FuXi-S2S: A machine learning model that outperforms conventional global subseasonal forecast models
Authors:
Lei Chen,
Xiaohui Zhong,
Hao Li,
Jie Wu,
Bo Lu,
Deliang Chen,
Shangping Xie,
Qingchen Chao,
Chensen Lin,
Zixin Hu,
Yuan Qi
Abstract:
Skillful subseasonal forecasts are crucial for various sectors of society but pose a grand scientific challenge. Recently, machine learning based weather forecasting models outperform the most successful numerical weather predictions generated by the European Centre for Medium-Range Weather Forecasts (ECMWF), but have not yet surpassed conventional models at subseasonal timescales. This paper intr…
▽ More
Skillful subseasonal forecasts are crucial for various sectors of society but pose a grand scientific challenge. Recently, machine learning based weather forecasting models outperform the most successful numerical weather predictions generated by the European Centre for Medium-Range Weather Forecasts (ECMWF), but have not yet surpassed conventional models at subseasonal timescales. This paper introduces FuXi Subseasonal-to-Seasonal (FuXi-S2S), a machine learning model that provides global daily mean forecasts up to 42 days, encompassing five upper-air atmospheric variables at 13 pressure levels and 11 surface variables. FuXi-S2S, trained on 72 years of daily statistics from ECMWF ERA5 reanalysis data, outperforms the ECMWF's state-of-the-art Subseasonal-to-Seasonal model in ensemble mean and ensemble forecasts for total precipitation and outgoing longwave radiation, notably enhancing global precipitation forecast. The improved performance of FuXi-S2S can be primarily attributed to its superior capability to capture forecast uncertainty and accurately predict the Madden-Julian Oscillation (MJO), extending the skillful MJO prediction from 30 days to 36 days. Moreover, FuXi-S2S not only captures realistic teleconnections associated with the MJO, but also emerges as a valuable tool for discovering precursor signals, offering researchers insights and potentially establishing a new paradigm in Earth system science research.
△ Less
Submitted 5 July, 2024; v1 submitted 15 December, 2023;
originally announced December 2023.
Adaptive locomotion of artificial microswimmers
Authors:
Henwei Huang,
Fazil Emre Uslu,
Panayiota Katsamba,
Qianwen Chao,
Eric Lauga,
Mahmut Selman Sakar,
Bradley J. Nelson
Abstract:
Bacteria can exploit mechanics to display remarkable plasticity in response to locally changing physical and chemical conditions. Compliant structures play a striking role in their taxis behavior, specifically for navigation inside complex and structured environments. Bioinspired mechanisms with rationally designed architectures capable of large, nonlinear deformation present opportunities for int…
▽ More
Bacteria can exploit mechanics to display remarkable plasticity in response to locally changing physical and chemical conditions. Compliant structures play a striking role in their taxis behavior, specifically for navigation inside complex and structured environments. Bioinspired mechanisms with rationally designed architectures capable of large, nonlinear deformation present opportunities for introducing autonomy into engineered small-scale devices. This work analyzes the effect of hydrodynamic forces and rheology of local surroundings on swimming at low Reynolds number, identifies the challenges and benefits of utilizing elastohydrodynamic coupling in locomotion, and further develops a suite of machinery for building untethered microrobots with self-regulated mobility. We demonstrate that coupling the structural and magnetic properties of artificial microswimmers with the dynamic properties of the fluid leads to adaptive locomotion in the absence of on-board sensors.
△ Less
Submitted 24 February, 2019;
originally announced February 2019.
