An autonomous cloud detection algorithm using single ground-based infrared radiometer for the Tibetan Plateau

Abstract. Accurate cloud detection over the Tibetan Plateau (TP) is crucial for understanding regional weather patterns and global climate dynamics. Yet, it remains challenging due to harsh environmental conditions and sparse observations. While ground-based infrared radiometers offer a promising solution through downwelling infrared brightness temperature (IRBT) measurements, existing algorithms require supplementary meteorological data often unavailable in remote TP regions. This study presents a novel cloud detection algorithm that operates solely on IRBT data from a single ground-based infrared radiometer, addressing the critical need for autonomous cloud monitoring in resource-limited environments. The algorithm integrates spectral and temporal analysis approaches: the spectral test identifies cloud presence by comparing observed IRBT against statistically derived clear-sky diurnal cycles, and the temporal test detects clouds through IRBT variability analysis using sliding standard deviation calculations. A key innovation includes a normalization procedure that effectively mitigates dust contamination effects – a persistent challenge in the arid TP environment that can introduce extremely large errors. Validation against 13 months of radiosonde data demonstrates robust performance with agreement rates exceeding 70 % in most months, with particularly effective performance during the wet season. This work provides a practical and cost-effective solution for autonomous cloud monitoring over the TP, with potential for application in other regions with limited observational data.