A Method for Clear-Sky Identification and Long-Term Trends Assessment Using Daily Surface Solar Radiation Records

Abstract

Time series of clear-sky irradiance are fundamental for the understanding of changes in the Earth Radiation budget, since they allow to examine radiative processes in the cloud-free atmosphere. Clear-sky data is usually derived from all-sky irradiances using one of several clear-sky methods proposed in the literature. However, most of the available clear-sky methods require additional in situ measurements and/or high temporal resolution (sub-daily), which restricts the derivation of clear-sky time series to a few well equipped stations. Here we propose a new clear-sky identification method that aims to overcome this problem, with the ultimate goal of deriving multidecadal clear-sky trends for many sites globally. The method uses site specific monthly transmittance thresholds to derive long term clear-sky time series for any station worldwide that has daily mean irradiance data. We exemplify the method for 24 stations. Transmittance thresholds are derived by combining 29 years (1990–2018) of satellite cloud cover data with in situ irradiance measurements. The thresholds are then applied to the whole time series (independent of satellite data availability) to screen out cloudy days. Comparison of our results with reference data derived using Long and Ackerman's (2000, https://doi.org/10.1029/2000jd900077) method shows good agreement after bias correction, especially for decadal trends. While limitations of the method, such as anomalies representation, are highlighted and discussed, validation results encourage its use to derive long term clear-sky time series and associated decadal-scale trends around the globe.