Received: 26 Mar 2020 – Discussion started: 16 Jun 2020
Abstract. This paper aims to find a possible ensemble method to combine the global climate models, providing an accuracy forecast of sea ice thickness. Conventional multimodel superensemble, the advanced method that is widely used in atmosphere, ocean and other fields, cannot be well performed in sea ice thickness simulation. Hence, an adaptive forecasting through exponential re-weighting (AFTER) algorithm is adopted to improve the conventional multimodel superensemble. Results show our proposed methods perform better than any other mainstream ensemble methods by using a multi-criteria evaluation. The proposed method is used to predict the future sea ice thickness in the period of 2020–2049, where the possible biases are discussed.
How to cite. Yangjun, W., Kefeng, L., Ren, Z., Longxia, Q., and Yu, Z.: Improved Multimodel Superensemble Forecast for Sea Ice Thickness using Global Climate Models, The Cryosphere Discuss. [preprint], https://doi.org/10.5194/tc-2020-86, 2020.
This paper aims to find an ensemble method that combines the global climate models, providing an accurate forecast of sea ice thickness (SIT). An improved multimodel superensemble is proposed in SIT prediction, showing better performance than other mainstream methods. Large biases between the proposed model and observations in SIT simulation are found along the coastline in the west Arctic, and in August, being in accordance with the largest SIT anomaly in time and space.
This paper aims to find an ensemble method that combines the global climate models, providing an...