Thesis (M.S.) - U.S. Naval Postgraduate School, 1966. / Bibliography: leaves 57-[58].

This dissertation presents a series of work related to the representation of the Hadley circulation (HC) in atmospheric reanalyses and general circulation models (GCMs), with connections to the underlying tropical and subtropical cloud systems that comprise the mean meridional circulation. An intercomparison of eight atmospheric reanalyses showed that significant variability exists in the mean state for HC intensity, with less variability in HC width. Ensemble trends were broadly consistent with previous work and suggest a strengthening and widening of the tropical circulation over the last 30 years. Composite profiles of the apparent heat source and moisture sink were calculated for the International Satellite Cloud Climatology Project (ISCCP) cloud regimes using sounding observations from 10 field campaigns. Distinct heating profiles were determined for each ISCCP cloud regime, ranging from strong, upper-tropospheric heating for mesoscale convective systems to integrated cooling for populations associated with marine stratus and stratocumulus clouds. The derived profiles were generally similar over land and ocean with the notable exception of the fair-weather cumulus regime, which leads to some uncertainty in the mid- and upper-level reconstruction of subtropical heating. An instrument simulator indicated that low-latitude cloud properties from the NASA MERRA reanalysis qualitatively matched the distributions of cloud-top pressure and optical thickness in the ISCCP data, though the tallest and thickest clouds were missing from the reanalysis. Simulator results were sensitive to the choice of cloud overlap parameterization and the reanalysis consistently underpredicted the observed cloud fractions for all regimes. The vertical velocity, temperature, and moisture for each regime in MERRA largely matched observations from previous studies, suggesting that the dynamic and thermodynamic properties of the cloud regimes are well captured by the reanalysis. Finally, HC interannual variability was examined as a function of the observed frequency of the ISCCP cloud regimes. The strongest HC overturning events were attributed to an El Niño response in the central Pacific Ocean in addition to links between the intensity and position of the Pacific ITCZ. The ISCCP regime describing the most vigorous and organized convection contributed the most towards the total anomalous heating during HC extremes, despite an overall low frequency of occurrence. Idealized GCM simulations forced with the observed three-dimensional diabatic heating from ISCCP data produced too strong a HC with some improvement in other fields. Overall, much progress has been made regarding the links between low-latitude cloud systems and the HC, though future work will continue to address the upscale feedbacks of regional cloud variations upon the tropical circulation.

The thesis comprises two separate journal articles that together form a coherent body of work. In this thesis, the key physical processes responsible for the 23-26 September 2012 UK floods are investigated using a case study approach. The cyclone responsible for the floods developed near the Azores on 20¬-22 September following the interaction between an equatorward-moving potential vorticity (PV) streamer and tropical storm Nadine. Convectively-driven latent heat release associated with the developing cyclone reduced upper-level PV and resulted in the fracture of the PV streamer into a discrete anomaly as the cyclone intensified. In Paper 1, convection-permitting model simulations and diabatic heating rate and PV tendency calculations along trajectories demonstrate that deposition heating strongly reduced upper-level PV in the vicinity of the PV streamer, contributing to its fracture into a discrete anomaly. The cyclone deepened further over the UK on 23-26 September, ahead of a second upper-level PV anomaly. In Paper 2, sensitivity simulations of the storm are presented. PV inversion is used to modify the strength and position of the PV anomaly in the initial conditions and to examine whether the event could have been even more extreme with different upper-level forcing. Results show that quasigeostrophic forcing for ascent ahead of the PV anomaly contributed to the maintenance of the rainfall band over the UK. Counterintuitively however, strengthening the upper-level forcing produced a shallower cyclone with lower rainfall totals. Instead of moving eastward over the UK to interact with the cyclone, the strengthened anomaly rotated cyclonically around a large-scale trough over Iceland, resulting in a fragmented rainfall band. The counterintuitive results suggest that the verifying analysis represents almost the highest-impact scenario possible for this flooding event.

The vertical structure of the tangential wind field in tropical cyclones is investigated through observations, theory, and numerical simulations. First, a dataset of Doppler radar wind swaths obtained from NOAA/AOML/HRD is used to create azimuthal mean tangential wind fields for 7 storms on 17 different days. Three conventional wisdoms of vertical structure are reexamined: the outward slope of the Radius of Maximum Winds (RMW) decreases with increasing intensity, the slope increases with the size of the RMW, and the RMW is a surface of constant absolute angular momentum (M). The slopes of the RMW and of M surfaces are objectively determined. The slopes are found to increase linearly with the size of the low-level RMW, and to be independent of the intensity of the storm. While the RMW is approximately an M surface, M systematically decreases with height along the RMW. The steady-state analytical theory of Emanuel (1986) is shown to make specific predictions regarding the vertical structure of tropical cyclones. It is found that in this model, the slope of the RMW is a linear function of its size and is independent of intensity, and that the RMW is almost exactly an M surface. A simple time-dependent model which is governed by the same assumptions as the analytical theory yields the same results. Idealized hurricane simulations are conducted using the Weather Research and Forecasting (WRF) model. The assumptions of Emanuel's theory, slantwise moist neutrality and thermal wind balance, are both found to be violated. Nevertheless, the vertical structure of the wind field itself is generally well predicted by the theory. The percentage rate at which the winds decay with height is found to be nearly independent of both size and intensity, in agreement with observations and theory. Deviations from this decay profile are shown to be due to gradient wind imbalance. The slope of the RMW increases linearly with its size, but is systematically too large compared to observations. Also in contrast to observations, M generally increases with height along the RMW.