Quantitative Assessment of Cloud Cover and Sky Brightness Based on All-Sky Camera Photometry

Abstract

All-sky cameras are the primary ground-based instruments used to monitor cloud cover for astronomical site testing. However, a quantitative, automated, and generalised method for cloud cover assessment using all-sky camera images has not yet been established. Photometric analysis of the images can address this issue by using relative extinction as a proxy for clouds, but the super-wide field of view and associated optical distortions present significant challenges for accurate photometry. In this paper, we present a new generalised photometric method for quantitatively estimating the spatial distribution of extinction, cloud cover, and sky background from any all-sky camera image. As an example, we demonstrate this method on images acquired by an all-sky camera operated at Dome A, Antarctica, and compare the results with those from the method of semi-quantitative visual inspection on the same data set. We reach a typical photometric accuracy of 0.1 mag for 4.5 mag stars, which is sufficient for reliable cloud detection. Through this method, we obtain extinction and sky background measurements in the Canon G band and the Tycho V band, respectively. Applying these measurements yields quantitative site statistics: a clear-sky rate of 72%, a median sky background of 21.09 mag, no aurorae within 65° of the zenith for 48% of the time, and widespread aurorae 9% of the time. Comparisons with previous studies confirm the robustness of our method, and reveal that visual inspection would overestimate the clear-sky rate because it is insensitive to thin clouds with extinction <0.45 mag.