4546225_monash34652.pdf (43.48 MB)
The classification and simulation of precipitating convective regimes over Darwin, Australia
thesisposted on 2017-01-13, 04:31 authored by Caine, Simon
The regime nature of tropical convection occurring over Darwin Australia is explored in an attempt to condense the large variety of cloud types that occur over a tropical region in to a discrete and manageable number. The regimes are defined by their precipitation structure as observed by a CPOL radar and their links with known features of the tropical atmosphere are investigated. Numerical simulations of tropical convection during the Tropical Warm Pool International Cloud Experiment were made using the Weather Research and Forecasting (WRF) model. The data used to provide the initial and boundary conditions for the model simulations was found to contain a large warm bias in the upper troposphere that detrimentally affected the simulated convection. The simulations were then evaluated against observations from a CPOL radar where it was found that the choice of microphysics scheme had a large impact on the quality of the simulations. One of the microphysics schemes used was found to have a significant problem simulating the precipitation coverage below the freezing level, while the other overestimated graupel coverage and underestimated snow. The precipitation regimes previously defined were used to evaluate the model simulations. It was found that by themselves the regimes were of limited use. The technique used to originally define the regimes was then extended to include model and radar data, which proved to be a more useful (and ob jective) method for evaluating the statistical representation of precipitation in the simulations. It was found that the biggest problem with the WRF simulations was the representation of weak convective time periods. The second biggest problem with the WRF model is its ability to simulate periods of strong deep convection with large coverage of stratiform precipitation, this was attributed to the incorrect forcing data.