Terrestrial Cloud Fraction derived from GOES-R Satellite

Cloud cover strongly influences the amount of solar radiation reaching the Earth's surface. Space-based and ground-based instruments have different fields of view and rely on different observation methods, therefore comparison of the observed quantities from each instrument can be challenging. Even so we need to establish the accuracy of satellite-based cloud cover estimates in order to utilize this data within our suite of meteorological observing and forecasting applications.

GOES-R data from 11.2 µm, 12.3 µm, 3.9 µm, 0.64 µm, 0.47 µm and 1.6 µm are used for the cloud fraction detection, which is compared with hourly METAR data quantitatively and qualitatively.

The study is done for hourly cloud fraction for overestimation and underestimation of Satellite data compare with METAR during Day and Night period for 500 stations in Canada and US. We found overestimation and underestimation of cloud fraction occurred in less than 3% of cases (day or night ). Overall error statistics for the satellite estimations were around -2% bias and 20% mean absolute error. Qualitatively, day period cloud fraction is more accurate than night period.