| 93 | 0 | 107 |
| 下载次数 | 被引频次 | 阅读次数 |
【目的】近年来鄱阳湖极端洪枯事件频现,为揭示鄱阳湖周边地区降雨变化规律,支撑鄱阳湖暴雨洪涝灾情快速处置,【方法】基于峡江等6个水文站点1998—2024年的降雨数据,采用小波分析和TimeSeriesKMeans时间序列聚类算法研究了不同尺度下鄱阳湖周边地区的降雨变化特性,使用CEEMD-Pearson-LSTM混合模型开展鄱阳湖周边地区降雨量预测研究。【结果】结果显示:鄱阳湖周边地区年均降雨量总体最高增速为12mm/a,最高减速为-6 mm/a,8~12a的长时间尺度上,存在由多到少4个周期,4~8a的中时间尺度上,少雨期和多雨期的循环交替激烈。年内月均降雨量分布可分为3类,聚类1月均降雨分布均匀,聚类2和聚类3中6月份降雨量显著增强;相比于LSTM模型,CEEMD-Pearson-LSTM混合模型的预测精度可提升20%。【结论】结果表明:鄱阳湖周边地区雨量多呈现增长趋势,且部分地区月均降雨量分布不平衡的趋势在逐步加剧。结果充分展示了深度学习技术在雨情快速预测中的独特作用,可为极端洪枯事件预测预警方案设计和技术准备提供参考。
Abstract:[Objective] In recent years, extreme flood and drought events have frequently occurred in Poyang Lake. The objective is to reveal the rainfall variation patterns in the surrounding areas of Poyang Lake and facilitate rapid responses to rainstorm and flood disasters in Poyang Lake. [Methods] Based on rainfall data from six hydrological stations, including Xiajiang station, from 1998 to 2024, wavelet analysis and the TimeSeriesKMeans time-series clustering algorithm were used to investigate the rainfall variation characteristics in the surrounding areas of Poyang Lake across various temporal scales. Furthermore, the CEEMD-Pearson-LSTM hybrid model was used to conduct research on rainfall prediction in the surrounding areas of Poyang Lake. [Results] The results showed that the maximum rate of increase in annual average rainfall in the surrounding areas of Poyang Lake was 12 mm/year, while the maximum rate of decrease was -6 mm/year. On a long timescale of 8-12 years, four cycles of decreasing rainfall were observed, whereas on a medium timescale of 4-8 years, there was a pronounced alternation between dry and wet periods. The distribution of monthly average rainfall throughout the year could be categorized into three distinct types. Cluster 1 exhibited a uniform distribution of monthly average rainfall, whereas Clusters 2 and 3 demonstrated a notable increase in rainfall during June. Compared to the LSTM model, the CEEMD-Pearson-LSTM hybrid model improved prediction accuracy by 20%. [Conclusion] The results indicate a growing trend in rainfall across the surrounding areas of Poyang Lake, with a gradual intensification of the uneven distribution of monthly average rainfall in certain regions. These findings underscore the distinctive role of deep learning technology in rapid rainfall prediction, and can provide a reference for the design and technical preparation of prediction and early warning schemes for extreme flood and drought events.
[1] 姚仕明, 范达福, 栾华龙, 等. 鄱阳湖近年极端洪枯情势分析及应对策略[J]. 长江科学院院报, 2024, 41(12): 171-179.
[2] 李文轩, 蒋名亮, 徐力刚, 等. 鄱阳湖浮游植物时空变化特征及其对极端洪枯事件的响应[J]. 湖泊科学, 2024, 36(4): 1001-1013.
[3] 唐国华, 胡坤琳, 甘黎黎. 鄱阳湖流域生态系统服务与人类福祉的时空耦合及空间效应[J]. 水利水电技术(中英文), 2025, 56(8): 78-90.
[4] 聂珉川, 姚静, 龚磊强, 等. 三峡水库调蓄与流域来水对2020年以来鄱阳湖极端洪/枯水情的影响[J]. 湖泊科学, 2026, 38(2): 817-830.
[5] 王容, 李相虎, 薛晨阳, 等. 1960—2012年鄱阳湖流域旱涝急转事件时空演变特征[J]. 湖泊科学, 2020, 32(1): 207-222.
[6] 傅静,胡振鹏,李国文. 鄱阳湖流域水旱灾害对气候变化的响应 [J]. 水利水电技术(中英文), 2025, 56 (11): 21-31.
[7] LIU Y, QIU H J, YANG D D, et al. Deformation responses of landslides to seasonal rainfall based on InSAR and wavelet analysis[J]. Landslides, 2022, 19(1): 199-210.
[8] THOMAS E, ABRAHAM N P. Relationship between sunspot number and seasonal rainfall over Kerala using wavelet analysis[J]. Journal of Atmospheric and Solar-Terrestrial Physics, 2022, 240: 105943.
[9] GAUTAM S P, SILWAL A, BARAL B, et al. Influence of the rainfall and temperature oscillation on air quality in Kathmandu valley: The wavelet analysis[J]. Environmental Engineering Research, 2023, 28(6): 220694.
[10] 原立峰, 杨桂山, 李恒鹏, 等. 近50年来鄱阳湖流域降雨多时间尺度变化规律研究[J]. 长江流域资源与环境, 2014, 23(3): 434-440.
[11] 符莎, 成静清, 刘章君. 鄱阳湖环湖区降水量演变特征分析[J]. 江西水利科技, 2021, 47(2): 105-115.
[12] 谢诗怡, 况润元, 宋子豪. 鄱阳湖水域面积变化特征及对气象因素的响应[J]. 中国农村水利水电, 2022(7): 103-109.
[13] 郭杨. 基于PSO优化BP神经网络的GNSS PWV和气象数据的尾矿库降雨预测研究[J]. 福建冶金, 2023, 52(5): 24-30.
[14] SUDARSAN PATRO B, BARTAKKE P P. Daily rainfall prediction using long short-term memory (LSTM) algorithm[J]. Journal of Agrometeorology, 2024, 26(4): 509-511.
[15] 张帅, 魏正英, 张育斌. 递归神经网络在降雨量预测中的应用研究[J]. 节水灌溉, 2017(5): 63-66.
[16] TIAN Y, CUI R N, FU W L, et al. Urban rainfall runoff prediction using LSTM with DTW-enhanced clustering[J]. Journal of Hydrologic Engineering, 2026, 31: 04025054.
[17] 姜淞川, 陆建忠, 陈晓玲, 等. 基于LSTM网络鄱阳湖抚河流域径流模拟研究[J]. 华中师范大学学报(自然科学版), 2020, 54(1): 128-139.
[18] 赵卉,张明顺,潘润泽.基于小波分析的海绵城市试点未来降雨变化预测分析研究[J].给水排水,2019,55(9):29-35.
[19] 凌铭, 肖丽英, 赵嘉, 等. 基于SVM-CEEMDAN-BiLSTM模型的日降水量预测[J]. 人民珠江, 2023, 44(9): 61-68.
[20] HIRANO T, ISHIKAWA M, NISHINO T, et al. Early heart-rate trajectory phenotypes predict short-term mortality in critically ill patients: A dynamic time-warping cluster analysis[J]. Journal of Clinical Monitoring and Computing, 2026, 40(3): 621-633.
[21] BUERKLE A, EATON W, LOHSE N, et al. EEG based arm movement intention recognition towards enhanced safety in symbiotic Human-Robot Collaboration[J]. Robotics and Computer-Integrated Manufacturing, 2021, 70: 102137.
[22] 马瑞强, 宋宝燕, 丁琳琳, 等. 面向时间序列事件的动态矩阵聚类方法[J]. 计算机科学与探索, 2021, 15(3): 468-477.
[23] GUPTA A K, SHARMA R, OJHA R P. Integrating clustering with LSTM diffusion for video anomaly detection[J]. Applied Soft Computing, 2026, 186: 114185.
[24] LING M W, GUAN Y S, LIAN Y Q, et al. An improved Hydrology-Informed attention LSTM(HIA-LSTM) model for runoff simulation with seasonal snowmelt[J]. Journal of Hydrology, 2026, 670: 135231.
[25] KASHIF K, ŚLEPACZUK R. LSTM-ARIMA as a hybrid approach in algorithmic investment strategies[J]. Knowledge-Based Systems, 2025, 320: 113563.
[26] HUH J, HA S, JEONG S. LSTM-based dynamic correlation forecasting with economic conditions[J]. Finance Research Letters, 2025, 86: 108425.
[27] 姚仕明, 姚立强. 人类活动对鄱阳湖流域水文干旱演变特征的影响[J]. 湖泊科学, 2026, 38(1): 339-352.
[28] 王栋, 魏加华, 章四龙, 等. 基于CEEMD-BP模型的水文时间序列月径流预测[J]. 北京师范大学学报(自然科学版), 2020(3): 376-386.
[29] ZHENG S H, HUANG Q, REN Y, et al. A diagnostic method for incipient cavitation based on CEEMD and optimized BPNN[J]. Measurement, 2026, 271: 120944.
[30] HUANG Y, HU Y F, WU L Z, et al. Price prediction of power transformer materials based on CEEMD and GRU[J]. Global Energy Interconnection, 2024, 7(2): 217-227.
[31] 殷剑敏, 苏布达, 陈晓玲, 等. 鄱阳湖流域气候变化影响评估报告[M]. 北京: 气象出版社, 2011.
[32] 鲁向晖, 张海娜, 胡羊羊. 赣江流域1966—2015年降水量多尺度时空变化特征分析[J]. 水土保持研究, 2021, 28(2): 168-175.
[33] 韩会明, 郭红虎, 游文荪, 等. 赣江流域降水时空分布特征[J]. 江西水利科技, 2020, 46(6): 449-455.
基本信息:
中图分类号:P426.614;P457.6
引用信息:
[1]廖轶鹏,赵金箫,李德龙,等.1998年—2024年鄱阳湖周边地区降雨变化特性与预测研究[J].水利水电技术(中英文)().
基金信息:
国家自然科学基金青年基金(52309029); 江西省水利科技项目(202426ZDKT23); 江西省水利科学院开放研究基金项目(2023SKSH07); 水利部太湖流域水治理重点实验室项目(Rd124005,Rd124006); 江苏省自然科学基金青年项目(BK20250286)
2026-08-04
2026-08-04
2026-08-04