Parametrization of net radiation at the surface using data from the Wangara experiment

Hourly Wangara boundary layer data (Clark et al., 1971) was used to substantiate previously derived empirical equations of the surface radiation budget. These equations only require measurements taken by a surface observer in order to be evaluated. A parameterization was made for infrared (IR) radiation. From this procedure, a residual radiation was obtained from the measured daytime net radiation data in order to approximate short wave (SW) radiation. Diurnal trends in radiation and other related meteorological parameters were also evaluated. For IR radiation, the Brunt equation was found to adequately represent effective radiation without clouds. IR attenuation (K-values) of 0.89, 0.64, and 0.51 were calculated for low, middle, and high clouds, respectively. A modification of Beer's Law was derived to evaluate the residual radiation under clear skies as suggested by Flowers et al. (1969) for SW radiation. With the surface albedo assumed at 0.15, a mean turbidity coefficient of 0.16 was determined. Also, the square root of the optical air mass was found to best represent the data in this equation. An analysis of SW cloud attenuation coefficients (C-values) from the residual radiation proved inconclusive. Of considerable importance is the finding that cloud attenuation of the residual radiation behaved nonlinearly. Whereas solutions as high as the fourth order fit the data at noon, quadratic solutions fit well near sunset and sunrise. Overall; relative deviations during the day of the combined parameterization were approximately 20% at noon and just less than 30% for periods up to one hour before sunset or sunrise. Predictions within these one-hour periods, however, contained considerable error, partially due, to the scatter and small magnitude of the measurements. Diurnal trends in surface radiation and other meteorological parameters were established