Download files
Abstract
The fundamental importance of the water cycle motivates our need for accurate forecasting and climate projections of regional precipitation. Predicting regional precipitation is a notoriously complicated task and improvements are necessary to better understand precipitation variability. In this thesis the interactions between large-scale atmospheric variability and precipitation are investigated in both observations and climate models. The influence of intraseasonal large-scale variability on observed Australian precipitation is investigated and the skill of current generation models to reproduce the observed relationships is assessed. Furthermore, the influence of parameterised convection on intraseasonal variability is examined in a simplified model.
An autonomous strategy is used to determine the dominant sources of interannual large-scale variability in Australia. Results provide observational evidence that natural variability in the tropical edge influences Australian precipitation where a poleward tropical edge reduces precipitation in southeastern and southwestern Australia. Current generation models successfully capture the observed drying however, their skill scores reflect that the interactions between large-scale variability and precipitation are not well reproduced.
Intraseasonal variability is poorly represented in global atmospheric models and in this thesis three mechanisms reported to improve intraseasonal variability are independently investigated in an aquaplanet model. These mechanisms include increasing the convection schemes sensitivity to moisture, inhibiting convection to allow instability to develop and cloud radiative effects. Inhibited convection and moisture-sensitivity are achieved by changes to the convection scheme parameters and cloud radiative effects are introduced into the aquaplanet.
The largest increase in intraseasonal variability occurs for inhibited temperature-sensitive convection, and for cloud radiative effects with the largest temperature gradient between cloudy and clear sky. MJO-like variability with a similar structure to observations is produced in the model however the large-scale circulation is shifted equatorward compared to the control without cloud radiative heating.
This thesis uses a new approach to identify precipitation drivers that provides a robust way of identifying model shortcomings. Furthermore, more realistic intraseasonal variability is achieved in an aquaplanet with improvements in the power and propagation of MJO-like waves. This thesis assists the weather and climate communities by providing new insight into model shortcomings and further exploring mechanisms which improve intraseasonal variability.