Extra-Tropical Cyclones (ETCs) are responsible for billions of dollars worth of economic damage every year, and affect the entire mid-latitude regions of both hemispheres_ They are often termed windstorms by policymakers and industry, as the damage associated wi th their intense surface windspeeds is often widespread and catastrophic_ It is therefo.re of paramount importance to tmderstand these hazardous phenomena_ In recent years vast amounts of observational ocean surface winds peed data have been made available, owing to the advent of the satellite era coupled with substantial improvements in remotely sensed ocean surface windspeed retrieval methods. These data have provided the opportunity to evaluate the current representation of ETC windspeeds in climate models. forecast analyses, and climate reanalyses. This had previously been difficult due to the obvious logistical constraints of obtaining reliable and continuous in-situ windspeed observations in ETC environments. This study utilises two satellite derived observational datasets in an attempt to provide verification of ETC surface windspeed structure of ECM\VF and NASA produced models, analyses and reanalyses. A particular emphasis is made to examine the ability of the datascts to simulate the most intense windspeeds within ETCs. These often occur on a small scale but cause the vast majority of damage associated with ETCs.

No description available.

Mid-tropospheric cyclones are observed to occur quite frequently over the western Mediterranean region during early winter. The reasons for this phenomenon are not fully understood and a recent case was chosen for detailed diagnosis. It was necessary to be able to deduce the vertical motion fields at all stages of the process with high accuracy and a technique (based on the kinematic method) was developed for this purpose.

Hurricanes are the most destructive of all weather systems. Conditions necessary for their formation are well known, but not those sufficient to guarantee their occurrence. A thorough understanding of their structure, the physical and dynamical processes which operate within them, and their interactions with the large-scale environmental flow are necessary if accurate forecasts are to be made of their development and tracks. In the present study, both 2-D (axisymmetric) and 3-D (asymmetric), two-layer (plus the boundary layer) numerical models have been developed. The 2 -D model is used to investigate the essential characteristics of the symmetric flow component of tropical cyclones. The heating rate, which is computed at the mid-level in the model, is related to the vortexflow itself. It consists of three parts: moisture convergence over the whole column, Ekman pumping and surface evaporation. The moisture convergence is found to be crucial in providing latent heating released by condensation of water vapour sufficient to compensate adiabatic ascent cooling. The Ekman pumping plays an important role as a 'contractor', i.e. to cause the maximum winds to concentrate towards the centre, and consequent intensification of the modelled vortex. The surface evaporation, although relatively S11UJU, spreads over a larger area than the other two. It is found to beessentialformaintenance of the mature stage.

Over the last decade a number of studies have detected a degree of tropospheric predictability from knowledge of the stratospheric polar vortex, for example whether or not it is disturbed by planetary waves that propagate vertically from the troposphere, often resulting in sudden stratospheric warmings (SSWs). This thesis has built a comprehensive set of diagnostics to characterise the state of the stratospheric polar vortex using a 2D moment technique. This has allowed the area, centroid and aspect ratio of the vortex to be diagnosed in both reanalysis datasets and stratosphere resolving chemistry-climate models (CCMs). The first part of this thesis extends the technique of moment analysis so that it can be applied to all conceivable states of the vortex. These techniques are applied to the ERA-40 dataset and composites are built that characterise the structure and evolution of the vortex throughout the Northern Hemisphere winter and specifically during SSW events. It is found that use of the moment based diagnostics alone can determine whether the vortex is in a stable, displaced or split phase. The second part of this thesis compares the vortex diagnostics calculated for CCM output with those calculated from ERA-40 using a mixture of Gaussian and extreme value distributions. It is found that the models are able to capture the essential characteristics of SSW extreme events, but in general under estimate the frequency and over estimate the magnitude of these events.

This Thesis develops understanding of the physical mechanisms responsible for the frictional effects observed in cyclone development. To do this, a first-order closure mixing-length boundary layer scheme has been added to a baroclinic life cycle model to accurately represent the frictional processes occurring in cyclone development. Life cycles simulated with the model consist of normal mode baroclinic growth with cyclone development followed by barotropic decay. By considering life cycles where friction is the only diabatic process, it is found that surface drag reduces rates of baroclinic growth and barotropic decay by 40%. The classical description of frictional effects in rotating geophysical flows involves the Ekman spin-down of a barotropic vortex. This mechanism is studied by considering the quasi-geostrophic w-equation with a frictional term. However, these barotropic vortex ideas do not account for the baroclinic processes occurring, especially within the frontal regions. To address these shortcomings, a potential vorticity (PV) approach is adopted. Large frictionally generated positive PV anomalies form close to developing warm and cold fronts, due to the relative alignment of surface and thermal wind vectors. These PV anomalies are advected upwards and polewards along the warm conveyor belt and then westwards. This results in a band of positive PV associated with high static stability in the lower troposphere above the surface low centre. Using Rossby edge wave theory, a mechanism is proposed to explain the reduced baroclinic development observed in terms of this positive PV anomaly. Hence the baroclinic dynamics are shown to play a crucial role in the frictional modification of cyclone development. The classical notion of Ekman spin-down is shown to be of secondary importance. This mechanism by which frictional processes reduce cyclone development is found also to be valid in the presence of sensible and latent heat fluxes.

No description available.

Tropical cyclones are large weather systems whose development is characterised by complex physical processes that are not yet fully understood. A better understanding of how cyclones work is vital to help forecasters make more accurate predictions. The objective of this study is to contribute to the understanding of the track and intensity evolution in relation to the β-effect. While the β-effect has been widely studied, its response to changes in environmental conditions has been investigated only marginally. To this end a series of idealised experiments using the Weather Research and Forecasting model is performed. The sensitivity of the β-drift is investigated with no background flow by changing uniformly sea surface and atmospheric temperature. With increasing temperature tropical cyclones become more intense, larger and move faster with a stronger westward deflection. In particular, there is a linear relationship between the β-drift speed and size above a gale-force wind radius of 250 km. Another aspect investigated is the impact of the β-effect on cyclone intensity. A comparison between an experiment with variable Coriolis parameter f and one on an f -plane shows that after the initial intensification the former has a lower intensity and undergoes a weakening not present in the latter. The main reason behind this weakening is the dry intrusion of a mid-upper tropospheric westerly flow. Back-trajectories show that the intrusion originates from the anticyclonic recirculation of the air expelled from the cyclone deep convection at earlier times, eventually causing a self-weakening of the cyclone. A real-case simulation of Hurricane Katrina (2005) and two sensitivity experiments are performed to investigate the results of the idealised simulations. The occurrence of both the β-drift sensitivity to temperature and the dry intrusion further demonstrates that the results presented here might contribute to the general understanding of the intensity and track evolution of tropical cyclones.

Local periodic behaviour of storm tracks is found in the North Atlantic, the North Pacific, and in an aquaplanet global circulation model. Such periodicity is suggested to be a result of baroclinic instability that is characterised by nonlinear interactions between eddy fluxes and the mean-flow structure. This periodicity is associated with different processes at the beginning and end parts of the storm track. The beginning part exhibits maxima in both storm track activity and its growth rate, both of which oscillate temporarily in intensity akin to a predator-prey relationship. A nonlinear oscillator model is proposed to describe this relationship quantitatively, yielding a good agreement with atmospheric observations. It is predicted and observed that on average stronger storm events occur less frequently but are triggered more rapidly. Examination of the associated energetics suggests additional importance of the mean overturning circulation and the transport of mean available potential energy from polar latitudes, neither of which were included in the present model. Towards the end of the storm track, the dynamics are characterised more by variability in eddy momentum fluxes and transient jets, the latter often exhibiting periodic latitudinal fluctuations. It is suggested that the above effect of cycling baroclinicity and heat flux induces changes in eddy anisotropy which are responsible for the periodic jet deflections further downstream. On average, low heat flux is associated with an equatorward deflection of the jet, and vice versa. This jetdeflecting effect is characterised by a transfer of the system to a lower-frequency variability, and a mechanism to explain the observed preferred transitions of the North Atlantic jet is proposed. The oscillations in the storm track activity, baroclinicity and eddy-driven jet are closely linked, and can be viewed as describing the spatio-temporal lifecycle of the storm track.

Soundings were made to measure the winds between the surface and 30 kilometre altitude using a lock-follow radar to track a reflector carried aloft by a free balloon. Where possible the descending reflector, or a cloud of radar window was tracked after burst of the balloon. To reduce error in the wind velocity measurements the radar dials were read at frequent intervals and several readings used to give each velocity. The accuracy of the radar was determined and the expected error in velocity shown to be significantly less than the amplitudes of wind fluctuations at ranges of less than 40 kilometres with a 54" corner reflector. At greater raves the error increased rapidly due to poor signal-to-noise ratio. Some statistical analysis of the wind fluctuations as components of a turbulent field was undertaken but is of uncertain significance as no suitable objective criteria could be devised for determining an appropriate mean wind field from the winds measured on a single sounding. Certain more pronounced irregularities on some of the soundings were observed to persist with little change for periods of several hours. In particular, layers of pronounced shear persisted in the troposphere and could be related in many cases to temperature inversions or frontal surfaces shown by temperature soundings at Crawley (50 kilometre distant). Two sequences of sounding made thrown jet streams gave information on the structure beneath the jet, but wind measurements within the core were subject to large errors. Wind fluctuations with amplitudes 1.11) to 7m/sec, horizontal wavelength of order of 300 kilometres, and vertical wavelengths between 1 and 2.5 kilometres were observed in the stratosphere. A model of the wind structure of these fluctuations is proposed which fits the observed features as found in our soundings and in those of the Meteorological Office. Their properties are shown to be consistent in some respects with the properties of inertial waves investigated theoretically by Sawyer (1961). Clouds of window were tracked by the radar and it is shown that differences in fall speed of pieces within the cloud accounted for far more expansion than small scale turbulence. Clouds of a new type of window made from copper wire wore found to provide a useful anemometer in the stratosphere with characteristics similar to those of a balloon. Estimates of the lateral spread of these clouds gave a value of diffusivity, K, of 0.1 m2/sec.