同济大学计算机科学与技术学院 AI +大气海洋研究团队

同济大学计算机科学与技术学院 AI +大气海洋研究团队主要由两名正教授、一名助理教授及其指导下的16名博士生和14名硕士生所组成。主要从事人工智能与大气海洋科学的交叉科学研究。研究方向涵盖：人工智能及其可解释性、机器学习、神经网络、大气及海洋大数据分析、智能数据同化、AI 多圈层耦合、大模型误差溯源等。当前主持国家自然科学基金【原创探索】计划项目一项、国家自然科学基金联合基金【重点】项目两项、国家重点研发计划课题一项。团队在该研究方向已发表高影响因子学术论文150余篇，获授国家发明专利17项。研发的“天行”气象大模型入选中国气象局人工智能天气预报大模型示范计划。研发的 SmaAt-UNet 海冰预测系统在国际海冰预测网络（SIPN）中预报精度排名全球第五，并为“雪龙2”号科考船2024航次提供了精准海冰预报。团队研究地点：同济大学嘉定校区济事楼316左实验室。

1、国家自然科学基金联合基金【重点】项目，“基于预报误差溯源的气象大模型优化研究”，项目编号：U2542212，2026.01-2029.12，主持。

2、国家自然科学基金联合基金【重点】项目，“基于因果推断和物理引导的面向天气预报与气候预测的深度学习理论算法及可解释性研究”，项目编号：U2142211，2022.01-2025.12，主持。

3. 国家自然科学基金【面上】项目，“多模态数据驱动的海气耦合台风概率预报模型”，项目编号：42075141，2021.01-2024.12，主持。

4. 国家重点研发计划“全球变化及应对”专项项目，“大数据与深度学习方法创新地球系统模式发展及应用研究”之课题四“深度神经网络预测模型发展与动力模式结果订正方法研究”，课题编号2020YFA0608004，2020-11至2025-04，第二负责人。

5. 上海市2020年度“科技创新行动计划”社会发展科技攻关“公共安全/突发公共安全事件应急处理处置”专题项目，“基于风云卫星智能精准观测针对极端天气事件的长三角航空运行安全应对研究”之课题二“针对CNOP的高效智能算法开发与应用“，课题编号 20dz1200702， 2020-09-01至2023-08-31，主持。

主持或参与科研项目（课题）情况：

1、国家自然科学基金联合基金【重点】项目，“基于预报误差溯源的气象大模型优化研究”，项目编号：U2542212，2026.01-2029.12，主持。

2、国家自然科学基金联合基金【重点】项目，“基于因果推断和物理引导的面向天气预报与气候预测的深度学习理论算法及可解释性研究”，项目编号：U2142211，2022.01-2025.12，主持。

3. 国家自然科学基金【面上】项目，“多模态数据驱动的海气耦合台风概率预报模型”，项目编号：42075141，2021.01-2024.12，主持。

4. 国家重点研发计划“全球变化及应对”专项项目，“大数据与深度学习方法创新地球系统模式发展及应用研究”之课题四“深度神经网络预测模型发展与动力模式结果订正方法研究”，课题编号2020YFA0608004，2020-11至2025-04，第二负责人。

5. 上海市2020年度“科技创新行动计划”社会发展科技攻关“公共安全/突发公共安全事件应急处理处置”专题项目，“基于风云卫星智能精准观测针对极端天气事件的长三角航空运行安全应对研究”之课题二“针对CNOP的高效智能算法开发与应用“，课题编号 20dz1200702， 2020-09-01至2023-08-31，主持。

主持研发出的模型与系统：

【1】“天行”气象大模型，入选中国气象局人工智能天气预报大模型示范计划。

【2】 SmaAt-UNet 海冰预测系统，在国际海冰预测网络（SIPN）中预报精度排名全球第五，并为“雪龙2”号科考船2024航次提供了精准海冰预报。

发表的学术论文：

【1】A DeepLearning–Based Bias Correction Model for Tropical Cyclone Track and Intensitytowards Forecasting of the TianXing Large Weather Model ，Yuan,Shijin; 王星洲; 穆斌; Wang,Guansong; Niu, Zeyi，Advances in AtmosphericSciences | 2026年 | 43卷 | 3期 | 612-630页

【2】Assessment ofTropical Cyclone Disaster Damage Based on Learnable Inter-City Interaction GNN ，Yuan,Shijin; Yang, Laiyu; 穆斌; 秦博;Huang, Yanjun，Journal of MeteorologicalResearch | 2025年 | 5期 | 1146-1166页

【3】AnInterpretable NAO Daily Prediction Model Considering Weighted Causal Effects ofPhysical Processes ，Yuan, Shijin; Wu, Haoyu; 穆斌; Cui, Yuehan; 秦博，Journal of MeteorologicalResearch | 2025年 | 39卷 | 05期 | 1126-1145页

【4】EAAC-S2S:East Asian Atmospheric Circulation S2S Forecasting with a Deep Learning ModelConsidering Multi-Sphere Coupling ，穆斌; 陈宇轩; 袁时金; 秦博; Liu,Zhenchen，Advances in AtmosphericSciences | 2025年 | 42卷 | 7期 | 1442 - 1462页

【5】TianXing: ALinear Complexity Transformer Model with Explicit Attention Decay for GlobalWeather Forecasting ，袁时金; Wang, Guansong; 穆斌; Zhou, Feifan，ADVANCES IN ATMOSPHERICSCIENCES | 2025年 | 42卷 | SI期 | 9-25页

【6】DevelopingIntelligent Earth System Models : An AI scheme of K-profile parameterizationand stable coupling into CESM with FTA ，穆斌; Yang, Kang; 秦博; Li, Hao; 袁时金，OceanModelling | 2025年 | 197卷

【7】Prediction ofthe summertime Northwest Pacific subtropical high based on ConvLSTM ，Yang,Fei; Ma, Jing; Lan, Hongxia; 穆斌; 袁时金，Atmospheric and Oceanic Science Letters | 2025年

【8】Incorporatingheat budget dynamics in a Transformer-based deep learning model for skillfulENSO prediction，穆斌; 崔悦涵; 袁时金; 秦博，NPJ CLIMATE AND ATMOSPHERIC SCIENCE | 2024年 | 7卷 | 1期

【9】基于深度学习的全球热带气旋生成预测模型及其可解释性分析，穆斌; 王馨; 袁时金; 陈宇轩; 王冠淞等7名作者，中国科学:地球科学 | 2024年 | 54卷 | 12期 | 3708-3733页

【10】MultivariateUpstream Kuroshio Transport (UKT) Prediction and Targeted Observation SensitiveArea Identification of UKT Seasonal Reduction，穆斌; Yang-Hu, Yifan; 秦博; 袁时金，Ocean Modelling | 2024年 | 189卷

【11】A generativeadversarial network-based unified model integrating bias correction anddownscaling for global SST，袁时金; 冯新; 穆斌; 秦博; 王馨等6名作者，Atmosphericand Oceanic Science Letters | 2024年 | 17卷 | 1期

【12】Toward aLearnable Climate Model in the Artificial Intelligence Era，Huang, Gang; Wang, Ya; Ham,Yoo-Geun; 穆斌; Tao, Weichen等6名作者，Advances in Atmospheric Sciences | 2024年 |41卷 | 7期 | 1281-1288页

【13】A deeplearning-based bias correction model for Arctic sea ice concentration towardsMITgcm，袁时金; 朱师辰; Luo, Xiaodan; 穆斌，Ocean Modelling | 2024年 | 188卷

【14】Developingintelligent Earth System Models: An AI framework for replacing sub-modulesbased on incremental learning and its application，穆斌; 赵紫君; 袁时金; 秦博; Dai,Guo-Kun等6名作者，AtmosphericResearch | 2024年 | 302卷

【15】An extensionto ensemble forecast of conditional nonlinear optimal perturbation consideringnonlinear interaction between initial and model parametric uncertainties ，穆斌; 赵紫君; 袁时金; Chen, Xing-Rong; 秦博等6名作者，Atmospheric Research | 2024年 | 311卷

【16】A deeplearning-based global tropical cyclogenesis prediction model and itsinterpretability analysis，穆斌; Wang, Xin; 袁时金; 陈宇轩; 王冠淞等7名作者，ScienceChina Earth Sciences | 2024年 | 67卷 | 12期 | 3671-3695页

【17】IceTFTv1.0.0:interpretable long-term prediction of Arctic sea ice extent with deeplearning，穆斌; 罗晓丹; 袁时金; Liang,Xi，GEOSCIENTIFIC MODEL DEVELOPMENT |2023年 | 16卷 | 16期 |4677-4697页

【18】Aparalleledembedding high-dimensional Bayesian optimization with additiveGaussian kernelsfor solving CNOP，袁时金; 刘娅璇; Qin,Bo; 穆斌; Zhang, Kun，OceanModelling | 2023年 | 184卷

【19】Aradiativetransfer deep learning model coupled into WRF with a generic fortrantorchadaptor，穆斌; 陈璐; 袁时金; Qin, Bo，FRONTIERS IN EARTH SCIENCE | 2023年 | 11卷

【20】Dimensionshiftingbased intelligent algorithm framework to solve conditional nonlinearoptimalperturbation，袁时金; 刘娅璇;Zhang,Huazhen; 穆斌，Computers and Geosciences | 2023年 | 176卷

【21】NAOSeasonalForecast Using a Multivariate Air–Sea CoupledDeepLearning Model Combined with Causal Discovery，穆斌; 姜欣; 袁时金; 崔悦涵; Qin, Bo，Atmosphere | 2023年 | 14卷 | 5期

【22】ErrorEvolutionsand Analyses on Joint Effects of SST and SL via Intermediate CoupledModels andConditional Nonlinear Optimal Perturbation Method，穆斌; 秦小云; 袁时金; Qin, Bo，JOURNALOF MARINE SCIENCE AND ENGINEERING | 2023年 | 11卷 | 5期

【23】Estimatingthetropical cyclone wind structure using physics-incorporated networks，袁时金; 尤钱湖; 穆斌; 秦博; Xu Jing，FRONTIERS IN EARTH SCIENCE | 2023年 | 10卷

【24】PIRT:APhysics-Informed Red Tide Deep Learning Forecast ModelConsideringCausal-Inferred Predictors Selection，穆斌; 秦博; 袁时金; Wang, Xin; Chen, Yuxuan，IEEE Geoscience and Remote Sensing Letters | 2023年 | 20卷

【25】CAU:ACausality Attention Unit for Spatial-temporal Sequence Forecast，Qin, Bo; Meng, Fanqing; Fang, Xianghui; Dai, Guokun; 袁时金等6名作者，IEEE Transactions on Multimedia | 2023年 | 1-15页

【26】ENSO-GTC:ENSODeep Learning Forecast Model With a Global Spatial-TemporalTeleconnectionCoupler，穆斌; 秦博; 袁时金，Journal of Advances in Modeling Earth Systems | 2022年 | 14卷 | 12期

【27】Featureextraction-basedintelligent algorithm framework with neural network forsolving conditionalnonlinear optimal perturbation，袁时金;张华桢; 刘娅璇; 穆斌，SoftComputing | 2022年 | 26卷 | 14期 | 6907-6924页

【28】Adeep learningurban traffic congestion forecast model blending the temporalcontinuity andperiodicity，穆斌; Huang, Yuxi，ACMInternational Conference Proceeding Series | 2022年 |602-607页

【29】EnsembleForecastfor Tropical Cyclone Based on CNOP-P Method: A Case Study of WRF Modeland TwoTyphoons，袁时金; Shi Bo; 赵紫君; 穆斌; Zhou Fei-fan等6名作者，JOURNALOF TROPICAL METEOROLOGY | 2022年 | 28卷 | 2期 | 121-138页

【30】Simulation,precursoranalysis and targeted observation sensitive area identification fortwo types ofENSO using ENSO-MC v1.0 ，穆斌; 崔悦涵; 袁时金; 秦博，GEOSCIENTIFICMODELDEVELOPMENT | 2022年 | 15卷| 10期 | 4105-4127页

【31】OptimalPrecursorsIdentification for North Atlantic Oscillation Using the ParallelIntelligenceAlgorithm，穆斌; 李婧; 袁时金; 罗晓丹; Dai, Guokun，ScientificProgramming| 2022年 | 2022卷

【32】TheNAOVariability Prediction and Forecasting with Multiple Time Scales Driven byENSOUsing Machine Learning Approaches，穆斌; 李婧; 袁时金; Luo, Xiaodan，ComputationalIntelligenceand Neuroscience | 2022年 | 2022卷

【33】GCNModelcombined with Bi-GRU for traffic prediction，穆斌;Zhen,Lin，Proceedings of SPIE - The InternationalSociety forOptical Engineering | 2022年 | 12259

【34】ENSO-ASC1.0.0:ENSO deep learning forecast model with a multivariate air-sea coupler，穆斌; 秦博; 袁时金，GEOSCIENTIFICMODELDEVELOPMENT | 2021年 | 14卷| 11期 | 6977-6999页

【35】TheELM Modelwith Residual Compensation Based on ARIMA for North AtlanticOscillation IndexPrediction，Luo, Xiaodan; 袁时金; 穆斌; Li, Jing，ACMInternational ConferenceProceeding Series | 2021年 |122-126页

【36】Animprovedcontinuous tabu search algorithm with adaptive neighborhood radiusandincreasing search iteration times strategies，袁时金; 徐运佳; 穆斌; Zhang, Linlin; Ren, Juhui等7名作者，International Journal onArtificialIntelligence Tools | 2021年 | 30卷 | 2期

【37】TyphoonIntensityForecasting Based on LSTM Using the Rolling Forecast Method，袁时金; Wang, Cheng; 穆斌; Zhou, Feifan; Duan,Wansuo，ALGORITHMS | 2021年 | 14卷 | 3期

【38】Efficientexecutionsof community earth system model onto accelerators using GPUs，袁时金; Wang, Cheng; 穆斌; 罗晓丹，ACM International Conference Proceeding Series | 2020年 | 192-199页

【39】CNOP-P-BasedParameterSensitivity Analysis for North Atlantic Oscillation in CommunityEarth SystemModel Using Intelligence Algorithms，穆斌; 李婧; 袁时金; 罗晓丹;Dai,Guokun，ADVANCES IN METEOROLOGY | 2020年 | 2020卷

【40】ApplyingConvolutionalLSTM Network to Predict El Ni?o Events: Transfer Learning fromthe Data ofDynamical Model and Observation，穆斌; Ma,Shaoyang; 袁时金; Xu, Hui，ICEIEC2020 - Proceedings of 2020IEEE 10th International Conference on ElectronicsInformation and EmergencyCommunication，2020年 | 215-219页

【41】DataAssimilationby Artificial Neural Network using Conventional Observation forWRF Model，袁时金; Shi, Bo; 穆斌，ACMInternational ConferenceProceeding Series | 2020年 |62-67页

【42】Multi-scaledownscalingwith bayesian convolution network for ENSO SST pattern，穆斌; 秦博; 袁时金，Proceedings-2020 5th International Conference on Electromechanical Control TechnologyandTransportation, ICECTT 2020 | 2020年 | 359-362页

【43】AClimateDownscaling Deep Learning Model considering the Multiscale SpatialCorrelationsand Chaos of Meteorological Events，穆斌; 秦博; 袁时金; 秦小云，MathematicalProblemsin Engineering | 2020年 | 2020卷

【44】Predictionofnorth atlantic oscillation index associated with the sea level pressureusingDWT-LSTM and DWT-ConvLSTM networks，穆斌; 李婧; 袁时金; 罗晓丹，MathematicalProblemsin Engineering | 2020年 | 2020卷

【45】ApplyingConvolutionalLSTM Network to Predict El Nino Events: Transfer Learning fromThe Data ofDynamical Model and Observation，穆斌; 马少阳; 袁时金; Xu, Hui，PROCEEDINGSOF2020 IEEE 10TH INTERNATIONAL CONFERENCE ON ELECTRONICS INFORMATION ANDEMERGENCYCOMMUNICATION (ICEIEC 2020) | 2020年 | 215-219页

【46】NAOIndexPrediction using LSTM and ConvLSTM Networks Coupled with DiscreteWaveletTransform，穆斌; 李婧; 袁时金; 罗晓丹; Dai, Guokun，Proceedingsofthe International Joint Conference on Neural Networks | 2019年 | 2019-July卷，匈牙利布达佩斯

【47】ENSOForecastingover Multiple Time Horizons Using ConvLSTM Network and RollingMechanism，穆斌; Peng, Cheng; 袁时金;Chen, Lei，Proceedings of the International JointConference on Neural Networks| 2019年 | 2019-July卷，匈牙利布达佩斯

【48】IdentifyingTyphoonTargeted Observations Sensitive Areas Using the Gradient DefinitionBased Method，穆斌; Ren, Juhui; 袁时金; Zhou, Feifan，ASIA-PACIFIC JOURNAL OFATMOSPHERIC SCIENCES | 2019年 | 55卷 | 2期 |195-207页

【49】Predictionofnorth atlantic oscillation index with convolutional LSTM based onensembleempirical mode decomposition，袁时金; 罗晓丹; 穆斌; Li, Jing; Dai, Guokun，Atmosphere | 2019年 | 10卷 | 5期

【50】INTELLIGENTALGORITHMSFOR SOLVING CNOP AND THEIR APPLICATIONS IN ENSO PREDICTABILITY ANDTROPICALCYCLONE ADAPTIVE OBSERVATIONS，穆斌; ZhangLin-lin; 袁时金; 钱一闻; 温仕成等7名作者，JOURNAL OF TROPICAL METEOROLOGY | 2019年 | 25卷 | 1期 | 63-81页

【51】TheOptimalPrecursors for ENSO Events Depicted Using the Gradientdefinition-basedMethodin an Intermediate Coupled Model ，穆斌; Ren, Juhui;袁时金; Zhang, Rong-Hua; Chen, Lei等6名作者，Advances in Atmospheric Sciences | 2019年 |36卷 | 12期 |1381-1392页

【52】Optimalprecursorsof double-gyre regime transitions with an adjoint-free method，袁时金; 李糜; Wang, Qiang; Zhang, Kun; 张华桢等6名作者，Journal of Oceanology and Limnology|2019年 | 37卷 | 4期 | 1137-1153页

【53】CNOP-P-basedparametersensitivity for double-gyre variation in ROMS with simulatedannealing algorithm，袁时金; 张华桢; 李糜; 穆斌，Journal of Oceanology and Limnology |2019年 |37卷 | 3期 | 957-967页

【54】Amodifieddirect search algorithm based on kernel density estimator with threemappingstrategies for solving nonlinear optimization，Zhang,Lin-Lin;穆斌; 袁时金，Journal ofComputers(Taiwan) | 2019年 | 30卷 | 4期 | 17-30页

【55】ParallelPCA-BasedBacterial Foraging Optimization Algorithm for Identifying OptimalPrecursors ofNorth Atlantic Oscillation，穆斌; Jing Li; 袁时金; 罗晓丹; Guokun Dai，2019IEEE21st International Conference on High Performance Computing andCommunications;IEEE 17th International Conference on Smart City; IEEE 5thInternationalConference on Data Science and Systems (HPCC/SmartCity/DSS).Proceedings | 2019年 | 1171-7页

【56】Anovelapproach for solving CNOPs and its application in identifying sensitiveregionsof tropical cyclone adaptive observations，Zhang,Linlin;穆斌; 袁时金; Zhou, Feifan，NONLINEAR PROCESSES IN GEOPHYSICS | 2018年|25卷 | 3期 | 693-712页

【57】Paralleldynamicsearch fireworks algorithm with linearly decreased dimension numberstrategy forsolving conditional nonlinear optimal perturbation，穆斌; 赵珺晖; 袁时金; 颜景豪，Proceedingsof the International Joint Conference on Neural Networks| 2017年 | 2017-May卷 | 2314-2321页，美国阿拉斯加

【58】CNOP-BasedSensitiveAreas Identification for Tropical Cyclone Adaptive Observations withPCAGAMethod ，Zhang, Lin-Lin; 袁时金; 穆斌; Zhou, Fei-Fan，ASIA-PACIFICJOURNAL OFATMOSPHERIC SCIENCES | 2017年 | 53卷 | 1期 | 63-73页

【59】An efficientapproachbased on the gradient definition for solving conditional nonlinearoptimalperturbation ，穆斌; Ren, Juhui; 袁时金，Mathematical Problems in Engineering | 2017年|2017卷

【60】CACO-LD:ParallelContinuous Ant Colony Optimization with Linear Decrease Strategy forSolvingCNOP，袁时金; 陈韵怡; 穆斌，Lecture Notes in Computer Science (including subseries LectureNotesin Artificial Intelligence and Lecture Notes in Bioinformatics) | 2017年 | 10637 LNCS卷 | 494-503页

【61】ParallelModifiedArtificial Bee Colony Algorithm for Solving Conditional NonlinearOptimalPerturbation，Ren, Juhui; 袁时金; 穆斌，Proceedings - 18th IEEE InternationalConference on High PerformanceComputing and Communications, 14th IEEEInternational Conference on Smart Cityand 2nd IEEE International Conference onData Science and Systems,HPCC/SmartCity/DSS 2016 | 2016年 | 333-340页，澳大利亚悉尼

【62】PCAFPforSolving CNOP in Double-Gyre Variation and Its Parallelization on Clusters，袁时金; 李糜; 穆斌;Wang,Jingpeng，Proceedings - 18th IEEEInternationalConference on High Performance Computing and Communications, 14thIEEEInternational Conference on Smart City and 2nd IEEE InternationalConference onData Science and Systems, HPCC/SmartCity/DSS 2016 | 2016年 | 284-291页，澳大利亚悉尼

【63】PCGD:Principalcomponents-based great deluge method for solving CNOP，温仕成; 袁时金; 穆斌;Li,Hongyu; Ren, Juhui，2015 IEEE CONGRESS ONEVOLUTIONARYCOMPUTATION (CEC) | 2015年 | 1513-1520页

【64】PCAGA:PrincipalComponent Analysis Based Genetic Algorithm for Solving ConditionalNonlinearOptimal Perturbation，穆斌; Zhang, Linlin; 袁时金; Li, Hongyu，2015 INTERNATIONALJOINTCONFERENCE ON NEURAL NETWORKS (IJCNN) | 2015年

【65】Paralleldynamicstep size sphere-gap transferring algorithm for solving conditionalnonlinearoptimal perturbation，袁时金; 颜景豪; 穆斌; Li, Hongyu，Proceedings- 2015 IEEE 17thInternational Conference on High Performance Computing andCommunications, 2015IEEE 7th International Symposium on Cyberspace Safety andSecurity and 2015 IEEE12th International Conference on Embedded Software andSystems, HPCC-CSS-ICESS2015 | 2015年 | 559-565页

【66】PPSO:PCA basedparticle swarm optimization for solving conditional nonlinearoptimalperturbation，穆斌; 温仕成; 袁时金; Li, Hongyu，Computers and Geosciences |2015年 | 83卷 | 65-71页

【67】AParallelSensitive Area Selection-Based Particle Swarm Optimization Algorithmfor FastSolving CNOP，Yuan Shijin, Ji Feng, Yan Jinghao,Mu Bin，22nd International Conference on NeuralInformation Processing(ICONIP)，土耳其伊斯坦布尔

【68】ParallelCooperativeCo-evolution Based Particle Swarm Optimization Algorithm forSolving ConditionalNonlinear Optimal Perturbation，YuanShijin, Zhao Li, MuBin，22nd International Conferenceon Neural InformationProcessing (ICONIP)，土耳其伊斯坦布尔

【69】Paralleldynamicstep size sphere-gap transferring algorithm for solving conditionalnonlinearoptimal perturbation，Yuan Shijin, YanJinghao, Mu Bin,Li Hongyu，17th IEEE InternationalConference on HighPerformance Computing and Communications, IEEE 7thInternational Symposium onCyberspace Safety and Security and IEEE 12thInternational Conference onEmbedded Software and Systems, 美国纽约

【70】PCAGA:PrincipalComponent Analysis Based Genetic Algorithm for Solving ConditionalNonlinearOptimal Perturbation，Bin Mu，LinlinZhang，Shijin Yuan，Hongyu Li，2015 International JointConference on Neural Networks (IJCNN)，爱尔兰基拉尼

【71】User-QoS-basedWebService Clustering for QoS Prediction，Fuxin Chen,ShijinYuan, Bin Mu，the 22nd IEEE InternationalConference onWeb Services， CCF-B，美国纽约

【72】PCGD:Principalcomponents-based great deluge method for solving CNOP，Wen,Shicheng，Yuan, Shijin，Mu, Bin，Li, Hongyu，Ren,Juhui，IEEE Congress on Evolutionary Computation, CEC2015，日本仙台