Earth is a cloudy planet—at every moment, 62–72% of the planet is covered by cloud. Its presence and properties affect the flow of energy between the sun, Earth and the atmosphere. More comprehensive information about cloudiness is important to improve our understanding of the impact of cloud on Earth’s radiation budget. The aim of our work is to assess the accuracy of groundbased high-level cloud detections over Poland compared with satellite data. This project is the very first evaluation of error in visual high-cloud observation. Specifically, we answer a fundamental question: how effective are ground-based observers in detecting high-level cloud?Our unprecedented analysis is based on two datasets: traditional, visual observations from 43 selected synoptic stations and satellite (lidar) data, for the years 2006–2017. Ground-based observations were obtained from the archives of the Polish National Weather Service (IMWM-NRI), while satellite data came from the CALIPSO mission. Our final dataset comprised 7548 ground-based and satellite observations. Accuracy was assessed using the following indicators: probability of detection (Pd), probability of false alarm (Pfa), and overall accuracy (ACC). Overall (day and night observations), in the absence of lower-level cloud, observers correctly detected 80% of high-level cloud. However, when lower-level cloud was present, observers could not detect around 70% of high-level cloud. In sum, ground-based, high-level cloud detection is a function of the presence of lower-level cloud, the time of day, and the size of the area over which cirrus are observed/detected.
Nguyen Huu, Ż., Kotarba, A.Z. (2021). Reliability of visual detections of cirrus over Poland. Theoretical and Applied Climatology