I do not know about Australia in particular, but most countries with weather data also have insolation data. Your Dept of Environment would be the place to start. Data is usually collected hourly as radiation on a horizontal surface by pyrheliometers or pyranometers at many weather stations. However, this data, in order to be useful to the solar engineer, must be altered (in conjunction with approximately the past 30 years worth of data) - changed statistically to give one set of applicable data - whether representative of hourly, diurnal, weekly, monthly, or annual insolation. Then, of course you must be able to work this into one or more of the currently accepted "sky models" (such as Perez et al). And, if there is glazing involved, you must be able to account for losses through different tilt angles, orientations, glazing types, low-e coating, and gas-fill configurations. I would recommend "Solar Engineering of Thermal Processes" by John Duffie and William Beckman (John Wiley & Sons, 1991. ISBN 0-471-51056-4)