Global irradiance measurements made at eleven Australian sites between 1928 and 1992 were analysed for evidence of secular change. Interannual variation in both annual and maximum and minimum monthly values was small and less than the accuracy of the measurements. Limitations of accuracy, together with the non-homogeneous and non-continuous nature of much of the data and the data processing procedures adopted at some periods, suggest that changes in measurement strategy are needed if secular, i.e. long-term, variations of the magnitude measured over much of the northern hemisphere are to be distinguished within a reasonable period.
Stanhill et al. (Wed,) studied this question.