Nonlinear dynamical systems theory has inspired a new set of useful tools to be applied in climate studies. In this work we presented specific examples where information has been gained by the application of methods from nonlinear dynamical systems theory. The main goal is to understand the relative importance of stochastic forcing versus deterministic coupling within the context of Coupled General Circulation Models. This work address this important subject by approaching this goal through the development of a hierarchy of models with increasing complexity that we assert contain the essential dynamics of ENSO. We examined the effect of noise in a low order model and found that it is not restricted to blurring the attractor trajectories in phase space, but includes important changes in the dynamics of the system. The main results indicate that the presence of noise in a nonlinear system has two different effects. The presence of noise acts to increase the maximum Lyapunov exponent and can result in noise induced chaos if the system was originally stable. However, the same arguments are not valid if the original system is already in the chaotic regime, where the noise inclusion acts to decrease the maximum Lyapunov exponent, therefore increasing the system stability. The system of interest includes coupled ocean-atmosphere interactions and here we mimic this interaction by coupling two low order models with very different dominant time scales. These subsystems interact in a complex, nonlinear way and the behavior of the whole system cannot be explained by a linear summation of dynamics of the system parts. We used information theory concepts to detect the influence of the slow system dynamics in synchronizing the fast system in coupled models. We introduced a fast-slow coupled system, where both the slowness of the ocean model and the intensity of the boundary forcing anomalies contribute to the asymmetry and phase locking of both subsystems. The mechanisms controlling the fast modelspread were uncovered revealing uncertainty dynamics depending on the location of ensemble members in the model’s phase space. As an intermediate step between low order models and CGCMs we study the effect of noise on an intermediate complexity model. The addition of gaussian noise to the Zebiak-Cane model in order to understand the effects of noise on its attractor led to a way of estimating the noise level based on the effects of noise on the correlation dimension curves. We investigate the intrinsic predictability of the coupled models used here, and the different time scales associated with fast and slow modes were detected using the Finite Size Lyapunov Exponents. We found new estimates for the prediction horizon of ENSO for the Zebiak-Cane model as well as for the NCAR CCSM3 model and observations. The whole analysis of observations and CCSM3 was possible after applying noise reduction techniques. We also improved our understanding of three different noise reduction techniques by comparing the Local Projective Noise Reduction, the Interactive Ensemble strategy, and a Random Interactive Ensemble applied to CCSM3. The main difference between these two noise reduction techniques is when the process is applied. The Local Projective Noise Reduction can be applied to both model and observations, and it is done a posteriori in phase space, therefore the trajectories to be adjusted already posses the physical mechanisms embedded in them. The Interactive Ensemble approach can only be applied to model simulations and has shown to be a very useful technique for noise reduction since its done a priori while the system evolves instead of a posteriori, besides the fact that it allows to retrieve the spatial distribution of the noise level in physical space.

A mesoscale non-hydrostatic atmospheric model was extended by including both a chemical transport module (CTM) for the chemical triade NO, N02, and 0 3, and an explicit surface-subgrid module (ESSM) for a subscale resolution of the topographical surface. CTEM: The simulated time-dependent concentration fields result from the following processes involved: anthropogenic emission at different heights, biogenic emission, dry deposition on the receptive surface, chemical reactions, turbulent diffusion, and passive transport according to the model dynamics. The calculations in the lowest model layer, usually treated as a constant-flux layer, are now performed on a vertical subgrid that was inserted to better resolve the often observed high concentration gradients within the surface layer. ESSM: Moreover, an equidistant horizontal-subgrid is introduced for finer resolving the topography. The surface fluxes of momentum, sensible and latent heat, long-wave radiation, soil heat flux and wetness as well as the surf ace-energy balance are calculated in the usual approximations, however, employing the individual surface and soil properties of the subgrid cells. The averaged subgrid quantities serve as boundary values required for the model-grid calculations. Within the CTM the ESSM method leads to an intersection of the horizontal ESSM subgrid and the vertical CTM subgrid. Preliminary results representing an interim realization state of the ESSM demonstrate partially strong changes of the dry deposition rates caused by subgrid-resolved surface properties. / Ein mesoskaliges nicht-hydrostatisches Atmosphärenmodell ist um ein Chemie-TransportModul (CTM) zur Berücksichtigung der Triaden-Komponenten NO, N02 und 03 sowie um ein Verfahren zur verfeinerten Auflösung der topographischen Unterlage (explicit surface-subgrid modul ESSM) erweitert worden. CTM: Die simulierten zeitabhängigen Konzentrationsfelder sind das Resultat folgender modellierter Prozesse: Anthropogene Emission in verschiedenen Höhenschichten, biogene Emission, trockene Deposition (Rezeption), die speziellen chemischen Umwandlungen, turbulente Diffusion und passiver Transport. Da der Schwerpunkt der Prozesse und die höchsten Konzentrationsgradienten innerhalb der bodennahen ersten Modellschicht vorliegen, werden die Berechnungen in dieser Schicht auf einem verfeinerten vertikalen Untergitter durchgeführt. ESSM: Unabhängig von den Eigenheiten des CTM wird für alle untergrundbezogenen meteorologischen Größen ein regelmäßiges horizontales Untergitter zwecks Berücksichtigung des subskalig aufgelösten topographischen Untergrundes eingeführt. Auf diesem Untergitter werden in den bisherigen Näherungen alle Oberflächenflüsse für Impuls, fühlbare und latente Wärme, langwellige Strahlung, der Bodenwärmefluß, die Bodenfeuchte sowie die Energiebilanz am Boden berechnet. Die über die Untergitterzellen gemittelten Werte dienen den weiteren Berechnungen im normalen Modellgitter als die erforderlichen Randwerte. Innerhalb des CTM führt die ESSM-Methode zu einer Überlagerung des vertikalen CTM-Untergitters mit dem horizontalen Untergitter des ESSM. Erste Simulationsergebnisse, die dem derzeitigen Stand in der Realisierung des ESSM entsprechen, erbringen teilweise stark veränderte Depositionsraten infolge der Berücksichtigung der horizontal feiner aufgelösten Topographie.

In dieser Arbeit wird die Hypothese getestet, ob ein Teil der atmosphärischen Komponente interdekadischer Klimavariabilität durch die Anregung interner, atmosphärischer Moden, speziell von singulären Moden eines linearen, stationären Atmosphärenmodell, begründet werden kann. Die Analysen basieren auf einem linearen, baroklinen quasigeostrophischen Modell, wobei der Grundzustand aus Daten einer Langzeitintegration des ECHAMl/LSG abgeleitet wird. Sie beziehen sich auf eine detektierte Oszillationsmode mit einer Periode von 18 Jahren in dieser GCM Integration. Es zeigt sich, daß der führende rechte singuläre Vektor des linearen baroklinen Modells signifikant mit der interdekadischen Anomalie der atmosphärischen Zirkulation über der Nordhemisphäre korreliert. Damit kann ein Anteil von über 40% der räumlichen Varianz dieser interdekadischen Mode erklärt werden. / In the present paper we examine the hypothesis that a part of the atmospheric component of interdecadal variability is manifested in the exitation of internal, atmospheric modes, in particular in singular modes of a linearized, steady-state atmospheric model. This hypothesis is tested by using a baroclinic quasigeostrophic model, for which data from the dimate model ECHAMl/LSG are utilized to define the long-term mean basic state. The analysis refers to the interdecadal oscillation with a period of 18 years in this GCM-integration. A significant projection is found of the first singular vector and the interdecadal mode of atmospheric circulation. This singular vector is able to explain about 40% of the spatial variance of the interdecadal anomaly over the Northern Hemisphere.

Two modules have been developed which qualify mesoscale atmospheric models for simulating the chemical transport at resolutions much higher than the model grid. Compared with total fine-grid application this method proves to be nearly so efficient but more economic. The modules solve the chemical transport equations (a) and submit the horizontal subgrid (b) for the meteorological and chemical calculations: (a) The chemical transport module considers the triad NO-N02-03 together with a simplified hydrocarbon chemistry. Involved are chemical reactions, anthropogenic and biogenic emission, dry deposition, passive transport, and turbulent diffusion. For these calculations a special vertical subgrid was introduced within the lowest atmospheric model layer. lt eliminates the frequently used approach of constant vertical particle fluxes near the surface. (b) The horizontal-subgrid module splits the horizontal model grid equidistantly into subgrid cells. The vertical surface fluxes of momentum, sensible and latent heat, radiation, soil heat and wetness, and chemical components are explicitly treated on this subgrid. The subgrid-averaged surface fluxes are employed for the (coarser) normal-grid calculations of the atmospheric meteorological variables. In contrast to the meteorological quantities, the chemical components and processes are perf ormed at all model layers on the horizontal subgrid. Several results are compared to conventional simulations of variable model resolution. / Zwei Programm-Module für mesoskalige Atmosphärenmodelle sind entwickelt worden, die Chemie-Transport-Vorgänge in höherer als der normalen Modellgitter-Auflösung simulieren. Im Vergleich zu hochaufgelösten Standardmodell-Anwendungen erweist sich diese Methode als effizienter. Die Module lösen die Chemie-Transport-Gleichungen (a) und schaffen das horizontale Untergitter für die meteorologischen und chemischen Berechnungen (b): (a) Im Chemie-Transport-Modul wird die Triade NO-N02-03 gemeinsam mit einer vereinfachten Kohlenwasserstoff-Chemie betrachtet. Berücksichtigt werden chemische Reaktionen, anthropogene und biogene Emissionen, trockene Deposition, passiver Transport und turbulente Diffusion. Für diese Berechnungen wurde innerhalb der untersten Modellschicht ein spezielles vertikales Untergitter eingeführt, um die in Oberflächennähe häufig angewendete Näherung konstanter Stoffflüsse zu eliminieren. (b) Das Untergitter-Modul unterteilt das horizontale Modellgitter in Unterzellen, auf welche die Berechnung der Boden- und Oberflächenflüsse bezogen wird. Die vertikalen Oberflächenflüsse von Impuls, sensibler und latenter Wärme, Strahlung, Bodenwärme und -feuchte sowie der chemischen Komponenten werden explizit im Untergitter bestimmt. Die über die Unterzellen gemittelten Flüsse werden für die im (gröberen) Modellgitter ablaufenden Berechnungen der meteorologischen Größen genutzt. Im Gegensatz dazu werden die chemischen Komponenten und Prozesse in allen Modellschichten vollständig auf dem Untergitter behandelt. Einige Ergebnisse dieser Methode werden im Vergleich mit Standard-Simulationen unterschiedlichen Auflösungsgrades gezeigt.

COMMA-LIM (Cologne Model of the Middle Atmosphere - Leipzig Institute for Meteorology) ist ein 3D-mechanistisches Gitterpuktsmodell, welches sich von ca. 0 bis 135 km in logarhitmischen Druckkordinaten z = -H ln(p=p0) erstreckt, wobei H=7 km und p0 den Referenzdruck am unteren Rand bezeichnet. Die vertikale Auflösung von COMMA-LIM wurde auf 48 Schichten erhöht. Zugleich wurde die Beschreibung des Strahlungsprozesses verbessert, zusammen mit den Beiträgen zur Temperaturbilanz durch atmosphärische Wellen und Turbulenz. Weitere Veränderungen betreffen die numerische Realisation der horizontalen Diffusion und des Filterproblems. Die Beschreibung ist unterteilt in den dynamischen Teil und die Strahlungsbeträge. Die jahreszeitlichen Klimatologien werden vorgestellt und diskutiert. / COMMA-LIM (Cologne Model of the Middle Atmosphere - Leipzig Institute for Meteorology) is a 3D-mechanistic gridpoint model extending up from 0 to 135 km with a logharithmic vertical coordinate z = -H ln(p=p0), where H=7 km and p0 is the reference pressure at lower boundary. The resolution of the 24 layer version has been increased to 48 layers and several improvements are made in the parameterisation of radiative processes, heating/cooling due to atmospheric waves and turbulence, as well as in the numerical realization of the horizontal diffusion and filtering. This description is divided into the section describing the changes in the dynamical part and the modifications in radiation routines. After all, the seasonal climatologies will be shown and discussed to demonstrate what the COMMA-LIM is capable of reproducing.

The  rst stars to illuminate the universe are said to belong to a group called Population III. Having knowledge of their atmospheric conditions would be useful in many endeavours. The aimof this project was to compile and run the FORTRAN code TLUSTY (Hubeny and Lanz 2017) inorder to create stellar atmospheres for Pop III stars. With a working version of TLUSTY, severalcontrol runs were then performed to make sure that everything worked as intended with the  final goal to create a large grid of calculated atmospheres in the parameter space of effective temperatureand surface gravity. Successful comparisons were made against earlier calculations made by Schaerer(2002) and Windhorst et al. (2019). Constructing such a grid required several codes to work togetherwith TLUSTY by constructing a shell script. The result is a grid  lled with many points that werewell converged, as well as a few that did not appear to converge. Comparing the converged part ofthe grid with stellar evaluations tracks made by Yoon et al. (2012) showed that heavy, rotating PopIII stars fell almost within the grid. One problem that arose, however, was that as the grid did notfully converge all the way to the tracks some could not be uniquely mapped to di erent points on thegrid. Thus, some di erent tracks would become degenerate which should not be physically expected. However this may not be a fatal problem.

Previsões meteorológicas para longos períodos de tempo estão se tornando cada vez mais importantes. A preocupação mundial com as consequências da mudança do clima tem estimulado pesquisas para determinar o seu comportamento nas próximas décadas. Ao mesmo tempo, os passos necessários para definir uma melhor modelagem e simulação do clima e/ou tempo estão longe da precisão desejada. Aumentar o refinamento da superfície terrestre e, consequentemente, aumentar o número de pontos discretos (utilizados para a representação da atmosfera) na modelagem climática e precisão das soluções computadas é uma meta que está em conflito com o desempenho das aplicações numéricas. Aplicações que envolvem a interação de longos períodos de tempo e incluem um grande número de operações possuem um tempo de execução inviável para as arquiteturas de computadores tradicionais. Para superar esta situação, um modelo climatológico pode adotar diferentes níveis de refinamento da superfície terrestre, utilizando mais pontos discretos somente em regiões onde uma maior precisão é requerida. Este é o caso de Ocean-Land-AtmosphereModel, que permite o refinamento estático de uma determinada região no início da execução do código. No entanto, um refinamento dinâmico possibilitaria uma melhor compreensão das condições climáticas específicas de qualquer região da superfície terrestre que se tivesse interesse, sem a necessidade de reiniciar a execução da aplicação. Com o surgimento das arquiteturas multi-core e a adoção de GPUs para a computação de propósito geral, existem diferentes níveis de paralelismo. Hoje há paralelismo interno ao processador, entre processadores e entre computadores. Com o objetivo de extrair ao máximo a performance dos computadores atuais, é necessário utilizar todos os níveis de paralelismo disponíveis durante o desenvolvimento de aplicações concorrentes. No entanto, nenhuma interface de programação paralela explora simultaneamente bem os diferentes níveis de paralelismo existentes. Baseado neste contexto, esta tese investiga como explorar diferentes níveis de paralelismo em modelos climatológicos usando interfaces clássicas de programação paralela de forma combinada e como é possível prover refinamento de malhas em tempo de execução para estes modelos. Os resultados obtidos a partir de implementações realizadas mostraram que é possível reduzir o tempo de execução de uma simulação atmosférica utilizando diferentes níveis de paralelismo, através do uso combinado de interfaces de programação paralela. Além disso, foi possível prover maior desempenho na execução de aplicações climatológicas que utilizam refinamento de malhas em tempo de execução. Com isso, uma malha de maior resolução para a representação da atmosfera terrestre pode ser adotada e, consequentemente, as previsões numéricas serão mais precisas. / Weather forecasts for long periods of time has emerged as increasingly important. The global concern with the consequences of climate changes has stimulated researches to determine the climate in coming decades. At the same time the steps needed to better defining the modeling and the simulation of climate/weather is far of the desired accuracy. Upscaling the land surface and consequently to increase the number of points used in climate modeling and the precision of the computed solutions is a goal that conflicts with the performance of numerical applications. Applications that include the interaction of long periods of time and involve a large number of operations become the expectation for results infeasible in traditional computers. To overcome this situation, a climatic model can take different levels of refinement of the Earth’s surface, using more discretized elements only in regions where more precision are required. This is the case of Ocean-Land- Atmosphere Model, which allows the static refinement of a particular region of the Earth in the early execution of the code. However, a dynamic mesh refinement could allow to better understand specific climatic conditions that appear at execution time of any region of the Earth’s surface, without restarting execution. With the introduction of multi-core processors and GPU boards, computers architectures have many parallel layers. Today, there are parallelism inside the processor, among processors and among computers. In order to use the best performance of the computers it is necessary to consider all parallel levels to distribute a concurrent application. However, nothing parallel programming interface abstracts all these different parallel levels. Based in this context, this thesis investigates how to explore different levels of parallelism in climatological models using mixed interfaces of parallel programming and how these models can provide mesh refinement at execution time. The performance results show that is possible to reduce the execution time of atmospheric simulations using different levels of parallelism, through the combined use of parallel programming interfaces. Higher performance for the execution of atmospheric applications that use online mesh refinement was also provided. Therefore, more mesh resolution to describe the Earth’s atmosphere can be adopted, and consequently the numerical forecasts are more accurate.

Previsões meteorológicas para longos períodos de tempo estão se tornando cada vez mais importantes. A preocupação mundial com as consequências da mudança do clima tem estimulado pesquisas para determinar o seu comportamento nas próximas décadas. Ao mesmo tempo, os passos necessários para definir uma melhor modelagem e simulação do clima e/ou tempo estão longe da precisão desejada. Aumentar o refinamento da superfície terrestre e, consequentemente, aumentar o número de pontos discretos (utilizados para a representação da atmosfera) na modelagem climática e precisão das soluções computadas é uma meta que está em conflito com o desempenho das aplicações numéricas. Aplicações que envolvem a interação de longos períodos de tempo e incluem um grande número de operações possuem um tempo de execução inviável para as arquiteturas de computadores tradicionais. Para superar esta situação, um modelo climatológico pode adotar diferentes níveis de refinamento da superfície terrestre, utilizando mais pontos discretos somente em regiões onde uma maior precisão é requerida. Este é o caso de Ocean-Land-AtmosphereModel, que permite o refinamento estático de uma determinada região no início da execução do código. No entanto, um refinamento dinâmico possibilitaria uma melhor compreensão das condições climáticas específicas de qualquer região da superfície terrestre que se tivesse interesse, sem a necessidade de reiniciar a execução da aplicação. Com o surgimento das arquiteturas multi-core e a adoção de GPUs para a computação de propósito geral, existem diferentes níveis de paralelismo. Hoje há paralelismo interno ao processador, entre processadores e entre computadores. Com o objetivo de extrair ao máximo a performance dos computadores atuais, é necessário utilizar todos os níveis de paralelismo disponíveis durante o desenvolvimento de aplicações concorrentes. No entanto, nenhuma interface de programação paralela explora simultaneamente bem os diferentes níveis de paralelismo existentes. Baseado neste contexto, esta tese investiga como explorar diferentes níveis de paralelismo em modelos climatológicos usando interfaces clássicas de programação paralela de forma combinada e como é possível prover refinamento de malhas em tempo de execução para estes modelos. Os resultados obtidos a partir de implementações realizadas mostraram que é possível reduzir o tempo de execução de uma simulação atmosférica utilizando diferentes níveis de paralelismo, através do uso combinado de interfaces de programação paralela. Além disso, foi possível prover maior desempenho na execução de aplicações climatológicas que utilizam refinamento de malhas em tempo de execução. Com isso, uma malha de maior resolução para a representação da atmosfera terrestre pode ser adotada e, consequentemente, as previsões numéricas serão mais precisas. / Weather forecasts for long periods of time has emerged as increasingly important. The global concern with the consequences of climate changes has stimulated researches to determine the climate in coming decades. At the same time the steps needed to better defining the modeling and the simulation of climate/weather is far of the desired accuracy. Upscaling the land surface and consequently to increase the number of points used in climate modeling and the precision of the computed solutions is a goal that conflicts with the performance of numerical applications. Applications that include the interaction of long periods of time and involve a large number of operations become the expectation for results infeasible in traditional computers. To overcome this situation, a climatic model can take different levels of refinement of the Earth’s surface, using more discretized elements only in regions where more precision are required. This is the case of Ocean-Land- Atmosphere Model, which allows the static refinement of a particular region of the Earth in the early execution of the code. However, a dynamic mesh refinement could allow to better understand specific climatic conditions that appear at execution time of any region of the Earth’s surface, without restarting execution. With the introduction of multi-core processors and GPU boards, computers architectures have many parallel layers. Today, there are parallelism inside the processor, among processors and among computers. In order to use the best performance of the computers it is necessary to consider all parallel levels to distribute a concurrent application. However, nothing parallel programming interface abstracts all these different parallel levels. Based in this context, this thesis investigates how to explore different levels of parallelism in climatological models using mixed interfaces of parallel programming and how these models can provide mesh refinement at execution time. The performance results show that is possible to reduce the execution time of atmospheric simulations using different levels of parallelism, through the combined use of parallel programming interfaces. Higher performance for the execution of atmospheric applications that use online mesh refinement was also provided. Therefore, more mesh resolution to describe the Earth’s atmosphere can be adopted, and consequently the numerical forecasts are more accurate.

Previsões meteorológicas para longos períodos de tempo estão se tornando cada vez mais importantes. A preocupação mundial com as consequências da mudança do clima tem estimulado pesquisas para determinar o seu comportamento nas próximas décadas. Ao mesmo tempo, os passos necessários para definir uma melhor modelagem e simulação do clima e/ou tempo estão longe da precisão desejada. Aumentar o refinamento da superfície terrestre e, consequentemente, aumentar o número de pontos discretos (utilizados para a representação da atmosfera) na modelagem climática e precisão das soluções computadas é uma meta que está em conflito com o desempenho das aplicações numéricas. Aplicações que envolvem a interação de longos períodos de tempo e incluem um grande número de operações possuem um tempo de execução inviável para as arquiteturas de computadores tradicionais. Para superar esta situação, um modelo climatológico pode adotar diferentes níveis de refinamento da superfície terrestre, utilizando mais pontos discretos somente em regiões onde uma maior precisão é requerida. Este é o caso de Ocean-Land-AtmosphereModel, que permite o refinamento estático de uma determinada região no início da execução do código. No entanto, um refinamento dinâmico possibilitaria uma melhor compreensão das condições climáticas específicas de qualquer região da superfície terrestre que se tivesse interesse, sem a necessidade de reiniciar a execução da aplicação. Com o surgimento das arquiteturas multi-core e a adoção de GPUs para a computação de propósito geral, existem diferentes níveis de paralelismo. Hoje há paralelismo interno ao processador, entre processadores e entre computadores. Com o objetivo de extrair ao máximo a performance dos computadores atuais, é necessário utilizar todos os níveis de paralelismo disponíveis durante o desenvolvimento de aplicações concorrentes. No entanto, nenhuma interface de programação paralela explora simultaneamente bem os diferentes níveis de paralelismo existentes. Baseado neste contexto, esta tese investiga como explorar diferentes níveis de paralelismo em modelos climatológicos usando interfaces clássicas de programação paralela de forma combinada e como é possível prover refinamento de malhas em tempo de execução para estes modelos. Os resultados obtidos a partir de implementações realizadas mostraram que é possível reduzir o tempo de execução de uma simulação atmosférica utilizando diferentes níveis de paralelismo, através do uso combinado de interfaces de programação paralela. Além disso, foi possível prover maior desempenho na execução de aplicações climatológicas que utilizam refinamento de malhas em tempo de execução. Com isso, uma malha de maior resolução para a representação da atmosfera terrestre pode ser adotada e, consequentemente, as previsões numéricas serão mais precisas. / Weather forecasts for long periods of time has emerged as increasingly important. The global concern with the consequences of climate changes has stimulated researches to determine the climate in coming decades. At the same time the steps needed to better defining the modeling and the simulation of climate/weather is far of the desired accuracy. Upscaling the land surface and consequently to increase the number of points used in climate modeling and the precision of the computed solutions is a goal that conflicts with the performance of numerical applications. Applications that include the interaction of long periods of time and involve a large number of operations become the expectation for results infeasible in traditional computers. To overcome this situation, a climatic model can take different levels of refinement of the Earth’s surface, using more discretized elements only in regions where more precision are required. This is the case of Ocean-Land- Atmosphere Model, which allows the static refinement of a particular region of the Earth in the early execution of the code. However, a dynamic mesh refinement could allow to better understand specific climatic conditions that appear at execution time of any region of the Earth’s surface, without restarting execution. With the introduction of multi-core processors and GPU boards, computers architectures have many parallel layers. Today, there are parallelism inside the processor, among processors and among computers. In order to use the best performance of the computers it is necessary to consider all parallel levels to distribute a concurrent application. However, nothing parallel programming interface abstracts all these different parallel levels. Based in this context, this thesis investigates how to explore different levels of parallelism in climatological models using mixed interfaces of parallel programming and how these models can provide mesh refinement at execution time. The performance results show that is possible to reduce the execution time of atmospheric simulations using different levels of parallelism, through the combined use of parallel programming interfaces. Higher performance for the execution of atmospheric applications that use online mesh refinement was also provided. Therefore, more mesh resolution to describe the Earth’s atmosphere can be adopted, and consequently the numerical forecasts are more accurate.

MENVSC / Department of Geography and Geo-Information Sciences / The process of cloud formation and distribution in the atmospheric circulation system is very important yet not easy to comprehend and forecast. Clouds affect the climate system by controlling the amount of solar radiation, precipitation and other climatic variables. Parameterised induced General Circulation Model (GCMs) are unable to represent clouds and aerosol particles explicitly and their influence on the climate and are thought to be responsible for most of the uncertainty in climate predictions. Therefore, the aim of the study is to investigate the climate of South Africa as simulated by Super Parameterised Community Atmosphere Model (SPCAM) for the period of 1987-2016. Community Atmosphere Model (CAM) and SPCAM datasets used in the study were obtained from Colorado State University (CSU), whilst dynamic and thermodynamic fields were obtained from the NCEP reanalysis ll. The simulations were compared against rainfall and temperature observations obtained from the South African Weather Service (SAWS) database. The accuracy of the model output from CAM and SPCAM was tested in simulating rainfall and temperature at seasonal timescales using the Root Mean Square Error (RMSE). It was found that CAM overestimates rainfall over the interior of the subcontinent during December - February (DJF) season whilst SPCAM showed a high performance in depicting summer rainfall particularly in the central and eastern parts of South Africa. During June – August (JJA), both configurations (CAM and SPCAM) had a dry bias with simulating winter rainfall over the south Western Cape region in cases of little rainfall in the observations. CAM was also found to underestimate temperatures during DJF with SPCAM results closer to the reanalysis. The study further analyzed inter-annual variability of rainfall and temperature for different homogenous regions across the whole of South Africa using both configurations. It was found that SPCAM had a higher skill than CAM in simulating inter-annual variability of rainfall and temperature over the summer rainfall regions of South Africa for the period of 1987 to 2016. SPCAM also showed reasonable skill simulating (mean sea level pressure, geopotential height, omega etc) in contrast to the standard CAM for all seasons at the low and middle levels (850 hPa and 500 hPa). The study also focused on major El Niño Southern Oscillation (ENSO) events and found that SPCAM tended to compare better in general with the observations. Although both versions of the model still feature substantial biases in simulating South African climate variables (rainfall, temperature, etc), the magnitude of the biases are generally smaller in the super parameterized CAM than the default CAM, suggesting that the implementation of the super parameterization in CAM improves the model performance and therefore seasonal climate prediction. / NRF