Submitted by Emanuel Varela Cardoso (emanuel.varela@ufcg.edu.br) on 2018-08-15T19:02:39Z No. of bitstreams: 1 RAFAEL CASTELO GUEDES MARTINS – TESE (PPGMet) 2014.pdf: 3362803 bytes, checksum: 5a99c28e73f6a95fef76f82f96d2edc4 (MD5) / Made available in DSpace on 2018-08-15T19:02:39Z (GMT). No. of bitstreams: 1 RAFAEL CASTELO GUEDES MARTINS – TESE (PPGMet) 2014.pdf: 3362803 bytes, checksum: 5a99c28e73f6a95fef76f82f96d2edc4 (MD5) Previous issue date: 2014-12-12 / Um dos principais desafios atuais da modelagem numérica da atmosfera trata da previsão quantitativa da precipitação e do posicionamento das nuvens de chuva. Este trabalho tem com o principal objetivo avaliar o desempenho das arametrizações de microfísicas na modelagem regional com ênfase no papel da informação de grande escala e sua influência sobre as simulações, e no uso de dados observados de radiossondagens como forma de acrescentar informação à modelagem . Inicialmente, duas reanálises (NCEP2 e ERAI) foram estatisticamente comparadas com dados de PCDs do Estado do Ceará. Verificou - se qu e a ERAI apresentou maior semelhança com as observações, principalmente para as variáveis diretamente ligadas à convecção. Em seguida, a ERAI foi utilizada como forçamento de grande escala em simulações com o modelo WRF. Observou- se que o uso de microfísica detalhada não melhora necessariamente a previsão do modelo, caso não sejam utilizados dados observados no local de estudo. Por último, duas simulações de alta resolução foram realizadas. Uma forçada pela reanálise sem modificação e outra forçada pela reanálise modificada utilizando o método de análise objetiva do WRF, para incluir as séries temporais de radiossondagens coletadas durante campanha experimental do Projeto CHUVA, em Fortaleza- CE. As duas simulações foram comparadas com dados observados pelo radiômetro para o mesmo local e período das radiossondagens . Observou - se que a inclusão das observações de sondagens na modelagem possibilita melhor modelagem de um sistema convectivo ocorrido em abril de 2011, principalmente para as variáveis ligadas à convecção. Este trabalho aponta, utilizando análises comparativas e estatísticas, que a utilização de uma maior densidade de dados observacionais válidos no modelo pode melhorar de forma muito mais eficiente o resultado da modelagem, do que mesmo a utilização do downscaling dinâmico do dado de grande escala ou a utilização de esquemas de microfísica detalhada, que, em algumas situações, pode inclusive inserir mais erros nos sistema s modelados. / The quantitative prediction of precipitation and the positioning of the rain clouds is one of the main challenges of numerical modeling of the atmosphere in present days. This work aims to evaluate the performance of the microphysical parameterizations in regional modeling, with emphasis on the role of large- scale information and its influence on the simulations, and the use of observational data from radiosondes as a way to add information to modeling. Initially, two reanalysis (NCEP2 and ERAI) were statistically compared with data from PCDs from the Ceará State. It was found that the ERAI showed similarity to the observations, especially for variables directly linked to convection. Then, the ERAI is used as large scale forcing in simulations with the WRF model. It was observed that the use of detailed microphysics does not necessarily improve the model performance, if in situ data were not used. Finally, two high resolution simulations were performed. The first f orced by reanalysis without modification and other forced by reanalysis using the modified method of objective analysis of the WRF, to include the time series of radiosonde observations collected during the experimental campaign of the CHUVA Project in Fortaleza- CE. The two simulations were compared with data observed by the radiometer to the same place and period of the radiosonde. It was observed that the inclusion of radiosonde observations in to the model leads to a better simulation of a convective system that occurred in April 2011, mostly for the variables related to convection. Using comparative statistical analysis, t his work points that the use of a higher density of valid observational data in the model can improve much more efficiently the model results than the use of a dynamic downscal ing of large- scale data or the use of schemes with detailed microphysics, which in some circumstances may even introduce more errors into the modeled system s.

The roughness of ice crystals, defined by small-scale surface roughness and large scale complexity, in high-altitude cloud, has been studied due to its important influence on the radiative properties of ice cloud. The Small Ice Detector 3 (SID-3) created at the University of Hertfordshire was used to measure the characteristics of individual ice crystals in situ. These are supplemented by a range of meteorological in situ measurements, including temperature, relative humidity, and wind velocity to investigate the influence of atmospheric conditions on ice crystal roughness/complexity. Since the method of roughness retrieval was novel, for atmospheric ice particles, laboratory experiments were setup to test and improve the characterization techniques. Criteria were set as a result of the laboratory experiments which data was expected to meet for it to be deemed reliable. These criteria and techniques were applied to data collected in situ on research aircraft. A range of degrees of ice crystal roughness were observed over five flights from two campaigns based out of Scotland in 2012 and 2015 (PIKNMIX and CIRCCREX). When all the flights were combined the majority of particles (51%) were categorised as lightly rough; the second most common roughness type was moderately rough (39%). Smooth particles made up 10% of the total particles, and < 0.02% were classed as severely rough. When considering a wave-cloud case separately, a similar range of roughness values were seen, however, smooth particles were only observed at the cloud leading-edge where nucleation was expected to occur during the only straight level run of the aircraft to probe this region. During the same wave-cloud flight smooth particles were more common in supersaturated regions and moderately rough crystals were more common in subsaturated regions, suggesting that crystals are more likely to tend towards rougher values when observed in subsaturated environments (a statistical T-test showed this hypothesis to be statistically significant). It was found that due to limitations associated with instantaneous measurements, it was challenging to observe how ice particle roughness evolved in situ, since the history of the individual crystals was unknown in most cases. Orographic cloud, however, was found to provide a more robust estimation of crystal evolution as a consequence of having sharp-leading edges where nucleation events were expected to occur, and since crystals then follow streamlines, the distance from the sharp-leading edge can act as a proxy for time since nucleation.

Le sondeur hyperspectral infrarouge IASI, dont le premier modèle vole depuis 2006 sur le satellite défilant météorologique Metop-A, a déjà conduit a des retombées scientifiques très spectaculaires, en prévision météorologie et pour l’étude de la composition atmosphérique et du climat. Les mesures du sondeurs sont toutefois largement sous exploitées en grande partie du a la présence des nuages dans l’atmosphère. Ces derniers interagissent avec le rayonnement incident de façon hautement non-linéaire rendant le traitement de la mesure du sondeur bien plus complexe, voire parfois rédhibitoire pour accéder depuis l’espace aux propriétés des couches atmosphériques situées au-dessus du nuage, mais également en dessous dans le cas de semi-transparence. Cependant, au vue de la quantité d’informations potentielles qu’offre les sondeurs, la communauté scientifique s’intéresse de près a l’exploitation des radiances nuageuse, c’est dans ce cadre que s’inscrivent les travaux de recherche de cette thèse. Nous proposons d’étudier deux schémas nuageux radicalement différents : la clarification nuageuse et un schéma permettant de simuler la radiance nuageuse en utilisant les propriétés optique et microphysique des nuages. La première de ces méthodes, initiée par Smith et al. (1968), permet sous certaines conditions, de faire abstraction du nuage dans le pixel IASI. La méthode est basée sur l’algorithme du logiciel Scenes Heterogenes du CNES. Apres une première étape de validation, les performances de la méthode sont évaluées a travers la quantité d’information indépendante qu’offre la clarification par rapport a une chaine de traitement des radiances nuageuses mise en place au CMS. Les résultats sont favorables à la méthode testée permettant de traiter les couches atmosphériques situées sous le nuage, possédant donc une quantité plus importante. Cependant la clarification repose a la fois sur une hypothèse forte d’homogénéité atmosphérique et ne s’applique qu’à 15% des situations nuageuses. La seconde méthode est une simulation de la radiance nuageuse par des modèles de transfert radiatif rapides utilisant les propriétés optique et microphysique du nuage. Cette méthode présente l’avantage majeur d’utiliser les mêmes profils nuageux que ceux produits par les modèles de prévision numérique, laissant entrevoir l’assimilation de ces profils à partir de la mesure IASI. Cependant, l’utilisation de ces modèles de transfert radiatif rapide dans le cadre d’une assimilation de données n’en est encore qu’à ces prémices, très peu d’études ont été menées sur ce sujet. Nous proposons une étude en trois étapes permettant une utilisation en opérationnel de ces modèles de transfert radiatif. La première étape est une compréhension des modèles et de leur validité en réalisant quelques études de cas s’appuyant sur la campagne de mesures de Lindenberg. Ensuite, dans le cadre de la campagne ConcordIasi, une statistique est réalisée mettant en place des filtrage pour sélectionner uniquement les profils nuageux cohérent avec l’observation IASI. La dernière étape est une application en global, les statistiques révèlent une nette amélioration des écarts a l’ébauche grâce aux filtres, passant de 8K a 2K. Nous proposons tout au long de l’étude une discussion sur les modèles utilises (RTTOV et HISCRTM), leurs points forts et leurs défaillances. Enfin l’ultime étape, permet d’évaluer les performances des profils nuageux issus des modèles de prévision numérique. / The IASI hyperspectral infrared sounding interferometer, the first model of which has been flown on board the meteorological polar orbiting satellite MetOp-A since 2006, has already led to spectacular scientific breakthroughs in both weather forecasting and research into atmospheric composition and the climate. Measurements from the sounders are however largely underutilised, mainly because of the presence of clouds in the atmosphere. The highly non-linear way in which the clouds interact with incident radiation makes analysis of the readings much more complex, and can sometimes even prohibit access from space to the properties of not only the atmospheric layers located above the cloud, but also below them in the case of semi-transparency. However, in view of the potential amount of information offered by the sounders, the scientific community is very interested in exploiting cloud radiance. The research for this thesis stems from this interest. We plan to study two radically different cloud schemes: cloud clarification and a scheme which allows for simulation of cloud radiance by using the optical and microphysical properties of clouds. The first of these methods, initiated by Smith et al. (1968), allows us under certain conditions to disregard the cloud in the IASI pixel. This method is based on the CNES Heterogeneous Scenes software algorithm. After a first validation step, method performance is evaluated by the amount of independent information offered by the clarification, compared to a cloud radiance process chain established at CMS. The results are favorable to the tested method allowing us to deal with atmospheric layers under the cloud, which have therefore larger quantities. However clarification is based on a strong assumption of atmospheric homogeneity and only applies to 15% of cloud situations. The second method is a simulation of cloud radiance by fast radiative transfer models using the optical and microphysical properties of the cloud. The major advantage of this method is that it uses the same cloud profiles as those produced by numerical weather prediction models, allowing assimilation of these profiles from the IASI measurement. However, the use of these fast radiative transfer models in the context of data assimilation is still in the early stages, very few studies have been conducted on this topic. We are proposing a three-phase study which will allow for an operational use of these radiative transfer models. The first step is validation. This is done by conducting several case studies based on Lindenberg’s measurement campaign. Then, within the framework of the ConcordIasi campaign, a statistical analysis will be carried out by introducing filtering, to select cloud profiles which are consistent with the IASI observations. The last step is an overall application, the statistics showing a clear improvement in deviation from the draft thanks to the filters, going from 8K to 2K. Throughout the study we will discuss the models used (RTTOV and HISCRTM), their strengths and weaknesses. Finally the last step allows us to evaluate the performance of the cloud profiles obtained by the digital forecasting models

Nuvens desempenham um papel fundamental no balanço radiativo terrestre, e o conhecimento de suas propriedades micro e macrofísicas é importante para o estudo do clima global. O desenvolvimento convectivo de nuvens está intimamente relacionado ao comportamento microfísico de seus hidrometeoros, os quais são influenciados pela variação nas concentrações de aerossóis disponíveis na atmosfera. Este trabalho utiliza o sensoriamento remoto por satélites para analisar a evolução diurna de propriedades ópticas de hidrometeoros de nuvens sobre a Amazônia. Para tanto, medidas de radiância efetuadas pelos canais 1, 2 e 4 do satélite geoestacionário GOES 13 para os anos de 2012, 2013, 2014 e 2015 foram aliadas a códigos computacionais de transferência radiativa visando a obtenção de estimativas de raios efetivos de gotas e partículas de gelo em nuvens convectivas. A variação temporal de parâmetros microfísicos ao longo do dia foi analisada durante as estações seca e úmida em dois locais prístinos e outros dois locais significativamente atingidos pela fumaça de queimadas na Amazônia. A profundidade óptica de aerossóis ( a em 550 nm) variou de 0,1 a 0,2 na maior parte do ano (estação úmida) sobre todos os locais. Na estação seca nos sítios prístinos observou-se um a em torno de 0,5 unidades, e de cerca de 0,8 nos sítios degradados. Os resultados mostram que para todos os locais analisados há 32% mais pixels de nuvens durante a estação úmida do que na seca. As distribuições relativas de refletâncias em 0,63 m e da temperatura de brilho em 11 m indicam que em todos os sítios e épocas do ano há predominância de nuvens menos espessas e mais quentes sobre a Amazônia. A análise da refletância em 3,90 m indicou que nos quatro locais ocorre uma redução do raio efetivo de hidrometeoros de nuvens quentes na estação seca em relação à estação úmida. A distribuição de raios efetivos é bimodal para todos os sítios e estações analisados, sendo a variação diurna dessa distribuição consistente com processos de desenvolvimento vertical de nuvens e crescimento de hidrometeoros. Esse mecanismo ocorre de maneira distinta em locais mais e menos poluídos, sendo que para regiões mais poluídas e desmatadas o desenvolvimento vertical de tamanhos de partículas na época seca se dá de maneira mais lenta do que na úmida. Para as áreas mais atingidas pela pluma de fumaça durante a estação seca os raios efetivos de gotas/cristais de gelo com temperatura de brilho maior que -20°C praticamente não mudam, sofrendo uma variação máxima de 2 m num período de 2 horas. Para o mesmo intervalo de temperaturas e de tempo a estação seca em ambientes mais limpos apresenta uma variação de até 6 m nos raios efetivos das partículas. Esse resultado é parcialmente compatível com modelos conceituais que procuram explicar efeitos microfísicos de aerossóis sobre o tamanho de hidrometeoros em nuvens. O atraso no crescimento vertical de hidrometeoros é mais pronunciado perto do meio dia solar e em locais onde as concentrações de aerossóis provenientes de queimadas são maiores. / Clouds play a key role in Earths radiative balance. The knowledge of its micro and macrophysical properties is important for the study of global climate. The life cycle of convective clouds is closely related to the microphysics of its hydrometeors, which are influenced by many factors including variations in the concentration of atmospheric aerosols. This study uses remote sensing by a satellite to analyze the diurnal evolution of reflective properties of clouds over Amazon. Radiance measurements performed by channels 1, 2 and 4 of the imager instrument onboard GOES-13 geostationary satellite, from 2012 to 2015, were analyzed using radiative transfer and computational codes. This allowed deriving estimates of the effective radius of cloud droplets and ice particles in convective clouds. The temporal variation of microphysical parameters throughout the day was analyzed during the dry and wet seasons at two pristine sites and two other sites significantly affected by biomass burning smoke in the Amazon. The aerosol optical depth ( a at 550 nm) ranged from 0.1 to 0.2 for most of the year (wet season) over all sites. In the dry season at the pristine sites a a of about 0.5 units was observed, while about 0.8 units were measured at the degraded sites. The results show that for all analyzed sites there were 32% more cloudy pixels during the wet season than in the dry season. The relative distribution of reflectance at 0.63 m and the brightness temperature at 11 m indicate that at all sites and times of the year there is a predominance of shallow warm clouds in the Amazon. The analysis of the reflectance at 3.90 m indicated that at the four sites a reduction of the effective radius of hydrometeors in warm clouds occurs in the dry season in comparison to the wet season. The distribution of effective radius is bimodal for all sites and seasons. The diurnal variation of this bimodal distribution is consistent with processes of vertical cloud development and hydrometeor growth. This mechanism occurs differently in the pristine and degraded sites. At polluted and deforested regions the vertical development of particle sizes in the dry season occurs more slowly than in pristine ones. For the areas more affected by smoke plumes during the dry season the effective radius of drops/ice crystals of clouds with brightness temperature greater than -20°C show small changes with height, undergoing a maximum variation of 2 m in 2 hours. For the same temperature range and time interval in the dry season, clouds in cleaner environments showed a variation up to 6 m in the effective radius of particles. This result is partly compatible with conceptual models that seek to explain microphysical effects of aerosols on the size of hydrometeors. The vertical growth delay of hydrometeors is more pronounced near local solar noon and in places where the concentration of smoke aerosols is higher.

This thesis addresses a series of questions related to the problem of achieving reliable and physically consistent representations of aerosol-cloud interaction in global circulation models (GCM). In-situ data and modeling tools are used to develop and evaluate novel parameterization schemes for the process of aerosol activation for applications in GCM simulations. Atmospheric models of different complexity were utilized, ranging from detailed Lagrangian parcel model simulations of the condensational growth of droplets, to one-dimensional single column model with aerosol and cloud microphysics, and finally GCM simulations performed with the Community Atmosphere Model (CAM). A scheme for mapping the sub-grid scale variability of cloud droplet number concentrations (CDNC) to a number of microphysical process rates in a GCM was tested, finding that neglecting this impact can have substantial influences in the integrated cloud properties. A comprehensive comparison and evaluation of two widely used, physically-based activation parameterizations was performed in the framework of CAM5.1. This was achieved by utilizing a numerical adjoint sensitivity approach to comprehensively investigate their response under the wide range of aerosol and dynamical conditions encountered in GCM simulations. As a result of this, the specific variables responsible for the observed differences in the physical response across parameterizations are encountered, leading to further parameterization improvement.

La prévision de l'intensité des cyclones tropicaux est aujourd'hui un enjeu scientifique majeur. Parmi de nombreux facteurs multi-échelle, l'impact de la microphysique nuageuse et des aérosols sur les variations d'intensité a été récemment mis en évidence. Cette problématique a motivé l'évaluation du schéma microphysique à 2-moments LIMA en milieu tropical et le développement d'un couplage avec le schéma d'aérosols ORILAM au sein du modèle atmosphérique Meso-NH. L'intérêt de ce développement numérique est d'inclure l'émission des aérosols marins en fonction des vents cycloniques et des paramètres océaniques. L'application de ce couplage aérosols-microphysique à la simulation du cyclone tropical Dumile (2013) montre que le modèle couplé tend à améliorer la représentation de l'intensité, la trajectoire, la structure microphysique du cyclone tropical et les précipitations associées, en comparaison avec les observations. La production secondaire des cristaux de glace est également un thème de recherche actif en microphysique nuageuse. Ainsi, une paramétrisation du processus de rupture collisionnelle de la glace a été implémentée dans le schéma microphysique LIMA. L'impact de ce processus a été testé sur le développement d'un orage des moyennes latitudes et sur le cyclone tropical Dumile. Les deux cas d'étude ont des réponses similaires vis-à-vis de ce processus : une augmentation de la concentration et de la masse des cristaux de glace et une diminution des cumuls de précipitations. La poursuite de ces travaux pourrait permettre de déterminer si ce processus de formation secondaire peut améliorer la modélisation de la couverture cirriforme des cyclones tropicaux. / Intensity forecast of tropical cyclones is a major scientific issue. Among many factors, the impact of cloud microphysics and aerosols on intensity variations has been recently underlined. This issue motivated the evaluation of the 2-moment microphysical scheme LIMA in a tropical context and the development of a coupling with the aerosol scheme ORILAM into the atmospheric model Meso-NH. The interest of this numerical development is to represent the emission of sea salt aerosols depending on cyclonic winds and oceanic parameters. The application of this aerosols-microphysics coupling to the simulation of tropical cyclone Dumile (2013) shows that the coupled model tends to improve the representation of the intensity, the track, the microphysical structure of the tropical cyclone and the associated precipitation, when comparing with observations. The secondary production of ice crystals is also an active research topic in cloud microphysics. A parameterization of the collisional ice break-up process is thus implemented into the microphysical scheme LIMA. The impact of this process has been analyzed on a mid-latitude storm and on tropical cyclone Dumile. Both case studies display similar results regarding this process: an increase of ice crystals concentration and mass, and a decrease of precipitation. The continuation of this work could allow to determine if this process of secondary formation could improve the cirrus modelling in tropical cyclones.

Nuvens desempenham um papel fundamental no balanço radiativo terrestre, e o conhecimento de suas propriedades micro e macrofísicas é importante para o estudo do clima global. O desenvolvimento convectivo de nuvens está intimamente relacionado ao comportamento microfísico de seus hidrometeoros, os quais são influenciados pela variação nas concentrações de aerossóis disponíveis na atmosfera. Este trabalho utiliza o sensoriamento remoto por satélites para analisar a evolução diurna de propriedades ópticas de hidrometeoros de nuvens sobre a Amazônia. Para tanto, medidas de radiância efetuadas pelos canais 1, 2 e 4 do satélite geoestacionário GOES 13 para os anos de 2012, 2013, 2014 e 2015 foram aliadas a códigos computacionais de transferência radiativa visando a obtenção de estimativas de raios efetivos de gotas e partículas de gelo em nuvens convectivas. A variação temporal de parâmetros microfísicos ao longo do dia foi analisada durante as estações seca e úmida em dois locais prístinos e outros dois locais significativamente atingidos pela fumaça de queimadas na Amazônia. A profundidade óptica de aerossóis ( a em 550 nm) variou de 0,1 a 0,2 na maior parte do ano (estação úmida) sobre todos os locais. Na estação seca nos sítios prístinos observou-se um a em torno de 0,5 unidades, e de cerca de 0,8 nos sítios degradados. Os resultados mostram que para todos os locais analisados há 32% mais pixels de nuvens durante a estação úmida do que na seca. As distribuições relativas de refletâncias em 0,63 m e da temperatura de brilho em 11 m indicam que em todos os sítios e épocas do ano há predominância de nuvens menos espessas e mais quentes sobre a Amazônia. A análise da refletância em 3,90 m indicou que nos quatro locais ocorre uma redução do raio efetivo de hidrometeoros de nuvens quentes na estação seca em relação à estação úmida. A distribuição de raios efetivos é bimodal para todos os sítios e estações analisados, sendo a variação diurna dessa distribuição consistente com processos de desenvolvimento vertical de nuvens e crescimento de hidrometeoros. Esse mecanismo ocorre de maneira distinta em locais mais e menos poluídos, sendo que para regiões mais poluídas e desmatadas o desenvolvimento vertical de tamanhos de partículas na época seca se dá de maneira mais lenta do que na úmida. Para as áreas mais atingidas pela pluma de fumaça durante a estação seca os raios efetivos de gotas/cristais de gelo com temperatura de brilho maior que -20°C praticamente não mudam, sofrendo uma variação máxima de 2 m num período de 2 horas. Para o mesmo intervalo de temperaturas e de tempo a estação seca em ambientes mais limpos apresenta uma variação de até 6 m nos raios efetivos das partículas. Esse resultado é parcialmente compatível com modelos conceituais que procuram explicar efeitos microfísicos de aerossóis sobre o tamanho de hidrometeoros em nuvens. O atraso no crescimento vertical de hidrometeoros é mais pronunciado perto do meio dia solar e em locais onde as concentrações de aerossóis provenientes de queimadas são maiores. / Clouds play a key role in Earths radiative balance. The knowledge of its micro and macrophysical properties is important for the study of global climate. The life cycle of convective clouds is closely related to the microphysics of its hydrometeors, which are influenced by many factors including variations in the concentration of atmospheric aerosols. This study uses remote sensing by a satellite to analyze the diurnal evolution of reflective properties of clouds over Amazon. Radiance measurements performed by channels 1, 2 and 4 of the imager instrument onboard GOES-13 geostationary satellite, from 2012 to 2015, were analyzed using radiative transfer and computational codes. This allowed deriving estimates of the effective radius of cloud droplets and ice particles in convective clouds. The temporal variation of microphysical parameters throughout the day was analyzed during the dry and wet seasons at two pristine sites and two other sites significantly affected by biomass burning smoke in the Amazon. The aerosol optical depth ( a at 550 nm) ranged from 0.1 to 0.2 for most of the year (wet season) over all sites. In the dry season at the pristine sites a a of about 0.5 units was observed, while about 0.8 units were measured at the degraded sites. The results show that for all analyzed sites there were 32% more cloudy pixels during the wet season than in the dry season. The relative distribution of reflectance at 0.63 m and the brightness temperature at 11 m indicate that at all sites and times of the year there is a predominance of shallow warm clouds in the Amazon. The analysis of the reflectance at 3.90 m indicated that at the four sites a reduction of the effective radius of hydrometeors in warm clouds occurs in the dry season in comparison to the wet season. The distribution of effective radius is bimodal for all sites and seasons. The diurnal variation of this bimodal distribution is consistent with processes of vertical cloud development and hydrometeor growth. This mechanism occurs differently in the pristine and degraded sites. At polluted and deforested regions the vertical development of particle sizes in the dry season occurs more slowly than in pristine ones. For the areas more affected by smoke plumes during the dry season the effective radius of drops/ice crystals of clouds with brightness temperature greater than -20°C show small changes with height, undergoing a maximum variation of 2 m in 2 hours. For the same temperature range and time interval in the dry season, clouds in cleaner environments showed a variation up to 6 m in the effective radius of particles. This result is partly compatible with conceptual models that seek to explain microphysical effects of aerosols on the size of hydrometeors. The vertical growth delay of hydrometeors is more pronounced near local solar noon and in places where the concentration of smoke aerosols is higher.

The Second Aerosol Characterization Experiment (ACE-2) data set along with ECMWF reanalysis meteorological fields provided the basis for the single column model (SCM) simulations, performed as part of the PACE (Parameterization of the Aerosol Indirect Climatic Effect) project. Six different SCMs were used to simulate ACE-2 case studies of clean and polluted cloudy boundary layers, with the objective being to identify limitations of the aerosol/cloud/radiation interaction schemes within the range of uncertainty in in situ, reanalysis and satellite retrieved data. The exercise proceeds in three steps. First, SCMs are configured with the same fine vertical resolution as the ACE-2 in situ data base to evaluate the numerical schemes for prediction of aerosol activation, radiative transfer and precipitation formation. Second, the same test is performed at the coarser vertical resolution of GCMs to evaluate its impact on the performance of the parameterizations. Finally, SCMs are run for a 24–48 hr period to examine predictions of boundary layer clouds when initialized with large-scale meteorological fields. Several schemes were tested for the prediction of cloud droplet number concentration (N). Physically based activation schemes using vertical velocity show noticeable discrepancies compared to empirical schemes due to biases in the diagnosed cloud base vertical velocity. Prognostic schemes exhibit a larger variability than the diagnostic ones, due to a coupling between aerosol activation and drizzle scavenging in the calculation of N. When SCMs are initialized at a fine vertical resolution with locally observed vertical profiles of liquid water, predicted optical properties are comparable to observations. Predictions however degrade at coarser vertical resolution and are more sensitive to the mean liquid water path than to its spatial heterogeneity. Predicted precipitation fluxes are severely underestimated and improve when accounting for sub-grid liquid water variability. Results from the 24–48 hr runs suggest that most models have problems in simulating boundary layer cloud morphology, since the large-scale initialization fields do not accurately reproduce observed meteorological conditions. As a result, models significantly overestimate optical properties. Improved cloud morphologies were obtained for models with subgrid inversions and subgrid cloud thickness schemes. This may be a result of representing subgrid scale effects though we do not rule out the possibility that better large-forcing data may also improve cloud morphology predictions.

With a nearly twice as strongly pronounced temperature increase compared to that of the Northern Hemisphere, the Arctic is especially susceptible to global climate change. The effect of clouds on the Arctic warming is especially uncertain, which is caused by misrepresented cloud microphysical processes in atmospheric models. This thesis aims at employing a scale- and definition-aware comparison of models and observations and will propose changes how to better parameterize Arctic clouds in atmospheric models. In the first part of this thesis, ECHAM6, which is the atmospheric component of the MPI-ESM global climate model, is compared to spaceborne lidar observations of clouds from the CALIPSO satellite. This comparison shows that ECHAM6 overestimates Arctic low-level, liquid containing clouds over snow- and ice-covered surfaces, which consequently leads to an overestimated amount of radiative energy received by the surface. Using sensitivity studies, it is shown that the probable cause of the model biases in cloud amount and phase is related to misrepresented cloud microphysical parameterization (i.e., parameterization of the Wegener-Bergeron-Findeisen process and of the cloud cover scheme) in ECHAM6. By revising those processes, a better representation of cloud amount and cloud phase is achieved, which helps to more accurately simulated the amount of radiative energy received by the Arctic in ECHAM6. The second part of this thesis will focus on a comparison of kilometer-scale simulation with the ICON model to aircraft observations from the ACLOUD campaign that took place in May/June 2017 over the sea ice-covered Arctic Ocean north of Svalbard, Norway. By comparing measurements of solar and terrestrial surface irradiances during ACLOUD flights to the respective quantities in ICON, it is shown that the model systematically overestimates the transmissivity of the mostly liquid clouds during the campaign. This model bias is traced back to the way cloud condensation nuclei get activated into cloud droplets in the two-moment, bulk microphysical scheme used. By parameterizing subgrid-scale vertical motion as a function of turbulent kinetic energy, a more realistic CCN activation into cloud droplets is achieved. This consequently results in an improved representation of cloud optical properties in the ICON simulations. Furthermore, the results of two studies to which contributions have been made during the Ph.D. will be summarized. In Petersik et al. 2018, the impact of subgrid-scale variability in clear-sky relative humidity on hygroscopic growth of aerosols in the aerosol-climate model ECHAM6-HAM2 has been explored. It was shown that the revised parameterization of hygroscopic growth of aerosols resulted in a stronger swelling of aerosol particles, which consequently causes an increased backscattering of solar radiation. In the study of Costa-Suros et al. 2019, it is explored whether it is possible to detect and attribute aerosol-cloud interactions in large-eddy simulation over Germany. It was shown that an increase in cloud droplet number concentration could be attributed to an increased aerosol load, while such an attribution was not possible for other cloud micro- and macrophysical variables.

We propose a mathematical procedure to derive a stochastic parameterization for the bulk warm cloud micro-physical properties of collision and coalescence. Unlike previous bulk parameterizations, the stochastic parameterization does not assume any particular droplet size distribution, all parameters have physical meanings which are recoverable from data, all equations are independently derived making conservation of mass intrinsic, the auto conversion parameter is finely controllable, and the resultant parameterization has the flexibility to utilize a variety of collision kernels. This new approach to modelling the kinetic collection equation (KCE) decouples the choice of a droplet size distribution and a collision kernel from a cloud microphysical parameterization employed by the governing climate model. In essence, a climate model utilizing this new parameterization of cloud microphysics could have different distributions and different kernels in different climate model cells, yet employ a single parameterization scheme. This stochastic bulk model is validated theoretically and empirically against an existing bulk model that contains a simple enough (toy) collision kernel such that the KCE can be solved analytically. Theoretically, the stochastic model reproduces all the terms of each equation in the existing model and precisely reproduces the power law dependence for all of the evolving cloud properties. Empirically, values of stochastic parameters can be chosen graphically which will precisely reproduce the coefficients of the existing model, save for some ad-hoc non-dimensional time functions. Furthermore values of stochastic parameters are chosen graphically. The values selected for the stochastic parameters effect the conversion rate of mass cloud to rain. This conversion rate is compared against (i) an existing bulk model, and (ii) a detailed solution that is used as a benchmark. The utility of the stochastic bulk model is extended to include hydrodynamic and turbulent collision kernels for both clean and polluted clouds. The validation and extension compares the time required to convert 50\% of cloud mass to rain mass, compares the mean rain radius at that time, and used detailed simulations as benchmarks. Stochastic parameters can be chosen graphically to replicate the 50\% conversion time in all cases. The curves showing the evolution of mass conversion that are generated by the stochastic model with realistic kernels do not match corresponding benchmark curves at all times during the evolution for constant parameter values. The degree to which the benchmark curves represent ground truth, i.e. atmospheric observations, is unknown. Finally, among alternate methods of acquiring parameter values, getting a set of sequential values for a single parameter has a stronger physical foundation than getting one value per parameter, and a stochastic simulation is preferable to a higher order detailed method due to the presence of bias in the latter. / Graduate / 0725 0608 0405 / davidc@uvic.ca