifié cette région comme un "hotspot" du couplage entre l'humidité du sol et l'atmosphère. Nous supposons que l'échec de ces prévisions est une conséquence des forts biais chauds et secs présents dans tous les modèles sur cette région en été, qui conduisent à un dessèchement excessif des sols. Pour le vérifier, nous avons développé une méthode qui corrige ces biais au cours de l'intégration des prévisions avec CNRM-CM6. Les prévisions qui en résultent sont nettement améliorées sur les Grandes Plaines. La compréhension de l'origine des biais continentaux en été et leur réduction dans les prochaines générations de modèles de climat sont des étapes essentielles pour tirer le meilleur parti de l'humidité du sol comme source de prévisibilité saisonnière dans les régions tempérées. / Severe heat waves and droughts that episodically hit temperate regions have detrimental consequences on health, economy and society. The design and deployment of efficient preparedness strategies foster high expectations for the prediction of such events a few weeks or months ahead. Their likely increased frequency throughout the 21st century, as envisaged by climate projections, further emphasizes these expectations. Nevertheless, the summer season is the most difficult to predict over mid-latitudes. Well-known sources of predictability are weaker than in winter and current climate prediction systems struggle to adequately represent associated teleconnection mechanisms. An increasing number of studies have shown a statistical link over some regions between spring soil moisture and subsequent summer temperature and precipitation. This link has been partly confirmed in climate numerical models, but many questions remain. The purpose of this PhD thesis is to better understand the role played by soil moisture onthe characteristics and predictability of the summer climate in temperate regions. By means of the CNRM-CM coupled general circulation model, we have designed a range of numerical simulations which help us evaluate the persistence level of spring soil moisture anomalies. Indeed, a long persistence is a necessary condition for these anomalies to influence the climate at the seasonal scale, through the process of evapotranspiration. By imposing in our model idealized initial and boundary soil moisture conditions, we have highlighted areas of the globe for which the average state and the variability of temperatures and precipitation in summer is particularly sensitive to these conditions. This is the case in particular for Europe and North America, including over high latitudes. Soil moisture is therefore a promising source of potential seasonal climate predictability for these regions, although the persistence of soil moisture anomalies remains locally very uncertain. An effective predictability coordinated experiment, bringing together several prediction systems, shows that a realistic soil moisture initialization improves the forecast skill of summer temperatures mainly over southeast Europe. In other regions, such as Northern Europe, the disagreement between models comes from uncertainty about the persistence of soil moisture anomalies. On the other hand, over the American Great Plains, even the forecasts with improved soil moisture initialization remain unsuccessful. Yet, the literature as well as our assessment of climate sensitivity to soil moisture have identified this region as a "hotspot" of soil moisture - atmosphere coupling. We assume that the failure of these predictions relates to the strong hot and dry bias present in all models over this region in summer, which leads to excessive soil drying. To verify this assumption, we developed a method that corrects these biases during the forecast integration based on the CNRM-CM6 model. The resulting forecasts are significantly improved over the Great Plains. Understanding the origin of continental biases in the summer and reducing them in future generations of climate models are essential steps to making the most of soil moisture as a source of seasonal predictability in temperate regions