The study of anthropogenic climate change is a research area of vital importance for the coming decades, with rainfall change and variability expected to be keenly felt in vulnerable regions of the world, including Africa. The focus of this study is daily rainfall during the short rains season over East Africa from October to December, which has one of the most coherent increasing rainfall projections in the Coupled Model Intercomparison Project (CMIP) dataset. This thesis aims to examine the fidelity of coupled climate models over an East African domain, with the question approached through focus on the daily (rather than monthly) rainfall field. The self-organising map is used as a clustering tool to establish contemporary characteristics of daily rainfall events in reanalysis (ERA-40) and satellite (TRMM) rainfall datasets between 1971-2000 and 1998-2010 respectively for the East African short rains. Moisture flux divergence is found to be the circulation parameter that is most closely related to the presence of rainfall events or dry conditions over East Africa. Coupled climate models are poor at replicating the daily rainfall field over the domain. A key result of the analysis is the consistent overestimation of daily rainfall by climate models for days where dry conditions of suppressed convection should prevail. In contrast, the moisture flux divergence field maps well to dry nodes for days of the self-organising map array for the models. Dry days are associated with widespread anomalous moisture flux divergence and rainfall events with co-located anomalous moisture flux convergence. This is in agreement with the moisture flux divergence field in the ERA-40 dataset; it is the rainfall field where there is disagreement for days of suppressed convection. Twenty to thirty-five percent of the projected rainfall increase towards the end of the twenty-first century results from an increase in the proportion of days assigned to nodes of suppressed convection in six of the seven models and the ensemble mean. There is an accompanying projected rainfall increase associated with days assigned to these nodes. Such days in the 2090s are characterised by projected increased strength moisture flux divergence over East Africa. Given that the moisture flux field was more successfully simulated in the coupled models under contemporary conditions than the daily rainfall field, this suggests that rainfall projections under a high emissions scenario at the end of the twenty-first century are overestimated and that an important part of the key increase in the projected rainfall may not be real.

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.

While exact methods, such as DDA or T-matrix, can be applied to particles withsizes comparable to the wavelength, computational demands mean that they are size limited. For particles much larger than the wavelength, the Geometric Optics approximation can be employed, but in doing so wave effects, such as interference and diffraction, are ignored. In between these two size extremes there exists a need for computational techniques which are capable of handling the wide array of ice crystal shapes and sizes that are observed in cirrus clouds. The Beam Tracing model developed within this project meets these criteria. It combines aspects of geometric optics and physical optics. Beam propagation is handled by Snell's law and the law of reflection. A beam is divided into reflected and transmitted components each time a crystal facet is illuminated. If the incident beam illuminates multiple facets it is split, with a new beam being formed for each illuminated facet. The phase-dependent electric field amplitude of the beams is known from their ampli- tude (Jones) matrices. These are modified by transmission and reflection matrices, whose elements are Fresnel amplitude coefficients, each time a beam intersects a crystal facet. Phase tracing is carried out for each beam by considering the path that its 'centre ray' would have taken. The local near-field is then mapped, via a surface integral formulation of a vector Kirchhoff diffraction approximation, to the far-field. Once in the far-field the four elements of the amplitude matrix are trans- formed into the sixteen elements of the scattering matrix via known relations. The model is discussed in depth, with details given on its implementation. The physical basis of the model is given through a discussion of Ray Tracing and how this leads to the notion of Beam Tracing. The beam splitting algorithm is described for convex particles followed by the necessary adaptations for concave and/or ab- sorbing particles. Once geometric aspects have been established details are given as to how physical properties of beams are traced including: amplitude, phase and power. How diffraction is implemented in the model is given along with a review of existing diffraction implementations. Comparisons are given, first against a modified Ray Tracing code to validate the geometric optics aspects of the model. Then, specific examples are given for the cases of transparent, pristine, smooth hexagonal columns of four different sizes and orientations; a highly absorbing, pristine, smooth hexagonal column and a highly absorbing, indented, smooth hexagonal column. Analysis of two-dimensional and one-dimensional intensity distributions and degree of linear polarisation results are given for each case and compared with results acquired through use of the Amster- dam Discrete-Dipole Approximation (ADDA) code; with good agreement observed. To the author's best knowledge, the Beam Tracer developed here is unique in its ability to handle concave particles; particles with complex structures and the man- ner in which beams are divided into sub-beams of quasi-constant intensity when propagating in an absorbing medium. One of the model's potential applications is to create a database of known particle scattering patterns, for use in aiding particle classification from images taken by the Small Ice Detector (SID) in-situ probe. An example of creating such a database for hexagonal columns is given.

The South Saskatchewan River Basin (SSRB) is amongst the largest watersheds in Canada. It is an ecologically diverse region, containing Montane Cordillera, Boreal Plains and Prairie ecozones. The SSRB is subject to chinooks, which bring strong winds, high temperatures and humidity deficits that alter the storage of water during winter. Approximately 40% of winter days experience chinooks. Ablation during chinooks has not been quantified; it is not known how much water evaporates, infiltrates or runs off. The aim of this thesis is to characterise the spatial variability of surface water fluxes as affected by chinooks over SSRB subbasins and ecozones. The objectives are addressed using detailed field observations and physically based land surface modelling. Eddy covariance was deployed at three prairie sites. During winter chinooks, energy for large evaporative fluxes were provided by downward sensible heat fluxes. There was no evidence of infiltration until March. The Canadian Land Surface Scheme (CLASS) coupled to the Prairie Blowing Snow Model (PBSM) was used as the modelling platform. A multi-physics version of CLASSPBSM was developed, consisting of two parameterisation options each for sixteen processes. Field observations were used to evaluate each of the configurations. Three parameterisations provide both best snow and best soil water simulations: iterative energy balance solution, air temperature and wind speed based fresh snow density and de Vries’ soil thermal conductivity. The model evaluation highlighted difficulties simulating evaporation and uncertainty in simulating infiltration into frozen soils at large scales. A single model configuration is selected for modelling the SSRB. Modelling showed that the SSRB generally experiences no net soil water storage change until March, confirming field observations. Chinooks generally reduce net terrestrial water storage, largely due to snowmelt and subsequent evaporation and runoff. The Prairie ecozone is that which is most strongly affected by chinooks. The Montane Cordillera ecozone is affected differently by chinooks; blowing snow transport increases during winter and runoff increases during spring. The Lower South Saskatchewan is the subbasin most affected by chinooks. The Red Deer is the subbasin least affected by chinooks.

The thesis is presented in alternative format, meaning that the results of the thesis take the form of three journal articles, each telling a distinct story within the subject matter, but collectively highlighting the sensitivity of bands to frictional and diabatic processes. Paper 1 is an idealised-modelling study with the Weather Research and Forecasting (WRF) model, in which moist baroclinic waves are simulated from an initial zonally uniform mid-latitude jet on an f-plane at 20-km grid spacing, and the sensitivity of the resulting precipitation bands is explored. Paper 2 employs further WRF idealised-baroclinic-wave simulations and takes a simulation from Paper 1, after the cold front has formed, as the initial condition. A nested domain at 4-km grid spacing is inserted when this simulation is re-initialised to invesigate the sensitivity of finer-scale precipitation cores along the surface cold front. In both Papers 1 and 2, friction and latent-heat release enhance multiple banding at the two distinct horizontal scales, while surface fluxes hinder multiple banding. Paper 3 studies post-frontal snowbands over the English Channel and Irish Sea during extreme cold-air outbreaks in the winters of 2009-10 and 2010-11, via a climatology of precipitation-radar, sounding, and SST data, and real-data WRF sensitivity simulations of one such band over the English Channel. The observational and modelling results show that strong winds and large differential heat fluxes between land and sea were necessary to generate banded precipitation. Coastal orography and the land-sea frictional contrast aided the morphology of bands, but banded precipitation did still form in the absence of these influences in the sensitivity simulations. These three studies and the thesis as a whole highlight the role of frictional and diabatic processes in modifying various types of precipitation bands within baroclinic waves, and in generating bands that would otherwise not exist.

Η διατριβή αυτή μελετά τα σταθερά ισότοπα στον υετό της Ελλάδας. η γεωγραφική κατανομή τους είναι η θεωρητικά αναμενόμενη, ενώ οι βαθμίδες με τη θερμοκρασία και το υψόμετρο και η τοπική μετεωρική γραμμή συμφωνούν με τη βιβλιογραφία για την ευρύτερη περιοχή. Η επίδραση της πορείας της αέριας μάζας, δεν είναι ευδιάκριτη λόγω της επίδρασης τοπικών διεργασιών κλασμάτωσης. Επίσης, η ισοτοπική σύσταση των πηγών ακολουθεί την κατανομή της σύστασης του υετού και είναι γενικά απεμπλουτισμένη σε σχέση με αυτόν. Ελέγχεται η καταλληλότητα των δεδομένων για κλιματολογική θεώρηση των ισότοπων του υετού στην κεντρική και ανατολική Μεσόγειο. Οι υπάρχουσες τάσεις σπανίως είναι σημαντικές και δεν έχουν χωρικά οργανωμένες μορφές. Επομένως, έχει νόημα ο υπολογισμός κλιματικών τιμών και η ανάπτυξη, με χρήση συνδυασμού προτύπων παλινδρόμησης και γεωστατιστικών μεθόδων, πλεγματικών κλιματικών ισοτοπικών δεδομένων και πλεγματικών ισοτοπικών χρονοσειρών. Τέλος, ερευνάται η ικανότητα των συνοπτικών ταξινομήσεων να οδηγούν σε ισοτοπικά διακριτές κλάσεις, αναπαράγοντας, ταυτόχρονα, σημαντικό μέρος της παρατηρούμενης ισοτοπικής μεταβλητότητας. Γενικά οι ταξινομήσεις βελτιστοποίησης με σχετικά μεγάλο αριθμό κλάσεων ικανοποιούν τα ανωτέρω. επιχειρείται η ανασύσταση ισοτοπικών χρονοσειρών χρησιμοποιώντας μέσες τιμές, σταθμισμένες βάσει των συχνοτήτων εμφάνισης των συνοπτικών κλάσεων και των υψών υετού. Η βέλτιστη επίδοση επιτυγχάνεται από εποχικές ταξινομήσεις με κλάσεις που αναπαράγουν τις βασικές ατμοσφαιρικές κυκλοφορίες, ακόμα και αν αυτό δεν συνεπάγεται βέλτιστους στατιστικούς δείκτες. / this thesis studies the stable isotpes in precipitation over greece. the geographical distribution is the one theoreticaly expected, while the temperature and altitude gradients, and the local meteoric line are onstistent with the literature for the area. the influence of air mass trajectory is not clear due to the effect of local fractionation processes. also, the composition of springs follows that of rain and is generaly depleted compared to it. the suitability of existing central and eastern mediterranean data, for climatological purposes is examined. existing trends are spatially inconsistent and rarely significant. thus, it is possible to calculate climatic values and obtain, using a combination of regression models and geostatistical methods, gridded isotopic climatologies and time series. finally, the ability of synoptic classifications to provide isotopicaly distinct classes, while at the same time reproducing the observed isotopic variability, is assessed. optimization classifications with a relatively large number of classes fulfill these criteria. reconstruction of isotope time series using mean values weighted according to synoptic class frequency and rain amount is possible. the best results are obtained from seasonal classifications with classes representing the basic atmospheric circulations, even if this does not lead to optimized statistical indices.

A detailed understanding of cloud microphysical processes is crucial for a large range of scientific disciplines that require knowledge of cloud particles for accurate climate and weather prediction. This thesis focuses on 3 measurement campaigns, encompassing both airborne and ground based measurements of the microphysical structures observed in cold, warm and occluded frontal systems around the United Kingdom, stratocumulus clouds in the Arctic and many different clouds observed over a 6 week period at a high-alpine site in the Swiss Alps. Particular attention was paid to the origin and distribution of both primary and secondary ice and the dominant features associated with ice phase processes. During investigation of cold, warm and occluded frontal systems associated with mid-latitude cyclones around the U.K., secondary ice was often found to dominate the number and mass concentrations of ice particles in all systems. The presence of large liquid droplets was sometimes observed in close proximity to regions of secondary ice production. The existence of these provides a possible mechanism by which rime-splintering is greatly enhanced through the creation of instant rimers as the large drops freeze. In-situ measurements during the cold frontal case were used to calculate rates of diabatic heating during a comparison between bin-resolved and bulk microphysics schemes. Observations in arctic stratocumulus clouds during spring and summer seasons revealed higher ice concentrations in the summer cases when compared to the spring season. This is attributed to secondary ice production actively enhancing ice concentrations in the summer, due to the higher temperature range the clouds spanned. At Jungfraujoch in the Swiss Alps, ground based measurements allowed us to obtain high spatial scale resolution measurements of cloud microphysics and we found transitions between high and low ice mass fractions that took place on differing temporal scales spanning seconds to hours. During the campaign measurements of aerosol properties at an out of cloud site, Schilthorn, were made. When analysing a Saharan Dust Event that took place a possible link between the number of U.V. fluorescent particles and the number of ice particles was found in the temperature range around -10 ºC.

A lack of quality data on high clouds has led to inadequate representations within global weather and climate models. Recent advances in spaceborne measurements of the Earth’s atmosphere have provided complementary information on the interior of these clouds. This study demonstrate how an array of space-borne measurements can be used and combined, by close co-located comparisons in space and time, to form a more complete representation of high cloud processes and properties. High clouds are found in the upper atmosphere, where sub-zero temperatures frequently result in the formation of cloud particles that are composed of ice. Weather and climate models characterise the bulk properties of these ice particles to describe the current state of the cloud-sky atmosphere. By directly comparing measurements with simulations undertaken at the same place and time, this study demonstrates how improvements can be made to the representation of cloud properties. The results from this study will assist in the design of future cloud missions to provide a better quality input. These improvements will also help improve weather predictions and lower the uncertainty in cloud feedback response to increasing atmospheric temperature. Most clouds are difficult to monitor by more than one instrument due to continuous changes in: large-scale and sub-cloud scale circulation features, microphysical properties and processes and characteristic chemical signatures. This study undertakes co-located comparisons of high cloud data with a cloud ice dataset reported from the Microwave Limb Sounder (MLS) instrument onboard the Aura satellite that forms part of the A-train constellation. Data from the MLS science team include vertical profiles of temperature, ice water content (IWC) and the mixing ratios of several trace gases. Their vertical resolutions are 3 to 6 km. Initial investigations explore the link between cloud-top properties and the longwave radiation budget, developing methods for estimating cloud top heights using; longwave radiative fluxes, and IWC profiles. Synergistic trios of direct and indirect high cloud measurements were used to validate detections from the MLS by direct comparisons with two different A-train instruments; the NASA Moderate-resolution Imaging Spectroradiometer (MODIS) and the Clouds and the Earth’s Radiant Energy System (CERES) onboard on the Aqua satellite. This finding focuses later studies on two high cloud scene types that are well detected by the MLS; deep convective plumes that form from moist ascent, and their adjacent outflows that emanate outwards several hundred kilometres. The second part of the thesis identifies and characterises two different high cloud scenes in the tropics. Direct observational data is used to refine calculations of the climate sensitivity to upper tropospheric humidity and high cloud in different conditions. The data reveals several discernible features of convective outflows are identified using a large sample of MLS data. The key finding, facilitated by the use of co-location, reveals that deep convective plumes exert a large longwave warming effect on the local climate of 52 ± 28Wm−2, with their adjacent outflows presenting a more modest warming of 33 ± 20Wm−2.

Dans cette thèse, nous nous intéressons à la représentation et à la prise en compte des incertitudes dans les systèmes de prévision hydrologique probabilistes à moyen-terme. Ces incertitudes proviennent principalement de deux sources : (1) de l’imperfection des prévisions météorologiques (utilisées en intrant de ces systèmes) et (2) de l’imperfection de la représentation du processus hydrologique par le simulateur pluie-débit (SPQ) (au coeur de ces systèmes).La performance d’un système de prévision probabiliste s’évalue par la précision de ses prévisions conditionnellement à sa fiabilité. L’approche statistique que nous suivons procure une garantie de fiabilité à condition que les hypothèses qu’elle implique soient réalistes. Nous cherchons de plus à gagner en précision en incorporant des informations auxiliaires.Nous proposons, pour chacune des sources d’incertitudes, une méthode permettant cette incorporation : (1) un post-traitement des prévisions météorologiques s’appuyant sur la propriété statistique d’échangeabilité et permettant la prise en compte de plusieurs sources de prévisions, ensemblistes ou déterministes ; (2) un post-traitement hydrologique utilisant les variables d’état des SPQ par le biais d’un modèle Probit arbitrant entre deux régimes hydrologiques interprétables et permettant ainsi de représenter une incertitude à variance hétérogène.Ces deux méthodes montrent de bonnes capacités d’adaptation aux cas d’application variés fournis par EDF et Hydro-Québec, partenaires et financeurs du projet. Elles présentent de plus un gain en simplicité et en formalisme par rapport aux méthodes opérationnelles tout en montrant des performances similaires. / In this thesis, we are interested in representing and taking into account uncertainties in medium term probabilistic hydrological prediction systems.These uncertainties mainly come from two sources: (1) from the imperfection of meteorological forecasts (used as inputs to these systems) and (2) from the imperfection of the representation of the hydrological process by the rainfall-runoff simulator (RRS) (at the heart of these systems).The performance of a probabilistic forecasting system is assessed by the sharpness of its predictions conditional on its reliability. The statistical approach we follow provides a guarantee of reliability if the assumptions it implies are complied with. We are also seeking to incorporate auxilary information to get sharper.We propose, for each source of uncertainty, a method enabling this incorporation: (1) a meteorological post-processor based on the statistical property of exchangeability and enabling to take into account several (ensemble or determistic) forecasts; (2) a hydrological post-processor using the RRS state variables through a Probit model arbitrating between two interpretable hydrological regimes and thus representing an uncertainty with heterogeneous variance.These two methods demonstrate adaptability on the various application cases provided by EDF and Hydro-Québec, which are partners and funders of the project. Those methods are moreover simpler and more formal than the operational methods while demonstrating similar performances.