The emergence of new-generation geostationary satellites (such as Himawari-8 or FenYun-4) provides us an opportunity to obtain more accurate global horizontal irradiance (GHI) data. In this study, instantaneous GHI estimates (with a spatio-temporal resolution of 10 min and 5 km) are produced using the cloud products from the new-generation geostationary satellite Himawari-8 with a physically based algorithm. The hourly, daily, and monthly GHI estimates are aggregated using "snapshots" (or instantaneous estimates) at two different sampling frequencies (i.e., 1 h and 10 min), and validated against surface radiometry measurements collected in China. The root mean square errors (RMSE) for hourly, daily and monthly GHI estimates calculated using 1-h snapshots were 106.6, 27.9, and 17.7 W m(-2), respectively, whereas those for hourly, daily and, monthly GHI calculated using 10-min snapshots decreased to 94.0, 24.0, and 16.6 W m(-2), respectively. This result demonstrates that the accuracies of hourly, daily, and monthly GHI estimates are improved by increasing the frequency of satellite observations from 1 h (frequency of the previous-generation geostationary satellites) to 10 min (frequency of the new-generation geostationary satellites).