Evaluation of the CMIP6 planetary albedo climatology using satellite observations

• Published in: Climate dynamics Vol. 54; no. 11-12; pp. 5145 - 5161
• Main Authors: Jian, Bida, Li, Jiming, Zhao, Yuxin, He, Yongli, Wang, Jing, Huang, Jianping
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.06.2020; Springer; Springer Nature B.V
• Subjects: Albedo; Analysis; Annual cycles; Annual variations; Arid lands; Arid regions; Atmospheric models; Climatology; Cloud albedo; Clouds; Computer simulation; Earth and Environmental Science; Earth Sciences; Electric waves; Electromagnetic radiation; Electromagnetic waves; Energy balance; Geophysics/Geodesy; Global radiation; Intercomparison; Ocean models; Oceanography; Oceans; Radiation; Radiation budget; Satellite observation; Seasonal variation; Statistical analysis; Stratocumulus clouds; Tropical climate; China; Satellite observations; CMIP6; Planetary albedo
• ISSN: 0930-7575; 1432-0894
• DOI: 10.1007/s00382-020-05277-4

The Earth’s planetary albedo (PA) has an essential impact on the global radiation budget. Based on 14 years of monthly data from the Clouds and the Earth’s Radiant Energy System energy balanced and filled (CERES-EBAF) Ed4.1 dataset and atmosphere-only simulations of the Coupled Model Intercomparison Project Phase6 (CMIP6/AMIP), this study investigates the ability of CMIP6/AMIP model in reproducing the observed inter-month changes, annual cycle and trend of PA at near-global and regional scales. Statistical results indicate that some persistent biases in the previous models continue to exist in the CMIP6 models; however, some progresses have been made. In CMIP6/AMIP, large negative correlations for PA between the model ensemble mean and observation are addressed over the subtropical stratocumulus regions. In addition, the simulation of PA in drylands and tropical oceans remains a challenge in CMIP6 models. Over the most regions, PA biases are governed by cloud albedo forcing biases. These results demonstrate the importance of improving cloud process simulations for accurately representing the PA in models. For the annual cycles, the model ensemble mean captures the difference in amplitude between the two peak values of PA (June and December), as well as the phase of the seasonal cycle, despite PA is systematically overestimated. The differences between different terrestrial climatic regions are also examined. Results indicate that the relative biases of PA are greatest in semi-arid (2.2%) and semi-humid (2.8%) regions, whereas the minimum relative bias occurs in arid regions (0.3%) due to compensating errors.