Seasonal changes in the fraction of global radiation retained as net all-wave radiation and their hydrological implications

Abstract
High linear correlation between concurrent measurements of net all-wave radiation Q*, and global incoming short-wave radiation Eg↓ was found for hourly, daytime and 24-h totals measured over a grassland catchment in the subhumid climate region of southeastern Australia. Five years of daily measurements were used to study the temporal variability in the fraction of incoming short-wave radiation retained at the surface as net all-wave radiation. When values of ξ = [Sgrave]Q*/[Sgrave]Eg↓ are calculated from 24-h totals and are plotted against date, the results reveal strong day-to-day and seasonal variability in ξ. Analysis indicated that the ξ ratio was largely determined by data, latitude and cloudiness, whilst the nature of the underlying surface appeared to play a less important role. Data from four other grassland sites in Australia, New Caledonia, Denmark and Ireland yielded a general relationship for estimating ξ from day length. Values of [Sgrave]Q* may thus be obtained from [Sgrave]Eg↓ observations using (a) a single regression between [Sgrave]Q* and [Sgrave]Eg↓ (b) monthly ξ values derived from measurements; or (c) monthly ξ values estimated from day length. All three approaches give estimates of [Sgrave]Q* with error terms similar to those of measurements of net all-wave radiation and those found when [Sgrave]Q* was estimated climatologically from the individual elements of the surface radiation balance—the standard method of obtaining this rarely measured but hydrologically important parameter. The hydrological impact of different methods of obtaining [Sgrave]Q* was demonstrated using the semi-distributed VIC catchment model to compute potential evapotranspiration for the 26 km2 Lockyersleigh catchment near Goulburn, New South Wales, Australia. The limited hydrological impact found is attributed to the fact that actual evapotranspiration in this subhumid catchment is limited by the strength of the soil water source rather than that of the atmospheric sink-potential evapotranspiration.