Skip to main content
SpringerLink
Log in

Adaptability of machine learning methods and hydrological models to discharge simulations in data-sparse glaciated watersheds

  • Research article
  • Published:
Journal of Arid Land Aims and scope Submit manuscript

Abstract

The accurate simulation and prediction of runoff in alpine glaciated watersheds is of increasing importance for the comprehensive management and utilization of water resources. In this study, long short-term memory (LSTM), a state-of-the-art artificial neural network algorithm, is applied to simulate the daily discharge of two data-sparse glaciated watersheds in the Tianshan Mountains in Central Asia. Two other classic machine learning methods, namely extreme gradient boosting (XGBoost) and support vector regression (SVR), along with a distributed hydrological model (Soil and Water Assessment Tool (SWAT) and an extended SWAT model (SWAT_Glacier) are also employed for comparison. This paper aims to provide an efficient and reliable method for simulating discharge in glaciated alpine regions that have insufficient observed meteorological data. The two typical basins in this study are the main tributaries (the Kumaric and Toxkan rivers) of the Aksu River in the south Tianshan Mountains, which are dominated by snow and glacier meltwater and precipitation. Our comparative analysis indicates that simulations from the LSTM shows the best agreement with the observations. The performance metrics Nash-Sutcliffe efficiency coefficient (NS) and correlation coefficient (R2) of LSTM are higher than 0.90 in both the training and testing periods in the Kumaric River Basin, and NS and R2 are also higher than 0.70 in the Toxkan River Basin. Compared to classic machine learning algorithms, LSTM shows significant advantages over most evaluating indices. XGBoost also has high NS value in the training period, but is prone to overfitting the discharge. Compared with the widely used hydrological models, LSTM has advantages in predicting accuracy, despite having fewer data inputs. Moreover, LSTM only requires meteorological data rather than physical characteristics of underlying data. As an extension of SWAT, the SWAT_Glacier model shows good adaptability in discharge simulation, outperforming the original SWAT model, but at the cost of increasing the complexity of the model. Compared with the oftentimes complex semi-distributed physical hydrological models, the LSTM method not only eliminates the tedious calibration process of hydrological parameters, but also significantly reduces the calculation time and costs. Overall, LSTM shows immense promise in dealing with scarce meteorological data in glaciated catchments.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

References

  • Aggarwal S K, Goel A, Singh V P. 2012. Stage and discharge forecasting by SVM and ANN techniques. Water Resources Management, 26: 3705–3724.

    Article  Google Scholar 

  • Ajami N, Gupta H, Wagener T, et al. 2004. Calibration of a semi-distributed hydrologic model for streamflow estimation along a river system. Journal of Hydrology, 298: 112–135.

    Article  Google Scholar 

  • Arnold J G, Srinivasan R, Muttiah R S, et al. 1998. Large area hydrologic modeling and assessment—Part 1: Model development. JAWRA Journal of the American Water Resources Association, 34: 73–89.

    Article  Google Scholar 

  • Bengio Y, Simard P, Frasconi P. 1994. Learning Long-Term dependencies with gradient descent is difficult. IEEE Transactions on Neural Networks, 5: 157–166.

    Article  Google Scholar 

  • Chen T Q, Guestrin C. 2016. XGBoost: A Scalable Tree Boosting System. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco: Special Interest Group on Management of Data, 785–794.

    Google Scholar 

  • Chen X, Long D, Hong Y, et al. 2017. Improved modeling of snow and glacier melting by a progressive two-stage calibration strategy with GRACE and multisource data: How snow and glacier meltwater contributes to the runoff of the Upper Brahmaputra River basin? Water Resources Research, 53: 2431–2466.

    Article  Google Scholar 

  • Chen Y N, Takeuchi K, Xu C C, et al. 2006. Regional climate change and its effects on river runoff in the Tarim Basin, China. Hydrological Processes, 20: 2207–2216.

    Article  Google Scholar 

  • Chen Y N, Li W H, Deng H J, et al. 2016. Changes in Central Asia’s Water Tower: Past, Present and Future. Scientific Reports, 6: 35458, doi: https://doi.org/10.1038/srep35458.

    Article  Google Scholar 

  • Chen Y N, Li W H, Fang G H, et al. 2017. Review article: Hydrological modeling in glacierized catchments of Central Asia — status and challenges. Hydrology and Earth System Sciences, 21: 669–684.

    Article  Google Scholar 

  • Cheng C T, Feng Z K, Niu W J, et al. 2015. Heuristic methods for reservoir monthly inflow forecasting: A case study of Xinfengjiang Reservoir in Pearl River, China. Water, 7: 4477–4495.

    Article  Google Scholar 

  • Duethmann D, Bolch T, Farinotti D, et al. 2015. Attribution of streamflow trends in snow and glacier melt-dominated catchments of the Tarim River, Central Asia. Water Resources Research, 51: 4727–4750.

    Article  Google Scholar 

  • Duethmann D, Menz C, Jiang T, et al. 2016. Projections for headwater catchments of the Tarim River reveal glacier retreat and decreasing surface water availability but uncertainties are large. Environmental Research Letters, 11: 054024, doi: https://doi.org/10.1088/1748-9326/11/5/054024.

    Article  Google Scholar 

  • Fan Y T, Chen Y N, Li W H. 2014. Increasing precipitation and baseflow in Aksu River since the 1950s. Quaternary International, 336: 26–34.

    Article  Google Scholar 

  • Fang K, Shen C P, Kifer D, et al. 2017. Prolongation of SMAP to spatiotemporally seamless coverage of continental U.S. using a deep learning neural network. Geophysical Research Letters, 44, doi: https://doi.org/10.1002/2017gl075619.

  • Fang K, Shen C P. 2019. Near-real-time forecast of satellite-based soil moisture using long short-term memory with an adaptive data integration kernel. Journal of Hydrometeorology, 21: 399–413.

    Article  Google Scholar 

  • Feng D P, Fang K, Shen C P. 2020. Enhancing streamflow forecast and extracting insights using Long-Short Term Memory networks with data integration at continental scales. Water Resources Research, 56: e2019WR026793, doi: https://doi.org/10.1029/2019WR026793.

    Article  Google Scholar 

  • Finger D, Vis M, Huss M, et al. 2015. The value of multiple data set calibration versus model complexity for improving the performance of hydrological models in mountain catchments. Water Resources Research, 51: 1939–1958.

    Article  Google Scholar 

  • Fu M L, Fan T C, Ding Z A, et al. 2020. Deep learning data-intelligence model based on adjusted forecasting window scale: Application in daily streamflow simulation. IEEE Access, 8: 32632–32651.

    Article  Google Scholar 

  • Gao S, Huang Y F, Zhang S, et al. 2020. Short-term runoff prediction with GRU and LSTM networks without requiring time step optimization during sample generation. Journal of Hydrology, 589: 125188, doi: https://doi.org/10.1016/j.jhydrol.2020.125188.

    Article  Google Scholar 

  • Greff K, Srivastava R K, Koutnik J, et al. 2017. LSTM: A search space odyssey. IEEE Transactions on Neural Networks and Learning Systems, 28: 2222–2232.

    Article  Google Scholar 

  • Greve P, Burek P, Wada Y. 2020. Using the Budyko framework for calibrating a global hydrological model. Water Resources Research, 56: e2019WR026280, doi: https://doi.org/10.1029/2019WR026280.

    Article  Google Scholar 

  • Hochreiter S, Schmidhuber J. 1997. Long short-term memory. Neural Computation, 9: 1735–1780.

    Article  Google Scholar 

  • Hsu K, Gupta H V, Sorooshia S. 1995. Artificial neural network modeling of the rainfall-runoff process. Water Resources Research, 31: 2517–2530.

    Article  Google Scholar 

  • Hu C H, Wu Q, Li H, et al. 2018. Deep learning with a Long Short-Term Memory networks approach for rainfall-runoff simulation. Water, 10: 1543, doi: https://doi.org/10.3390/w10111543.

    Article  Google Scholar 

  • Hu Y C, Yan L, Hang T, et al. 2020. Stream-flow forecasting of small rivers based on LSTM. arXiv e-prints. arXiv:2001.05681.

  • Immerzeel WW, Van Beek L, Bierkens M. 2010. Climate change will affect the Asian water towers. Science, 328: 1382–1385.

    Article  Google Scholar 

  • Immerzeel W W, Lutz A F, Andrade M, et al. 2020. Importance and vulnerability of the world’s water towers. Nature, 577: 364–369.

    Article  Google Scholar 

  • Ji H P, Fang G H, Yang J, et al. 2019. Multi-objective calibration of a distributed hydrological model in a highly glacierized watershed in Central Asia. Water, 11: 554, doi: https://doi.org/10.3390/w11030554.

    Article  Google Scholar 

  • Jodar J, Carpintero E, Martos-Rosillo S, et al. 2018. Combination of lumped hydrological and remote-sensing models to evaluate water resources in a semi-arid high altitude ungauged watershed of Sierra Nevada (Southern Spain). Science of the Total Environment, 625: 285–300.

    Article  Google Scholar 

  • Kingma D, Ba J. 2015. Adam: A Method for Stochastic Optimization. In: Proceedings of the 3rd International Conference on Learning Representations. San Diego: Computational and Biological Learning Society.

    Google Scholar 

  • Kratzert F, Klotz D, Brenner C, et al. 2018. Rainfall-runoff modelling using Long Short-Term Memory (LSTM) networks. Hydrology and Earth System Sciences, 22: 6005–6022.

    Article  Google Scholar 

  • Kratzert F, Klotz D, Herrnegger M, et al. 2019a. Toward improved predictions in ungauged basins: Exploiting the power of machine learning. Water Resources Research, 55: 11344–11354.

    Article  Google Scholar 

  • Kratzert F, Klotz D, Shalev G, et al. 2019b. Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large-sample datasets. Hydrology and Earth System Sciences, 23: 5089–5110.

    Article  Google Scholar 

  • Kratzert F, Klotz D, Hochreiter S, et al. 2020. A note on leveraging synergy in multiple meteorological datasets with deep learning for rainfall-runoff modeling. Hydrology and Earth System Sciences Discussion. [Preprint]. https://doi.org/10.5194/hess-2020-221, in review.

  • Li W, Kiaghadi A, Dawson C. 2021. High temporal resolution rainfall-runoff modeling using long-short-term-memory (LSTM) networks. Neural Computing and Applications, 33: 1261–1278.

    Article  Google Scholar 

  • Li Z L, Shao Q X, Xu Z X, et al. 2010. Analysis of parameter uncertainty in semi-distributed hydrological models using bootstrap method: A case study of SWAT model applied to Yingluoxia watershed in northwest China. Journal of Hydrology, 385: 76–83.

    Article  Google Scholar 

  • Luo X G, Yuan X H, Zhu S, et al. 2019. A hybrid support vector regression framework for streamflow forecast. Journal of Hydrology, 568: 184–193.

    Article  Google Scholar 

  • Maier H R, Jain A, Dandy G C, et al. 2010. Methods used for the development of neural networks for the prediction of water resource variables in river systems: Current status and future directions. Environmental Modelling & Software, 25: 891–909.

    Article  Google Scholar 

  • Michaud J, Sorooshian S. 1994. Comparison of simple versus complex distributed runoff models on a midsized semiarid watershed. Water Resources Research, 30: 593–605.

    Article  Google Scholar 

  • Ni L L, Wang D, Wu J F, et al. 2020. Streamflow forecasting using extreme gradient boosting model coupled with Gaussian mixture model. Journal of Hydrology, 586: 124901, doi: https://doi.org/10.1016/j.jhydrol.2020.124901.

    Article  Google Scholar 

  • Qin J, Yang K, Liang S L, et al. 2009. The altitudinal dependence of recent rapid warming over the Tibetan Plateau. Climatic Change, 97: 321, doi: https://doi.org/10.1007/s10584-009-9733-9.

    Article  Google Scholar 

  • Razavi T, Coulibaly P. 2013. Streamflow prediction in ungauged basins: Review of regionalization methods. Journal of Hydrologic Engineering, 18: 958–975.

    Article  Google Scholar 

  • Reichstein M, Camps-Valls G, Stevens B, et al. 2019. Deep learning and process understanding for data-driven earth system science. Nature, 566: 195–204.

    Article  Google Scholar 

  • Rudin C. 2019. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence, 1: 206–215.

    Article  Google Scholar 

  • Shen C P. 2018. A transdisciplinary review of deep learning research and its relevance for water resources scientists. Water Resources Research, 54: 8558–8593.

    Article  Google Scholar 

  • Shen C P, Laloy E, Elshorbagy A, et al. 2018. HESS Opinions: Incubating deep-learning-powered hydrologic science advances as a community. Hydrology and Earth System Sciences, 22: 5639–5656.

    Article  Google Scholar 

  • Shen Y J, Shen Y J, Fink M, et al. 2018a. Unraveling the hydrology of the glacierized Kaidu Basin by integrating multisource data in the Tianshan Mountains, Northwestern China. Water Resources Research, 54: 557–580.

    Article  Google Scholar 

  • Shen Y J, Shen Y J, Fink M, et al. 2018b. Trends and variability in streamflow and snowmelt runoff timing in the southern Tianshan Mountains. Journal of Hydrology, 557: 173–181.

    Article  Google Scholar 

  • Tarasova L, Knoche M, Dietrich J, et al. 2016. Effects of input discretization, model complexity, and calibration strategy on model performance in a data-scarce glacierized catchment in Central Asia. Water Resources Research, 52: 4674–4699.

    Article  Google Scholar 

  • Viviroli D, Kummu M, Meybeck M, et al. 2020. Increasing dependence of lowland populations on mountain water resources. Nature Sustainability, 3: 917–928.

    Article  Google Scholar 

  • Xiang Z R, Yan J, Demir I. 2020. A rainfall-runoff model with LSTM-based sequence-to-sequence learning. Water Resources Research, 56, doi: https://doi.org/10.1029/2019wr025326.

  • Yang T, Sun F B, Gentine P, et al. 2019. Evaluation and machine learning improvement of global hydrological model-based flood simulations. Environmental Research Letters, 14: 114027, doi: https://doi.org/10.1088/1748-9326/ab4d5e.

    Article  Google Scholar 

  • Yang T T, Asanjan A A, Welles E, et al. 2017. Developing reservoir monthly inflow forecasts using artificial intelligence and climate phenomenon information. Water Resources Research, 53: 2786–2812.

    Article  Google Scholar 

  • Zheng Z S, Ma Q, Jin S C, et al. 2019. Canopy and terrain interactions affecting snowpack spatial patterns in the Sierra Nevada of California. Water Resources Research, 55: 8721–8739.

    Article  Google Scholar 

Download references

Acknowledgments

This research was supported by the National Natural Science Foundation of China (U1903208, 41630859, 42071046). The authors wish to express great thanks to Prof. YANG Jing from National Institute of Water and Atmospheric Research in New Zealand for his guidance on hydrological models. We are grateful to the anonymous reviewers and editors for their insightful and constructive comments that helped improve the manuscript.

Author information

Authors and Affiliations

  1. State Key Laboratory of Desert and Oasis Ecology, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, Urumqi, 830011, China

    Huiping Ji, Yaning Chen, Gonghuan Fang, Zhi Li, Weili Duan & Qifei Zhang

  2. University of Chinese Academy of Sciences, Beijing, 100049, China

    Huiping Ji, Yaning Chen, Gonghuan Fang, Zhi Li, Weili Duan & Qifei Zhang

Authors
  1. Huiping Ji
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Yaning Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Gonghuan Fang
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Zhi Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Weili Duan
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Qifei Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Yaning Chen.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ji, H., Chen, Y., Fang, G. et al. Adaptability of machine learning methods and hydrological models to discharge simulations in data-sparse glaciated watersheds. J. Arid Land 13, 549–567 (2021). https://doi.org/10.1007/s40333-021-0066-5

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s40333-021-0066-5

Keywords

Search

Navigation