Abstract
Causal discovery from river runoff data aids flood prevention and mitigation strategies, garnering attention in climate and earth science. However, most climate causal discovery methods rely on conditional independence approaches, overlooking the non-stationary characteristics of river runoff data and leading to poor performance. In this paper, we propose a river runoff causal discovery method based on deep reinforcement learning, called RCD-DRL, to effectively learn causal relationships from non-stationary river runoff time series data. The proposed method utilizes an actor-critic framework, which consists of three main modules: an actor module, a critic module, and a reward module. In detail, RCD-DRL first employs the actor module within the encoder-decoder architecture to learn latent features from raw river runoff data, enabling the model to quickly adapt to non-stationary data distributions and generating a causality matrix at different stations. Subsequently, the critic network with two fully connected layers is designed to estimate the value of the current encoded features. Finally, the reward module, based on the Bayesian information criterion (BIC), is used to calculate the reward corresponding to the currently generated causal matrix. Experimental results obtained on both synthetic and real datasets demonstrate the superior performance of the proposed method over the state-of-the-art methods.
Graphical abstract
Similar content being viewed by others
Data Availability
The real river runoff data can download from the Bavarian Environmental Agency (Bayerisches Landesamt für Umwelt, https://www.lfu.bayern.de) at https://www.gkd.bayern.de/en/rivers/discharge.
References
Wang C, Liu L, Sun S, Wang W (2022) Rethinking the framework constructed by counterfactual functional model. Appl Intell 52(11):12957–12974
Cai R, Huang L, Chen W, Qiao J, Hao Z (2023) Learning dynamic causal mechanisms from non-stationary data. Appl Intell 53(5):5437–5448
McGraw MC, Barnes EA (2018) Memory matters: A case for granger causality in climate variability studies. J Clim 31(8):3289–3300
Runge J, Tibau XA, Bruhns M, Muñoz-Marí J, Camps-Valls G (2020) The causality for climate competition. In: NeurIPS 2019 Competition and Demonstration Track, pp. 110–120. PMLR
Nowack P, Runge J, Eyring V, Haigh JD (2020) Causal networks for climate model evaluation and constrained projections. Nat Commun 11(1):1415
Peng Z, Liu W, An S (2020) Haze pollution causality mining and prediction based on multi-dimensional time series with ps-fcm. Inf Sci 523:307–317
Zhou W, Yu S, Chen B (2022) Causality detection with matrix-based transfer entropy. Inf Sci 613:357–375
Böhnisch AF, Felsche E, Ludwig R (2023) European heatwave tracks: Using causal discovery to detect recurring pathways in a single-regional climate model large ensemble. Environmental Research Letters
Qu Y, Montzka C, Vereecken H (2021) Causation discovery of weather and vegetation condition on global wildfire using the pcmci approach. In: 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS, pp. 8644–8647. IEEE
Shi H, Zhao Y, Liu S, Cai H, Zhou Z (2022) A new perspective on drought propagation: Causality. Geophys Res Lett 49(2):2021–096758
Song C, Simmons BI, Fortin MJ, Gonzalez A (2022) Generalism drives abundance: A computational causal discovery approach. PLoS Comput Biol 18(9):1010302
Runge J (2023) Modern causal inference approaches to investigate biodiversity ecosystem functioning relationships. nature communications 14(1):1917
Lehmann J, Kretschmer M, Schauberger B, Wechsung F (2020) Potential for early forecast of moroccan wheat yields based on climatic drivers. Geophys Res Lett 47(12):2020–087516
Chopra R, Magazzino C, Shah MI, Sharma GD, Rao A, Shahzad U (2022) The role of renewable energy and natural resources for sustainable agriculture in asean countries: do carbon emissions and deforestation affect agriculture productivity? Resour Policy 76
Mhalla L, Chavez-Demoulin V, Dupuis DJ (2020) Causal mechanism of extreme river discharges in the upper danube basin network. J R Stat Soc: Ser C: Appl Stat 69(4):741–764
Gnecco N, Meinshausen N, Peters J, Engelke S (2021) Causal discovery in heavy tailed models. Ann Stat 49(3):1755–1778
Sheng Z, Wen S, Feng Zk, Gong J, Shi K, Guo Z, Yang Y, Huang T (2023) A survey on data-driven runoff forecasting models based on neural networks. IEEE Transactions on Emerging Topics in Computational Intelligence 7(4):1083–1097. https://doi.org/10.1109/TETCI.2023.3259434
Liu N, Qian L, Yan D, Hu W, Hong M (2023) A nonlinear dynamical model for monthly runoff forecasting in situations of small samples. Mathematical Geosciences 1–21
Liu Q, Li S, Zhou G, Feng Z, Wang G, Qiu Q (2022) Attribution of nonstationary changes in the annual runoff of the Weihe River using the de-nonstationarity method. Hydrol Res 53(3):407–418. https://doi.org/10.2166/nh.2022.106, https://iwaponline.com/hr/article pdf/53/3/407/1029010/nh0530407.pdf
Ebert-Uphoff I, Deng Y (2012) Causal discovery for climate research using graphical models. J Clim 25(17):5648–5665
Zhou X, Huang W, Zhang N, Hu W, Du S, Song G, Xie K (2015) Probabilistic dynamic causal model for temporal data. In: 2015 International Joint Conference on Neural Networks (IJCNN), pp. 1–8. IEEE
Zhao W, Zhang R, Cao H, Tan W (2019) Factor contribution to soil organic and inorganic carbon accumulation in the loess plateau: Structural equation modeling. Geoderma 352:116–125
Wang Y, Liu X, Wang T, Zhang X, Feng Y, Yang G, Zhen W (2021) Relating land-use/land-cover patterns to water quality in watersheds based on the structural equation modeling. CATENA 206
Garland G, Edlinger A, Banerjee S, Degrune F, García-Palacios P, Pescador DS, Herzog C, Romdhane S, Saghai A, Spor A et al (2021) Crop cover is more important than rotational diversity for soil multifunctionality and cereal yields in european cropping systems. Nature Food 2(1):28–37
Runge J, Bathiany S, Bollt E, Camps-Valls G, Coumou D, Deyle E, Glymour C, Kretschmer M, Mahecha MD, Muñoz-Marí J et al (2019) Inferring causation from time series in earth system sciences. Nat Commun 10(1):2553
Runge J, Nowack P, Kretschmer M, Flaxman S, Sejdinovic D (2019) Detecting and quantifying causal associations in large nonlinear time series datasets. Sci Adv 5(11):4996
Harries D, O’Kane TJ (2021) Dynamic bayesian networks for evaluation of granger causal relationships in climate reanalyses. Journal of Advances in Modeling Earth Systems 13(5):2020–002442
Silva FN, Vega-Oliveros DA, Yan X, Flammini A, Menczer F, Radicchi F, Kravitz B, Fortunato S (2021) Detecting climate teleconnections with granger causality. Geophys Res Lett 48(18):2021–094707
Kumar P, Kuttippurath J, Mitra A (2022) Causal discovery of drivers of surface ozone variability in antarctica using a deep learning algorithm. Environmental science: Processes & impacts 24(3):447–459
Ye Z, Tozuka T (2022) Causal relationship between sea surface temperature and precipitation revealed by information flow. Frontiers in Climate 4:1024384
Runge J (2020) Discovering contemporaneous and lagged causal relations in autocor-related nonlinear time series datasets. In: Conference on Uncertainty in Artificial Intelligence, pp. 1388–1397. PMLR
Gerhardus A, Runge J (2020) High-recall causal discovery for autocorrelated time series with latent confounders. Adv Neural Inf Process Syst 33:12615–12625
Lu NY, Zhang K, Yuan C (2021) Improving causal discovery by optimal bayesian network learning. Proceedings of the AAAI Conference on Artificial Intelligence. 35:8741–8748
Zhu S, Ng I, Chen Z (2020) Causal discovery with reinforcement learning. In: International Conference on Learning Representations (2020). https://openreview.net/forum?id=S1g2skStPB
Wang X, Du Y, Zhu S, Ke L, Chen Z, Hao J, Wang J (2021) Ordering-based causal discovery with reinforcement learning. In: IJCAI, pp. 3566–3573. https://doi.org/10.24963/ijcai.2021/491
Yang D, Yu G, Wang J, Wu Z, Guo M (2023) Reinforcement causal structure learning on order graph. Proceedings of the AAAI Conference on Artificial Intelligence. 37:10737–10744
Liu Y, Wu H, Wang J, Long M (2022) Non-stationary transformers: Exploring the stationarity in time series forecasting. In: Advances in Neural Information Processing Systems
Zheng X, Aragam B, Ravikumar PK, Xing EP (2018) Dags with no tears: Continuous optimization for structure learning. Advances in neural information processing systems 31
Asadi P, Davison AC, Engelke S (2015) Extremes on river networks. The Annals of Applied Statistics, 2023–2050
Engelke S, Hitz AS (2020) Graphical models for extremes. J R Stat Soc Ser B Stat Methodol 82(4):871–932
Papagiannopoulou C, Miralles DG, Decubber S, Demuzere M, Verhoest NE, Dorigo WA, Waegeman W (2017) A non-linear granger-causality framework to investigate climate-vegetation dynamics. Geoscientific Model Development 10(5):1945–1960
Weichwald S, Jakobsen ME, Mogensen PB, Petersen L, Thams N, Varando G (2020) Causal structure learning from time series: Large regression coefficients may predict causal links better in practice than small p-values. In: NeurIPS 2019 Competition and Demonstration Track, pp. 27–36. PML
Bussmann B, Nys J, Latré S (2021) Neural additive vector autoregression models for causal discovery in time series. In: Discovery Science: 24th International Conference, DS 2021, Halifax, NS, Canada, October 11–13, 2021, Proceedings 24:446–460. Springer
Sanchez P, Liu X, O’Neil, AQ, Tsaftaris SA (2023) Diffusion models for causal discovery via topological ordering. In: The Eleventh International Conference on Learning Representations. https://openreview.net/forum?id=Idusfje4-Wq
Acknowledgements
This work was partly supported by National Natural Science Foundation of China Research Program (62276010, 62106009), in part by R &D Program of Beijing Municipal Education Commission (KZ202210005009, KM202210005030).
Author information
Authors and Affiliations
Contributions
Junzhong Ji led and designed the study. Ting Wang drafted and revised the article with guidance from Jinduo Liu. Muhua Wang and Wei Tang provided guidance for this study. All authors have read and agreed to the published version of the manuscript.
Corresponding author
Ethics declarations
Conflicts of interest
The authors declare that they have no conflict of interest.
Ethical and informed consent for data used
This study did not involve human or animal subjects. The data used in this paper were obtained from publicly available sources and comply with relevant laws and regulations. The references are clearly indicated.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Ji, J., Wang, T., Liu, J. et al. River runoff causal discovery with deep reinforcement learning. Appl Intell 54, 3547–3565 (2024). https://doi.org/10.1007/s10489-024-05348-7
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10489-024-05348-7