Comparison of Cloud Amounts Retrieved with Three Automatic Methods and Visual Observations

• Published in: Atmosphere 2022-06, Vol.13 (6), p.937
• Main Authors: Utrillas, María Pilar, Marín, María José, Estellés, Víctor, Marcos, Carlos, Freile, María Dolores, Gómez-Amo, José Luis, Martínez-Lozano, José Antonio
• Format: Article
• Language: English
• Subjects: Algorithms; automatic cloud observations; Clear sky; Cloud amount; cloud cover; Clouds; Extreme values; Low clouds; Methods; Pyrgeometers; Radiation; Remote sensing; visual cloud observations; Visual observation
• ISSN: 2073-4433; 2073-4433
• DOI: 10.3390/atmos13060937

Four methods have been used for the estimation of the total cloud amount and cloud amount for low clouds: visual observations, the Long method applied on pyranometer measurements, the Automatic Partial Cloud Amount Detection Algorithm (APCADA) method applied on pyrgeometers measurements, and ceilometer measurements of the cloud base height. Records from meteorological observers indicate that clear days (0–1 octa) represent the most frequent cloud amount for low clouds. In contrast, the total cloud amount is more aleatory. Results obtained from the Long method show maximum frequency in the extreme cloud amount values. The APCADA method also indicates the predominance of cloudless skies. The ceilometer method shows a predominance of completely clear skies, but the completely cloudy (8 octas) is the second most frequent case. Automatic methods report more cloudless and overcast skies than the observer. Automatic methods agree with the visual method or differ in ±1 octa for 60–76% cases for low cloud amount and for 56–63% cases for total cloud amount. In general, low cloud amount agrees more with observer measurements than total cloud amount and the automatic methods underestimated total cloud amount observer values possibly due to the difficulty in monitoring high clouds.