本文已被：浏览 372次 下载 324次
投稿时间：2025-07-01 修订日期：2026-04-17
投稿时间：2025-07-01 修订日期：2026-04-17
中文摘要: 本文利用1961—2022年西北地区东部155个气象站降水量，全球大气、海温、海冰等历史数据资料，基于长短期记忆LSTM深度学习算法，融合了TCN模块和CBAM注意力机制，建立了西北地区东部夏季降水预测模型（CBAM-TCN-LSTM），开展了预测效果检验,并与多种深度学习算法预测能力对比。结果表明：多种深度学习算法融合的降水预测模型预测技巧优于单一深度学习算法模型，独立样本检验阶段（2018-2022年）夏季降水预测PS评分在60%～80%之间，均值达73.8%，ACC除2020年为负值外，其余各年份均为正值，平均值为0.14，较其他模型有显著提升；对比了目前7种主流的机器学习算法、深度学习网络和时序网络模型，CBAM-TCN-LSTM模型在PS、PC、ACC、MAE和RMSE5项指标均优于其他模型；在2023年汛期预测业务中CBAM-TCN-LSTM模型得到成功应用，准确预测了西北地区东部夏季大部地区降水偏少的特征，PS评分90%。通过采用具有强大时序预测能力的LSTM模型基础，融合时间卷积网络模块和嵌入卷积注意力模块，形成的降水预测模型，可以为区域降水预测提供科学依据和技术支持，具有较好的推广应用前景。
Abstract:Based on the precipitation data of 155 meteorological stations in the eastern part of Northwest China from 1961 to 2022, as well as historical datasets including global atmospheric data, sea surface temperature (SST), and sea ice data, this study integrates the Temporal Convolutional Network(TCN) module and the Convolutional Block Attention Module(CBAM) into the Long Short-Term Memory (LSTM) deep learning algorithm. A climate-smart prediction model for summer precipitation in the eastern part of Northwest China(named CBAM-TCN-LSTM) based on the fusion of deep learning algorithms was thereby established. The predictive performance of the model was verified, and its predictive capability was compared with that of multiple other deep learning algorithms.The results show that the intelligent prediction model based on the fusion of multiple deep learning algorithms outperforms single-algorithm deep learning models. During the independent sample validation period(2018–2022), the PS score for summer precipitation prediction ranged from 60% to 80%, with an average value of 73.8%.The Anomaly Correlation Coefficient (ACC) was positive for all years except 2020, with a mean value of 0.14, representing a significant improvement over other models. When compared with 7 current mainstream models (including machine learning algorithms, deep learning networks, and time-series networks), the CBAM-TCN-LSTM model exhibited superior performance across all five evaluation metrics: PS, PC, ACC, Mean Absolute Error (MAE), and Root Mean Square Error (RMSE). Furthermore, the CBAM-TCN-LSTM model was successfully applied in the 2023 flood season forecasting operation, accurately predicting the characteristic of below-normal summer precipitation in most areas of the eastern part of Northwest China, with a PS score of 90%.By building a precipitation prediction model that combines the TCN module and the CBAM module on the basis of the LSTM model (which has strong time-series predictive capability), this study provides a scientific basis and technical support for regional precipitation prediction, and the model holds good prospects for popularization and application.
文章编号：202507010182 中图分类号： 文献标志码：
基金项目:国家自然基金联合基金项目（U22A20577）、宁夏重点研发计划（2022BEG02020）、宁夏青年拔尖人才工程、宁夏智能数字预报技术与应用创新团队和中国气象局创新发展专项（CXFZ2021J024，CXFZ2025Q022）共同资助。
| Author Name | Affiliation | Address |
| zhuxiaowei | Ningxia Climate Center | 宁夏银川市金凤区新昌西路71号 |
| yang Jianling | 宁夏银川市金凤区新昌西路71号 | |
| zhang Suzhao |
引用文本：
zhuxiaowei,yang Jianling,zhang Suzhao,0.Research on Objective Prediction Method for Summer Precipitation in Eastern Northwest China Based on Multi-Deep-Learning Algorithm Fusion[J].Meteor Mon,():-.
zhuxiaowei,yang Jianling,zhang Suzhao,0.Research on Objective Prediction Method for Summer Precipitation in Eastern Northwest China Based on Multi-Deep-Learning Algorithm Fusion[J].Meteor Mon,():-.
zhuxiaowei,yang Jianling,zhang Suzhao,0.Research on Objective Prediction Method for Summer Precipitation in Eastern Northwest China Based on Multi-Deep-Learning Algorithm Fusion[J].Meteor Mon,():-.
zhuxiaowei,yang Jianling,zhang Suzhao,0.Research on Objective Prediction Method for Summer Precipitation in Eastern Northwest China Based on Multi-Deep-Learning Algorithm Fusion[J].Meteor Mon,():-.