Uncertainties in ocean-mixing parameterizations are primary sources for ocean and climate modeling biases.Due to lack of process understanding,traditional physics-driven parameterizations perform unsatisfactorily in the tropics.Recent advances in the deep-learning method and the new availability of long-term turbulence measurements provide an opportunity to explore data-driven approaches to parameterizing oceanic vertical-mixing processes.Here,we describe a novel parameterization based on an artificial neural network trained using a decadal-long time record of hydrographic and turbulence observations in the tropical Pacific.This data-driven parameterization achieves higher accuracy than current parameterizations,demonstrating good generalization ability under physical constraints.When integrated into an ocean model,our parameterization facilitates improved simulations in both ocean-only and coupled modeling.As a novel application of machine learning to the geophysical fluid,these results show the feasibility of using limited observations and well-understood physical constraints to construct a physics-informed deep-learning parameterization for improved climate simulations.

CAS Key Laboratory of Ocean Circulation and Waves,Institute of Oceanology,and Center for Ocean Mega-Science,Chinese Academy of Sciences,Qingdao 266071,China;Pilot National Laboratory for Marine Science and Technology(Qingdao),Qingdao 266237,China;University of Chinese Academy of Sciences,Beijing 100049,China;Center for Excellence in Quaternary Science and Global Change,Chinese Academy of Sciences,Xi'an 710061,China;College of Earth,Ocean and Atmospheric Sciences,Oregon State University,Corvallis,OR 97331,USA

[1]Yuchao Zhu;Rong-Hua Zhang;James N.Moum;Fan Wang;Xiaofeng Li;Delei Li-.Physics-informed deep-learning parameterization of ocean vertical mixing improves climate simulations)[J].国家科学评论（英文版）,2022(08):160-167

unsatisfactorily

Physics,informed,deep,learning,vertical,mixing,improves,climate,simulations,Uncertainties,parameterizations,are,primary,sources,modeling,biases,Due,lack,understanding,traditional,physics,driven,perform,tropics,Recent,advances,method,new,availability,long,term,turbulence,measurements,provide,opportunity,explore,data,approaches,parameterizing,oceanic,processes,Here,describe,novel,artificial,neural,network,trained,using,decadal,record,hydrographic,observations,tropical,Pacific,This,achieves,higher,accuracy,than,current,demonstrating,good,generalization,constraints,When,integrated,into,facilitates,improved,both,only,coupled,application,machine,geophysical,fluid,these,results,show,feasibility,limited,well,understood,construct

0.563012