An eddy-resolving upper ocean model is developed to study the dynamics and thermo-dynamics of Baie des Chaleurs (BdC, 47.5-48.5N, 65.5-66.5W), Gulf of St. Lawrence (GSL), Canada. The model has primitive equation dynamics with two active layers embedded with a Kraus-Niiler type mixed layer model at the top. / Forced by observed wind, atmospheric heat fluxes, river runoff and appropriate remote forcing (in particular, the Gaspe Current, GC), the model demonstrates that the mean cyclonic general circulation pattern in the bay is a consequence of the intrusion of the GC. In the mixed layer, atmospheric heat fluxes and horizontal thermal advection play a key role in the thermal balance at the eastern part of the bay. The local mixed layer fluctuations are controlled by wind and GC induced divergence. The entrainment (and its corresponding heat flux) is important at the western part of the bay and changes the mean mixed layer depth on a time scale of more than a week. Varying GC intensified the flow variations induced by the wind in the bay and improved simulation results as compared with observations. / Sensitivity runs are conducted to study the effects of external forcing, important physical processes and the internal physical parameterisation on the model results and to compare these with the main model run. Experiments show that nonlinearity is very important in determining the circulation pattern in the bay. Changing external thermal forcing also modifies dynamical processes in the BdC. The fluctuations in the near surface temperature are mainly due to latent and sensible heat changes. The parameter study indicates that, the model is not overly sensitive to changes in most of the parameters, but suggests that sensitivity of the mixed layer physical parameters depends on the dynamical and thermodynamic system applied. / Hydrographic and current meter data are used first to study the variability of both the dynamics and thermodynamics in the BdC and its relation to the separation/intrusion of the unsteady GC. A numerical model is then applied to gain insight into the problem. The time scales of interest range from tidal to seasonal. / The results show that the kinetic energy in the BdC is dominated by the semi-diurnal tide (M$ sb2$) and periods of 5-10 days for high and low frequency bands, respectively. Most of the energy in the low frequency band is found to be induced by wind-related forcing. / Both observations and model results indicate that seasonal variations in the BdC are strongly related to the characteristics of separation/intrusion of the GC, which is mainly controlled by its transport magnitude as well as phase, duration and strength of its acceleration (or deceleration). The separation occurs when (adverse) vorticity having an opposite sign from that existing upstream is generated near the separation area. Although the separation can be generated in a decelerating GC, it can also occur in an accelerating GC when the GC is strong enough to advect upstream vorticity necessary to form a recirculation and the related adverse vorticity downstream. Nonlinearity is critical to the separation. Nevertheless, separation can be generated in a linear current with strong deceleration. The GC intrudes either along the coastline (attachment) into the bay by a non-separated GC or following the separation of the GC (reattachment). Effects of various physical processes on the separation/intrusion and variability of eddies in the BdC are examined.

The plane parallel homogeneous (PPH) bias is defined as the difference between the plane parallel cloud albedo, calculated for homogeneous cloud optical depth distributions, and the independent pixel (IP) albedo, which allows for optical depth variability, but assumes that individual cloudy columns transfer solar radiation as plane parallel slabs (horizontal photon transport is neglected). Estimates of the PPH bias from extensive Advanced Very High Resolution Radiometer (AVHRR) observations are provided for areas similar in size to weather and climate model gridboxes. The goal is to investigate conditions and assumptions influencing the PPH bias, and to suggest methods to correct for it. / Visible PPH biases vary from about 0.02 to 0.30, depending on area size, view/sun geometry, and other factors influencing optical depth retrievals and albedo calculations. Broadband PPH biases are slightly smaller than visible biases; broadband absorptance biases are about an order of magnitude smaller. Approximate estimates of the bias in broadband reflected flux at cloud top often exceed 30 Wm$ sp{-2}$ for near-nadir measurements, suggesting that the assumption of cloud homogeneity produces errors that cannot be ignored in climate studies. / Solar geometry affects the PPH bias not only through the direct dependence of albedo on solar zenith angle, but also through systematic changes in the apparent mean and variance of optical depth arising from the neglect of 3-d effects in satellite radiance inversions. PPH biases decrease with data resolution, increase when atmospheric radiative effects are accounted for in optical depth retrievals, and are only slightly affected by water cloud microphysics. PPH biases are also shown to differ substantially between the forward and backward scattering directions, and between large-scale models with and without provision for fractional cloudiness. / A large fraction of the PPH bias is removable by: (1) adjusting regionally averaged optical depths with the reduction factor of Cahalan et al. (1994a), and (2) fitting observed optical depth distributions with lognormal and (to a lesser degree) gamma distributions. These methods require the logarithmic mean and variance of optical depth, which are parameterized as a function of regional mean optical depth and cloud fraction, quantities routinely available in climate models.

A study of feasibility is performed for the development of a nowcasting method of aircraft icing conditions. The supercooled cloud water content is a key parameter for the determination of hazardous regions. The production of supercooled water in clouds is possible only where sufficiently strong updraft produces supersaturation with respect to water. However, to maintain the presence of supercooled droplets, the generation of moisture excess during the air uplift must exceed the rate of vapor deposition on snow and the rate of snow riming. The "storm of the century" (March 13-14, 1993) has been chosen as a good case of stratiform precipitation event and its microphysics was simulated using a high resolution three-dimensional kinematic cloud model with a full microphysical scheme. The winds driving the development of hydrometeors are obtained by a three-dimensional wind retrieval technique applied to single Doppler radar data. The output of the cloud model provides the field of supercooled (since T $<$ 0$ sp circ$C) cloud water. In this way a potentially operational method for detection of icing conditions is suggested. / Once implemented, the method was used on a second case (October 21, 1993), where aircraft observations provided a mean of evaluating its performance. The comparison of retrieved cloud liquid water with in situ measurements, shows a good potential for the retrieval method.

An atomic oxygen target suitable for absolute differential cross sections has been developed. Molecular oxygen gas is partially dissociated in a microwave discharge and flowed through a collision cell to serve as a fast beam target. Determination of the atomic oxygen density is carried out using an electron impact time-of-flight mass spectrometer, which uses the exit aperture of the cell as a sample leak.

The problem of creating synoptic maps from asynoptically gathered data has prompted the development of a number of schemes. Most notable among these schemes are the Kalman filter, the Salby-Fourier technique, and constituent reconstruction. This thesis presents a new technique, called trajectory mapping. Trajectory mapping employs a simple model of air parcel motion to create synoptic maps from asynoptically gathered data. To assess the applicability of the technique, four sources of trajectory mapping errors were analyzed. The analysis revealed that (1) the computational error is negligible; (2) measurement uncertainties can result in errors which grow with time scales on the order of a week; (3) isentropic approximations lead to errors characterized by time scales of a week; and (4) wind field inaccuracies can cause significant errors in individual parcel trajectories in a matter of hours. All the studies, however, indicated that while individual trajectory errors can grow rapidly, constituent distributions, such as those depicted in trajectory maps, are much more robust, maintaining a high level of accuracy for periods on the order of several weeks. The trajectory mapping technique has been successfully applied to a variety of problems. First, trajectory mapping was employed in the study of dynamical wave-breaking events. Second, trajectory mapping was applied in satellite data validation studies, both for the determination of instrument accuracy and precision. Third, trajectory mapping was used to assess the accuracy of the meteorological wind fields. Such demonstrations imply that trajectory mapping will become an important tool in answering questions of global change, particularly the issue of ozone depletion.

Because of the environmental importance, ozone has long been one of the prominent research topics in the scientific community. Surprisingly, in spite of all the research, the UV photodissociation dynamics of ozone has not been completely understood yet. More specifically, the origin of the small structural features overlapping the broad-band feature in the UV absorption spectrum of ozone still remains a mystery. However, theoretical calculations done in our lab predict, subject to UV radiation, great majority of ozone molecules fall apart in roughly 6 femtoseconds while a very minute portion, about 1% corresponding to small structural features take much longer to dissociate (up to 150 femtoseconds). Even with the fastest lasers currently available, it would not be possible to learn about the photodissociation dynamics of ozone using Femtosecond chemistry. However, a unique technique called continuous-scan Raman Excitation Profiles developed and tested on iodobenzene in our lab has proven to be a powerful method in learning about extremely fast dissociating processes such as ozone. One of the features that make REP's a powerful tool is that, it provides valuable knowledge of time domain behavior based on only the intensity and frequency of the emitted photons. However, primarily because of its extremely low photon yield, one must overcome several significant and challenging experimental problems associated with the apparatus before achieving any reliable ozone REPs. The main objective of this thesis is to demonstrate how these problems were solved and to discuss and analyze some recently obtained preliminary data to be the guide for those pursuing full-scale ozone REP'S in the future.

Ground-level ozone is of a growing concern in many areas of the United States. Ozone is a significant health concern, particularly for people with asthma and other respiratory diseases. Ozone is rarely emitted directly into the air but is formed by the reaction of volatile organic compounds (VOCs) and nitrogen oxides (NOx) in the presence of sunlight. VOCs are emitted from a variety of sources, including motor vehicles, chemical plants, refineries, factories, consumer and commercial products, other industrial sources, and biogenic sources. NOx is emitted from motor vehicles, power plants, and other combustion sources. Ozone and ozone precursors also can be transported into an area from pollution sources found hundreds of miles away. In accordance with the 1990 Clean Air Act Amendments, EPA has required more extensive monitoring of ozone and its precursors in areas with persistently high ozone levels. In these areas, the States have established ambient air monitoring networks consisting of CAMS (continuous air monitoring system) sites, which collect and report detailed data for volatile organic compounds, nitrogen oxides, ozone and meteorological parameters. Analyses of these data help the regulatory agencies to better understand the underlying causes of ozone pollution, to devise effective remedies and to measure air quality trends. This thesis focuses on how to integrate these measurements of VOCs with the receptor modeling techniques in order to identify the sources of VOCs and to attribute ambient VOC concentrations to their original sources. The measurements taken from three CAMS stations in Houston, TX serve as the basis of this research. After presenting the source attribution of volatile organic compounds, where the contribution from different sources to ambient VOC levels are determined, the methods to identify the source regions associated with elevated VOC levels are described. The quantitatively reconstructed emissions from a recently prepared VOC emissions inventory are compared with the receptor model calculations of ambient VOC measurements. Finally, a separate growing concern in the US, the particulate matter pollution, is addressed. The impacts of regional wild fires in Texas on the secondary particulate matter formation are examined. The results of the laboratory investigations on the formation of the secondary sulfate particles through heterogeneous surface reactions are presented.

The climate system exhibits behavior on a wide range of time and spatial scales. Computer models that simulate climate must be faithful to these matters of scale in order to achieve meaningful results. This study involves the coupling of two models: a model that simulates the seasonal fluctuations of the atmosphere and ocean with a time constant on the order of one day and a continental ice sheet model designed to simulate the behavior of ice sheets over tens of thousands of years. These models are asynchronously coupled in both time and space because computational requirements limit the minimum grid size and time step size. Experiments are presented which examine the long time constant of the continental ice sheet and the fast approach to quasi-equilibrium of the seasonal model. Simulations of present day conditions from the coupled model are compared with climate observations and simulations of other climate models. These studies also show the sensitivity of model simulations of past climate to albedo changes, differing ice flow parameters, aerosol dust, lapse rate changes, and changes in carbon dioxide. The coupled model is then used to simulate a 120,000 year ice age cycle and mechanisms important over long time scales are discussed.

An idealized box model of the Indian Ocean is used to examine the hypothesis proposed by Godfrey and Weaver (1991) that the buoyancy-forced Leeuwin Current off the west coast of Australia is a manifestation of a basinwide thermohaline circulation driven by the Indonesian throughflow. / The stronger Sverdrup circulation dominates the thermohaline circulation in most of the model ocean except near the eastern boundary. / The western boundary currents apparently play a very minor role in this basinwide thermohaline circulation. This differs from the warm water route proposed by Gordon (1986), and supports the alternative hypothesis that the heat from the equatorial Pacific is returned to the South Atlantic via the eastward-flowing Antarctic Circumpolar Current rather than past the Agulhas Retroflection. / The Indonesian throughflow is shown to significantly affect the surface heat fluxes and the meridional heat transport in the Indian Ocean. The role of the throughflow in maintaining the very warm climate of the Indian Ocean (a net exporter of heat) is described. / Large-scale, fairly long period ($>$100 days) barotropic eddies are found in the western portion of the basin for some solutions.

A review of studies tackling problems on marine stratiform clouds is presented. The strong effect of marine stratocumulus clouds on the top-of-the-atmosphere radiation budget is demonstrated using 5 years of Earth Radiation Budget Experiment (ERBE) data. A relationship between albedo and sea surface temperature (SST) is shown to exist in two areas of the globe that tend to be covered by marine stratocumulus. Albedo increases when SST decreases and vice-versa in both regions when examined on an annual, interannual or spatial basis. The magnitude of the albedo response to a given SST change (1) varies within the regions; (2) differs between the two regions; (3) depends on the type of variability examined; (4) depends on the SST. No useful relations were found between albedo and other meteorological variables. Climatic implications on a global scale arising from the albedo-SST anticorrelation are also discussed.