Damage from large hail is a costly problem in Canada and other parts of the world. With continuous increase in computer power, numerical weather prediction (NWP) models are getting closer to resolving the convective scale, where it is appropriate to employ explicit microphysics schemes. / This dissertation examines the potential to predict hail using an explicit microphysics scheme in an NWP model. To this end, the requirements of a bulk scheme to model the hydrometeor size distribution are investigated. A number of schemes in the literature use the double-moment method with a three-parameter gamma distribution function to represent the size spectrum. In general, two of the parameters vary with the predicted moments while the third, normally the shape parameter, is held constant. In a simple 1D context, the role of the shape parameter is analyzed by comparing results from bulk schemes using different numbers of predicted moments and an analytic bin model. It is shown that this parameter is important in the overall prediction of the size distribution by affecting both the instantaneous growth rates and the sedimentation. In view of this, two alternatives to the fixed-value approach are presented. One is a double-moment method, where the shape parameter is diagnosed from the predicted moments; the other is a triple-moment approach, where all distribution parameters are fully prognosed. / A new microphysics scheme using the proposed approaches has been designed and interfaced with the Mesoscale Compressible Community model (MC2). High-resolution (1 km) simulations of a severe hailstorm are conducted. The control simulation using the full triple-moment version of the scheme is compared to radar observations and is shown to realistically simulate the observed storm, including the spatial distribution and sizes of hail at the ground. Experiments are performed to determine the sensitivity of the different approaches on the simulation of hail. The results indicated that the triple-moment scheme gives the best results. For a double-moment scheme, the diagnostic approach for the shape parameter exhibits distinct improvement over the fixed-value approach. It is also shown that double-moment schemes are dramatically better in reproducing the control simulation than single-moment schemes, owing largely to the ability of multi-moment schemes to account for the effects of size-sorting.

Winter precipitation is an important issue in Canada because of its common occurrence and associated destructive consequences. Prediction of the precipitation type when temperatures are near 0°C is often difficult because so many types can occur. This study examines the microphysics of ice pellet formation, in particular the ability of these to form aggregates and the consequences of these aggregates. This issue was examined by modelling the freezing of a distribution of precipitation particles as they fall through the atmosphere and interact through collisions. Three mechanisms for aggregation were examined, collisions among the particles involved in these mechanisms were modelled and the relative importance of each mechanism was determined. It is shown that, for the conditions considered, aggregates are often able to collect freezing rain drops and that aggregation can sometimes be very effective at eliminating freezing rain but the conditions need to be precise for this to occur. / Les divers types de précipitations observées durant les tempêtes hivernales sont souvent la cause d'inconvénients durant cette période au Canada. Il est difficile de prédire ces divers types de précipitations du fait de leur sensibilité à certaines conditions atmosphériques, en particulier à des températures près de 0°C. Cette étude examine la microphysique de la formation des granules de glace. Plus précisément, la capacité de ces granules de former des agrégats et les conséquences de ces agrégats sur les autres types de précipitation présents. Cette recherche repose sur l'étude du regel d'une distribution de pluie verglaçante dans une atmosphère sous le point de congélation à l'aide de simulations incluant les interactions entre particules. Une attention particulière a été prêtée sur trois principaux aspects. Premièrement, trois mécanismes formant des agrégats de particules ont été étudiés. Deuxièmement, les collisions parmi les particules entraînées dans ces mécanismes ont été modélisées. Finalement, l'importance relative de chaque mécanisme a été déterminée. Les résultats illustrent que pour les conditions atmosphériques considérées, la pluie verglaçante est souvent collectée par les agrégats de particules formés durant leur descente dans l'atmosphère. De plus, l'agrégation de particules s'avère efficace à l'élimination complète de la pluie verglaçante dans des conditions atmosphériques précises et favorables.

Intercomparison among three sensors for ground-based measurement of DSD's are presented, the Joss-Waldvogel Distrometer (JWD), the Optical Spectro-Pluviometer (OSP), and the Precipitation Occurrence Sensor System (POSS). This is done in order to quantifying any systematic bias that can be identified in their measurements, and to evaluate a new data acquisition system developed for the JWD. / Data corresponding to stratiform, continental rain is analyzed, and a function (delta) is defined in order to provide a quantification of the relative differences between pairs of instruments. Significant differences in drop concentrations simultaneously measured by the instruments are found, being these about 100% at the extremes of the diameter spectra. Bulk rain quantities, rain rate (R) and reflectivity (Z), do not show significant biases in POSS and OSP, although the Z-R relationships present a strong dependence with the sensor and the method for obtaining them. / Analyzing an event of convective precipitation, it is concluded that neither POSS nor JWD can be used unattended without a criterion that filters observations during non-liquid precipitation. / A problem in the new data acquisition system for the MID is identified, and a future project for its calibration is proposed.

Two cases of supercooled precipitation were observed with the McGill wind profiler and X-band Vertically Pointing Radar, respectively. All the relevant spectra were displayed to show some characteristics of the supercooled water, and the interaction between different precipitation components. The empirical relationships between reflectivity and precipitation content are employed to retrieve the snow content and the content of the supercooled liquid water. An alternative method for the supercooled liquid water was also tried and provided a similar result. The results of the calculations lie in the reasonable range of values as compared to previous research result. Some special attention is focused on the study of the precipitation trail, attempting to get some additional information.

Spectra for various meteorological fields are computed using the two-dimensional Fourier transform technique on a limited-area grid of the Regional finite-element (RFE) model and the Canadian Regional Climate Model (CRCM). Limitations of removing the linear trend from the fields, and thereby removing the large-scale variance, are discussed. The spectra of difference fields are calculated, yielding the error variance for different scales. / Spectral methods are used widely in the evaluation of global models. In this study, the same method is used for evaluating the CRCM in its ability to correctly reproduce the mesoscale systems in short-term integrations, when low-resolution GCM-like initial and lateral boundary conditions are provided. Two cases have been chosen for this study, the first one over the Montreal region and second one over the Mackenzie River Basin (MRB). It is found that the relative error variance growth at most scales, particularly the small scales, is less for the MRB region possibly due to the topographic forcing. In both cases and all experiments, the maximum relative error variance is found to be at a wavelength of about 350-km. / Root mean square (rms) error and relative rms error for the geopotential height field for both cases are very small and show little or no growth, when scale decompositions are not made. However, the relative error variance when examined according to scale, show considerable differences. The relative errors at different scales show different growth rates and that of the meso-alpha and synoptic scales are found to be growing with time.

The representativeness of North Atlantic cyclone activity is assessed for the 15 January-15 March 1992 period of the second Canadian Atlantic Storms Program (CASP II). Explosive cyclogenesis is found to shift northeastward to a locus east of the Canadian seaboard along a sustained anomalous cyclone track. / Active (inactive) periods are defined as comprising the quarter of the study period with the strongest (weakest) cyclone deepening: Of 24 Atlantic bombs, 14 (1) occur(s) during active (inactive) periods. During active periods, storms develop along strong baroclinic zones on the leading edge of southeasterly advancing Arctic air masses, and lead low-frequency oscillations in the position of an intensified Atlantic jet stream. Cyclone deepening shows heightened response to 500 hPa vorticity advection during active periods. Maximum deepening positions of explosive (weak) cases are preconditioned by negative (positive) 1000-500 hPa thickness anomalies 36-12 h ahead of the developing cyclones.

In this thesis we examine the possible causes of the model-to-model variability observed in the Comparison of Mesoscale Prediction and Research Experiments (COMPARE). A series of sensitivity experiments is performed involving changes to model initial conditions, physical parameterizations, and numerics. / Secondary cyclogenesis is most sensitive to latent heat release and moisture fluxes from the ocean surface. Initial differences are found to decay with time, and more rapidly so, in a 'dry' simulation. Sea-level pressures for the low center are lower in integrations with a longer time step, owing to greater rainrates and low-level diabatic heating at the storm center from 12 to 18 h. / Moisture fluxes in the first 12 h are crucial in creating grid-box saturation for initiation of precipitation, and the generation of a low-level potential vorticity maximum, with fluxes during the subsequent 24-h period of rapid deepening not as crucial. A westward storm track bias in COMPARE is related to a more robust low-level potential vorticity maximum along the coast, as compared to the maximum offshore.

The responses to positive and negative sea surface temperature (SST) anomalies in the North Pacific are computed with a time-dependent, quasi-geostrophic, global spectral model with a T21 horizontal resolution and three levels in the vertical. The simplicity of the model allows a large number of cases starting from different initial conditions to be run. The model produces on average a ridge (low) downstream of the warm (cold) SST anomaly, but the average response to the warm anomaly is much weaker and statistically less significant than that to the cold anomaly. In the case of the warm SST anomaly, the storm track is displaced northward into the high-pressure atmospheric anomaly, whereas in the case of the cold SST anomaly, the storm track is moved southward, away from the atmospheric low-pressure anomaly. The higher level of atmospheric nonlinearity in the warm cases leads to more case-to-case variability in the model response to the SST anomaly than for the cold SST anomaly. The results are compared with those of previous work in the literature where the response of a GCM to a warm SST anomaly was found to be weaker and statistically less significant than that to a cold anomaly.

The transformation of 45 tropical cyclones into extratropical cyclones over the western Noah Atlantic Ocean between 1963 and 1996 is studied. Cases are selected from the National Hurricane Center's "best track" archive. National Centers, for Environmental Prediction (NCEP) reanalyses of geopotential height data are used to construct a synoptic-dynamic climatology of extratropical transition, or "ET". The Forecast Systems Laboratory (FSL) upper-air archive of six near-track stations is used to produce sounding composites. / Primary results of the study follow. (1) A statistically-significant 1000--500-hPa warm anomaly (with respect to the 1963--96 climatology) persists for the one-week period prior to the passage of the tropical systems into the Canadian Maritime provinces. (2) A northwestward extension of the surface subtropical anticyclone exists over the Canadian Atlantic Provinces during the two-day period prior to the arrival of the cyclone. (3) The tropical cyclone's warm core and conditionally-unstable tropical airmass are maintained after transition. (4) The presence of quasi-geostrophic forcing for ascent, typically seen in extratropical cyclones, is observed during periods in which the systems are still classified as tropical cyclones. This forcing for ascent continues during the extratropical transformation, and typically occurs ahead and to the left of the storm track.

Despite advances in mesometeorology and computer technology, high-resolution numerical simulations of small-scale severe weather events remain extremely challenging. This is primarily due to insufficient initial conditions and inadequate convective parameterization schemes (CPSs). This thesis serves to illustrate how these difficulties may be overcome in a real-data simulation of the Garden City, Kansas, tornadic thunderstorm, which occurred during the VORTEX (Verification of the Origins of Rotation in Tornadoes Experiment) field experiment of 1995. / Using a sophisticated mesoscale model at 18 km horizontal resolution, a successful reproduction of the environment over southwestern Kansas is obtained. However, modifications to the CPS are required to trigger the Garden City storm at the correct time and location. Utilizing severe weather parameters, it is found that the simulated atmosphere is susceptible to tornadic supercells. The results of a sensitivity study also indicate that a neighbouring storm may have influenced tornadogenesis in the Garden City supercell.