LAUSR

We used parallel sunshine duration datasets obtained with a Jordan sunshine recorder and three automatic sunshine duration sensors to investigate the differences between these instruments. We used measurements obtained at Shangdianzi (SDZ) regional Global Atmosphere Watch (GAW) station with a Jordan sunshine recorder, a DFC2 photoelectric sunshine meter, a CHP1 pyrheliometer, and two CMP11 pyranometers from 1 January to 5 July 2019 and from 3 November 2020 to 28 February 2021. The results showed that the daily sunshine duration measurements obtained from the Jordan sunshine recorder were comparable with those from the DFC2 meter and the CMP11 pyranometers under all-sky conditions, but were considerably different from those observed by the CHP1 pyrheliometer. An analysis of potential influencing factors showed that the solar zenith angle, the spectral range of the automatic sensors, the relative humidity, and the sky conditions were the main factors affecting the measurements of sunshine duration between the Jordan sunshine recorder and three automatic sensors. We proposed a simple linear regression function—the DFC2-equivalent sunshine duration estimation (DFCESD) model—to guarantee the consistency of the long-term sunshine duration series observed by the Jordan sunshine recorder at SDZ and the measurements from the DFC2 meter. Validation of the DFCESD model showed that the mean absolute difference (MAD) between the daily sunshine duration observed by the Jordan sunshine recorder and those from the DFC2 meter improved from −0.7 to −0.2 h day −1 , the relative deviation (RD) improved from −9.3% to −2.3%, and the root-mean-square deviation (RMSD) decreased from 1.0 to 0.8 h day −1 .