• 全国中文核心期刊
  • 中国科技核心期刊
  • 中国科学引文数据库收录期刊
  • 世界期刊影响力指数(WJCI)报告来源期刊
  • Scopus, CA, DOAJ, EBSCO, JST等数据库收录期刊

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于机器学习的西南岩溶泉流量模拟研究

马从文 张志才 陈喜 程勤波 彭韬 张林

马从文,张志才,陈 喜,等. 基于机器学习的西南岩溶泉流量模拟研究[J]. 中国岩溶,2024,43(1):48-56 doi: 10.11932/karst2023y013
引用本文: 马从文,张志才,陈 喜,等. 基于机器学习的西南岩溶泉流量模拟研究[J]. 中国岩溶,2024,43(1):48-56 doi: 10.11932/karst2023y013
MA Congwen, ZHANG Zhicai, CHEN Xi, CHENG Qinbo, PENG Tao, ZHANG Lin. Modelling karst spring flow in Southwest China based on machine learning[J]. CARSOLOGICA SINICA, 2024, 43(1): 48-56. doi: 10.11932/karst2023y013
Citation: MA Congwen, ZHANG Zhicai, CHEN Xi, CHENG Qinbo, PENG Tao, ZHANG Lin. Modelling karst spring flow in Southwest China based on machine learning[J]. CARSOLOGICA SINICA, 2024, 43(1): 48-56. doi: 10.11932/karst2023y013

基于机器学习的西南岩溶泉流量模拟研究

doi: 10.11932/karst2023y013
基金项目: 自然科学基金重点项目(41571130071);面上项目(41971028,41571020)
详细信息
    作者简介:

    马从文(1999-),男,硕士研究生,主要研究方向为岩溶流域水文。E-mail:756073751@qq.com。

    通讯作者:

    张志才(1980-),男,博士,教授,研究方向为岩溶水文学。E-mail:zhangzhicai_0@hhu.edu.cn。

  • 中图分类号: P641.134

Modelling karst spring flow in Southwest China based on machine learning

  • 摘要: 岩溶泉对西南岩溶区生态系统稳定和经济社会发展具有重要意义。受岩溶区独特水文地质结构与多重水流过程控制,岩溶泉流量具有复杂的动态变化特征,机器学习模型为其模拟和预测提供了有效手段。然而,岩溶泉域降雨−泉流量过程及其时空变异特征对机器学习模型结构与模拟精度的影响仍不明晰。本文选取西南典型岩溶泉,基于长短期记忆网络(LSTM)建立岩溶泉流量模拟模型,利用泉域实测逐小时降雨与泉流量序列进行模型训练与验证。在此基础上,分析了不同降雨−泉流量过程对岩溶泉流量模拟精度的影响,以及岩溶水文地质结构对降雨−泉流量响应滞时的控制作用。研究结果显示,山坡岩溶泉与流域出口岩溶泉训练期纳什效率系数(NSE)分别为0.942与0.951,验证期分别为0.831与0.834。对于山坡岩溶泉与流域出口岩溶泉,利用全年实测序列训练的模型预测雨季泉流量存在较大偏差,NSE分别为0.793与0.798,而利用雨季实测序列训练的模型预测雨季泉流量,精度显著提升,NSE分别为0.956与0.962,且此差异在暴雨频繁的5、6、7月尤为显著。受浅薄土壤与表层岩溶带分布影响,山坡岩溶泉LSTM模型时序步长显著小于流域出口岩溶泉。

     

  • 图  1  陈旗流域地形、岩性及水文气象观测

    Figure  1.  Topography, lithology and hydrometeorological observations of the Chenqi watershed

    图  2  LSTM模型原理示意图

    Figure  2.  Schematic diagram of LSTM model: neural network structure (a); LSTM hidden layer structure (b)

    图  3  山坡岩溶泉与出口岩溶泉LSTM模型训练期与验证期泉流量模拟结果

    Figure  3.  Simulation results of spring flow in the training and validation period of LSTM model: hillside karst spring (a); outlet karst spring (b)

    图  4  基于雨季实测序列训练的模型模拟结果

    Figure  4.  Simulation results from the model trained based on the sequence in the rainy season: hillside karst spring (a); outlet karst spring (b)

    图  5  基于不同降雨−泉流量形成特征模型的泉流量模拟结果

    Figure  5.  Simulation results of spring flow based on different models: hillside karst spring (a); outlet karst spring (b)

    图  6  时序步长与LSTM模型精度的关系

    Figure  6.  Relationship between timing steps and LSTM

    表  1  研究期降雨、泉流量统计特征

    Table  1.   Statistical characteristics of rainfall and spring flow during the study period

    站点降雨量泉流量/m3·s−1
    时段总量/mm最大值/mm·h−1最大值最小值均值
    山坡岩溶泉 3 584.2 59.9 0.002 3 0 0.000 2
    出口岩溶泉   0.164 6 0 0.006 8
    下载: 导出CSV

    表  2  超参数取值

    Table  2.   Hyperparameter values

    超参数名称含义取值
    隐藏层数(layer)神经网络结构层数2
    单个隐藏层神经元数(hidden size)隐藏层节点的数量10
    训练轮数(epoch)模型进行完整训练的次数100
    批量大小(batch size)一次训练所选取的样本数32
    学习率(learning rate)神经网络的学习速度0.001
    时序步长(time step)山坡岩溶泉数据传递时长大小15
    出口岩溶泉28
    下载: 导出CSV

    表  3  基于全年数据训练的验证期雨、枯季LSTM模拟结果

    Table  3.   LSTM simulation results for the rainy and dry seasons in validation period trained based on annual data

    时间山坡岩溶泉出口岩溶泉
    NSER2NSER2
    训练期0.9420.9880.9510.989
    验证期枯季0.9730.9870.9820.993
    雨季0.7930.9420.7980.935
    下载: 导出CSV
  • [1] Hartmann A, Goldscheider N, Wagener T, Lange J, Weiler M. Karst water resources in a changing world: Review of hydrological modeling approaches[J]. Reviews of Geophysics, 2014, 52(3): 218-242. doi: 10.1002/2013RG000443
    [2] 李潇, 漆继红, 许模. 西南典型紧窄褶皱小尺度浅层岩溶水系统特征及隧道涌水分析[J]. 中国岩溶, 2020, 39(3):375-383.

    LI Xiao, QI Jihong, XU Mo. Analysis on the characteristics of small-scale shallow karst water systems in typical tight-narrow folds and tunnel water inrush in Southwestern China[J]. Carsologica Sinica, 2020, 39(3): 375-383.
    [3] Zhang Zhicai, Chen Xi, Soulsby Chris. Catchment-scale conceptual modelling of water and solute transport in the dual flow system of the karst critical zone[J]. Hydrological Processes, 2017, 31(19): 3421-3436. doi: 10.1002/hyp.11268.
    [4] 蒙海花, 王腊春, 苏维词, 霍雨. 基于落水洞的岩溶半分布式水文模型的构建及其应用[J]. 地理科学, 2009, 29(4):550-554. doi: 10.3969/j.issn.1000-0690.2009.04.014

    MENG Haihua, WANG Lachun, SU Weici, HUO Yu. Development of a karst sinkhole-based semi-distributed hydrological model and its application[J]. Scientia Geographica Sinica, 2009, 29(4): 550-554. doi: 10.3969/j.issn.1000-0690.2009.04.014
    [5] 王宇. 西南岩溶区岩溶水系统分类、特征及勘查评价要点[J]. 中国岩溶, 2002, 21(2):114-119. doi: 10.3969/j.issn.1001-4810.2002.02.008

    WANG Yu. Classification, features of karst water system and key point for the evaluation to karst water exploration in Southwest China karst area[J]. Carsologica Sinica, 2002, 21(2): 114-119. doi: 10.3969/j.issn.1001-4810.2002.02.008
    [6] 王宇. 岩溶高原地下水径流系统垂向分带[J]. 中国岩溶, 2018, 37(1):1-8.

    WANG Yu. Vertical zoning of groundwater runoff system in karst plateau[J]. Carsologica Sinica, 2018, 37(1): 1-8.
    [7] 王在高, 徐萍莉. 浅析喀斯特流域水文地貌过程—响应系统[J]. 贵州师范大学学报(自然科学版), 2002, 20(1):36-39.

    WANG Zaigao, XU Pingli. An elementary analysis of hydro-geomorphological process—response system[J]. Journal of Guizhou Normal University (Natural Sciences), 2002, 20(1): 36-39.
    [8] Hu Caihong, Wu Qiang, Li Hui, Jian Shengqi, Li Nan, Lou Zhengzheng. Deep learning with a long short-term memory networks approach for rainfall-runoff simulation[J]. Water, 2018, 10(11): 1543. doi: 10.3390/w10111543
    [9] Goodfellow I, Bengio Y, Courville A. Deep learning[M]. Cambridge, USA: The MIT Press, 2016.
    [10] Kratzert F, Klotz D, Brenner C, Schulz K. Rainfall-runoff modelling using Long Short-Term Memory(LSTM) networks[J]. Hydrology and Earth System Sciences, 2018, 22(11): 6005-6022. doi: 10.5194/hess-22-6005-2018
    [11] 党池恒, 张洪波, 陈克宇, 支童, 卫星辰. 长短期记忆神经网络在季节性融雪流域降水-径流模拟中的应用[J]. 华北水利水电大学学报(自然科学版), 2020, 41(5):10-18, 33. doi: 10.19760/j.ncwu.zk.2020057

    DANG Chiheng, ZHANG Hongbo, CHEN Keyu, ZHI Tong, WEI Xingchen. Application of the long-short-term memory neural network for rainfall-runoff simulation in seasonal snowmelt basin[J]. Journal of North China University of Water Resources and Electric Power (Natural Science Edition), 2020, 41(5): 10-18, 33. doi: 10.19760/j.ncwu.zk.2020057
    [12] Y Bengio, P Simard, P Frasconi. Learning long-term dependencies with gradient descent is difficult[J]. IEEE Transactions on Neural Networks, 1994, 5(2): 157-166. doi: 10.1109/72.279181
    [13] Hochreiter S. The vanishing gradient problem during learning recurrent neural nets and problem solutions[J]. International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems, 1998, 6(2): 107-116. doi: 10.1142/S0218488598000094
    [14] Hochreiter S, Schmidhuber J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780. doi: 10.1162/neco.1997.9.8.1735
    [15] Gao Shuai, Huang Yuefei, Zhang Shuo, Han Jingcheng, Wang Guangqian, Zhang Meixin, Lin Qingsheng. Short-term runoff prediction with GRU and LSTM networks without requiring time step optimization during sample generation[J]. Journal of Hydrology, 2020, 589: 125188. doi: 10.1016/j.jhydrol.2020.125188
    [16] Zhang Di, Lin Junqiang, Peng Qidong, Wang Dongsheng, Yang Tiantian, Soroosh Sorooshian, Liu Xuefei, Zhuang Jiangbo. Modeling and simulating of reservoir operation using the artificial neural network, support vector regression, deep learning algorithm[J]. Journal of Hydrology, 2018, 565: 720-736. doi: 10.1016/j.jhydrol.2018.08.050
    [17] Tennant C, Larsen L, Bellugi D, Moges E, Zhang Liang, Ma Hongxu. The utility of information flow in formulating discharge forecast models: A case study from an arid snow-dominated catchment[J]. Water Resources Research, 2020, 56(8): e2019WR024908.
    [18] 顾逸. 基于长短期记忆循环神经网络及其结构约减变体的中长期径流预报研究[D]. 武汉: 华中科技大学, 2018.

    GU Yi. Research on mid-long-term runoff forecasting based on long-short-term memory recurrent networks and its structural reduction variant[D]. Wuhan: Huazhong University of Science and Technology, 2018.
    [19] Cho K, Kim Y. Improving streamflow prediction in the WRF-Hydro model with LSTM networks[J]. Journal of Hydrology, 2022, 605: 127297. doi: 10.1016/j.jhydrol.2021.127297
    [20] Lu Dan, Goutam Konapala, Scott L Painter, Shih Chieh Kao, Sudershan Gangrade. Streamflow simulation in data-scarce basins using bayesian and physics-Informed machine learning models[J]. Journal of Hydrometeorology, 2021, 22(6): 1421-1438.
    [21] 张志才, 陈喜, 刘金涛, 彭韬, 石朋, 严小龙. 喀斯特山体地形对表层岩溶带发育的影响:以陈旗小流域为例[J]. 地球与环境, 2012, 40(2):137-143.

    ZHANG Zhicai, CHEN Xi, LIU Jintao, PENG Tao, SHI Peng, YAN Xiaolong. Influence of topography on epikarst in karst mountain areas: A case study of Chenqi catchment[J]. Earth and Environment, 2012, 40(2): 137-143.
    [22] Zhang Zhicai, Chen Xi, Cheng Qinbo, Soulsby Chris. Using Storage Selection(SAS) functions to understand flow paths and age distributions in contrasting karst groundwater systems[J]. Journal of Hydrology, 2021, 602: 126785
    [23] Zhang Zhicai, Chen Xi, Cheng Qinbo, Soulsby Chris. Storage dynamics, hydrological connectivity and flux ages in a karst catchment: Conceptual modelling using stable isotopes[J]. Hydrology and Earth System Sciences, 2019, 23(1): 51-71. doi: 10.5194/hess-23-51-2019
    [24] Cai Lianbin, Chen Xi, Huang Richao, Keith Smettem. Runoff change induced by vegetation recovery and climate change over carbonate and non-carbonate areas in the karst region of South-west China[J]. Journal of Hydrology, 2022, 604: 127231.
    [25] 侯怡萍. 流域生态水文敏感度及其影响因子分析[D]. 成都: 电子科技大学, 2019.

    HOU Yiping. Research on ecohydrological sensitivity of watershed and its influencing factors analysis[D]. Chengdu: University of Electronic Science and Technology of China, 2019.
    [26] 刘炜, 焦树林, 李银久, 莫跃爽, 赵宗权, 张洁, 赵梦. 喀斯特地表植被覆盖变化及其与气候因子相关性分析[J]. 水土保持研究, 2021, 28(3):203-215. doi: 10.13869/j.cnki.rswc.2021.03.024

    LIU Wei, JIAO Shulin, LI Yinjiu, MO Yueshuang, ZHAO Zongquan, ZHANG Jie, ZHAO Meng. Analysis on the correlation between vegetation cover of land surface and climatic factors in karst area[J]. Soil and Water Conservation Research, 2021, 28(3): 203-215. doi: 10.13869/j.cnki.rswc.2021.03.024
    [27] 李林立. 西南典型岩溶区生态环境对表层岩溶水调蓄功能的影响研究[D]. 重庆: 西南大学, 2009.

    LI Linli. Study of effects of ecological environment on regulated function of epikarst water in typical karst area of Southwest, China[D]. Chongqing: Southwest University, 2009.
    [28] Zhang Zhicai, Chen Xi, Chen Xunhong, Shi Peng. Quantifying time lag of epikarst-spring hydrograph response to rainfall using correlation and spectral analyses[J]. Hydrogeology Journal, 2013, 21(7): 1619-1631. doi: 10.1007/s10040-013-1041-9
  • 加载中
图(6) / 表(3)
计量
  • 文章访问数:  1301
  • HTML浏览量:  1116
  • PDF下载量:  66
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-07-27
  • 网络出版日期:  2023-03-09
  • 刊出日期:  2024-02-01

目录

    /

    返回文章
    返回