Identification of clear skies from broadband pyranometer measurements and calculation of downwelling shortwave cloud effects

Abstract

We present an automated method to identify periods of clear skies for a 160° field of view using only 1‐min measurements of surface downwelling total and diffuse shortwave irradiance. The clear‐sky detection method is verified using Whole Sky Imager and lidar data, observer reports, and model comparisons. Identified clear‐sky irradiance measurements are then used to empirically fit clear‐sky irradiance functions using the cosine of the solar zenith angle as the independent variable. These fitted functions produce continuous estimates of clear‐sky total, diffuse, and direct component shortwave irradiances. While this method ignores diurnal changes in such variables as column water vapor and aerosol amounts and changes between clear‐sky days, it is shown that the resultant clear‐sky irradiance estimates have RMS uncertainty comparable to the uncertainty of the measuring instruments themselves. The estimated clear‐sky irradiances are used to estimate the effect of clouds on the downwelling shortwave irradiance as a difference between the measured and clear‐sky amounts. We show that the cloud effect calculations from this method appear to decrease the uncertainty due to systematic pyranometer offsets and cosine response errors. Thus any data set that includes downwelling diffuse and total shortwave measurements can be processed to identify clear‐sky periods and produce estimates of clear‐sky irradiance and cloud effects.