A long-term monthly dataset of cloud fraction over the Arctic based on multiple satellite products using cumulative distribution function matching and Bayesian maximum entropy

The low accuracy of satellite Cloud fraction (CF) over the Arctic seriously restricts accurate assessment of regional and global radiant energy balance under the changing climate. Previous studies have reported that not a single satellite CF product could satisfy the needs of accuracy and spatio-temporal coverage simultaneously for long-term applications over the Arctic. Merging multiple CF products with complementary properties is an effective way to produce more spatiotemporally complete and accurate CF data record. This study proposed a spatiotemporal statistical data fusion framework based on cumulative distribution function (CDF) matching and Bayesian maximum entropy (BME) method to produce a syncretic 1°×1° CF dataset in the Arctic during 2000-2020. The original datasets contain CF from MOD08/MYD08, CERES-SSF Terra/Aqua, CLARA-A2 AM/PM, PATMOS-x AM/PM, ISCCP-H AM/PM. The fused CF product is more consistent with the active satellite data GEWEX-CALIPSO and the ground-based observation data CRU TS4.05.北极地区卫星云量（Cloud Fraction, CF）产品精度偏低，严重制约了气候变化背景下区域与全球辐射能量平衡的精准评估。既往研究表明，目前尚无任何一款卫星CF产品能够同时满足北极地区长期应用所需的精度与时空覆盖需求。融合多组具备互补特性的CF产品，是生成时空完整性与精度更优的CF数据记录的有效途径。本研究提出了一种基于累积分布函数（Cumulative Distribution Function, CDF）匹配与贝叶斯最大熵（Bayesian Maximum Entropy, BME）方法的时空统计数据融合框架，用于生成2000-2020年北极地区1°×1°分辨率的融合CF数据集。原始数据集涵盖了来自MOD08/MYD08、CERES-SSF Terra/Aqua、CLARA-A2 AM/PM、PATMOS-x AM/PM以及ISCCP-H AM/PM的CF数据。该融合CF产品与主动卫星数据GEWEX-CALIPSO以及地面观测数据CRU TS4.05的一致性更佳。