Research progress on coupling artificial intelligence and eco-environmental models

doi:10.13287/j.1001-9332.202301.019

Abstract

Abstract: Artificial intelligence (AI) has been widely used in the eco-environment field, but with shortcomings in revealing the laws of natural science, such as insufficient generalization ability and poor interpretability. In order to overcome these shortages and tap into complementary advantages, coupling AI and eco-environmental models containing physical mechanism has been a new research method with fast development in recent years. We introduced the classifications of AI used in eco-environmental field, outlined its applications, and mainly illustrated the progresses, status and inadequacies for the coupling research. Based on all the summaries, we proposed a new coupling method of physical mechanism and AI for reconstructing mechanism processes, followed by analyses of theoretical significance of partial parameters, feasibility of better generalization and interpretability, as well as prospection of imitating physical mechanism. At the end of the review, we discussed the trend of the coupling method of AI and eco-environment models.

Key words: artificial intelligence, eco-environmental model, coupling, reconstruction of mechanism.

HU Yu-cong, LI Na, JIANG Yan, BAO Xin, LI Xu-yong. Research progress on coupling artificial intelligence and eco-environmental models[J]. Chinese Journal of Applied Ecology, 2023, 34(1): 257-263.

References

[1] de Bézenac E, Pajot A, Gallinari P. Deep learning for physical processes: Incorporating prior scientific know-ledge. Journal of Statistical Mechanics: Theory and Experiment, 2019, 2019: 124009
[2] 刘思峰, 杨英杰, 吴利丰. 灰色系统理论及其应用. 第7版. 北京: 科学出版社, 2014: 3
[3] Nearing GS, Gupta HV. The quantity and quality of information in hydrologic models. Water Resources Research, 2015, 51: 524-538
[4] Blöschl G, Bierkens MFP, Chambell A, et al. Twenty-three unsolved problems in hydrology (UPH): A community perspective. Hydrological Sciences Journal, 2019, 64: 1141-1158
[5] 周信哲, 陈永娟, 蒲文鹏. 机器学习在水质预测中的应用. 2020中国环境科学学会科学技术年会, 南京, 2020: 3948-3953
[6] Nearing GS, Kratzert F, Sampson AK, et al. What role does hydrological science play in the age of machine learning? Water Resources Research, 2021, 57: e2020WR028091
[7] Reichstein M, Camps-Valls G, Stevens B, et al. Deep learning and process understanding for data-driven Earth system science. Nature, 2019, 566: 195-204
[8] Sun AY, Scanlon BR. How can Big Data and machine learning benefit environment and water management: A survey of methods, applications, and future directions. Environmental Research Letters, 2019, 14: 073001
[9] Liu ZL, Peng CH, Xiang WH, et al. Application of artificial neural networks in global climate change and ecological research: An overview. Chinese Science Bulletin, 2010, 55: 3853-3863
[10] Makridakis S, Spiliotis E, Assimakopoulos V. Statistical and machine learning forecasting methods: Concerns and ways forward. PLoS One, 2018, 13(3): e0194889
[11] 方精云. 生态学学科体系的再构建. 大学与学科, 2021, 2(4): 61-73
[12] Hubbard SS, Varadharajan C, Wu YX, et al. Emerging technologies and radical collaboration to advance predictive understanding of watershed hydrobiogeochemistry. Hydrological Processes, 2020, 34: 3175-3182
[13] Mital U, Dipankar D, Brown JB, et al. Sequential imputation of missing spatio-temporal precipitation data using random forests. Frontiers in Water, 2020, 2: 20
[14] Nguyen DH, Bae DH. Correcting mean areal precipita-tion forecasts to improve urban flooding predictions by using long short-term memory network. Journal of Hydrology, 2020, 584: 124710
[15] Cubuk ED, Zoph B, Mane D, et al. Autoaugment: Learning augmentation strategies from data. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, CA, USA, 2019: 113-123
[16] Shorten C, Khoshgoftaar TM. A survey on image data augmentation for deep learning. Journal of Big Data, 2019, 6: 60
[17] 陶阳明. 经典人工智能算法综述. 软件导刊, 2020, 19(3): 276-280
[18] Yahya Y, Ismail R. Using uncertainty of Bayesian theorem to predict mortality of tree in forest growth simulation system. Proceedings of the 12th International Conference on Ubiquitous Information Management and Communication, New York, USA, 2018: 1-5
[19] 廖杰, 王文圣, 丁晶. 贝叶斯公式在河流水质综合评价中的应用. 四川师范大学学报: 自然科学版, 2007, 30(4): 519-522
[20] 闫香慧, 赵志模, 刘怀, 等. 应用马尔可夫链理论对褐飞虱和白背飞虱发生程度的预测. 生态学报, 2009, 29(11): 5799-5806
[21] 薛亮, 任志远. 基于空间马尔科夫链的关中地区生态安全时空演变分析. 生态环境学报, 2011, 20(1): 114-118
[22] 何丹, 金凤君, 周璟. 基于Logistic-CA-Markov的土地利用景观格局变化——以京津冀都市圈为例. 地理科学, 2011, 31(8): 903-910
[23] Chen C, He W, Zhou H, et al. A comparative study among machine learning and numerical models for simulating groundwater dynamics in the Heihe River Basin, northwestern China. Scientific Reports, 2020, 10: 3904
[24] Liu R, Gu Y, Shen M, et al. Predicting postmortem interval based on microbial community sequences and machine learning algorithms. Environmental Microbiology, 2020, 22: 2273-2291
[25] 毕温凯, 袁兴中, 唐清华, 等. 基于支持向量机的湖泊生态系统健康评价研究. 环境科学学报, 2012, 32(8): 1984-1990
[26] Wei L, Yuan Z, Zhong Y, et al. An improved gradient boosting regression tree estimation model for soil heavy metal (Arsenic) pollution monitoring using hyperspectral remote sensing. Applied Sciences, 2019, 9: 1943
[27] Guio Blanco CM, Brito Gometz VM, Crespo P, et al. Spatial prediction of soil water retention in a Páramo landscape: Methodological insight into machine learning using random forest. Geoderma, 2018, 316: 100-114
[28] 张文强, 罗格平, 郑宏伟, 等. 基于随机森林模型的内陆干旱区植被指数变化与驱动力分析: 以北天山北坡中段为例. 植物生态学报, 2020, 44(11): 1113-1126
[29] 方馨蕊, 温兆飞, 陈吉龙, 等. 随机森林回归模型的悬浮泥沙浓度遥感估算. 遥感学报, 2019, 23(4): 756-772
[30] 李德, 陈文涛, 乐章燕, 等. 基于随机森林算法和气象因子的砀山酥梨始花期预报. 农业工程学报, 2020, 36(12): 143-151
[31] 芮孝芳, 蒋成煜, 张金存. 流域水文模型的发展. 水文, 2006, 26(3): 22-26
[32] Sit M, Demiray BZ, Xiang ZR, et al. A comprehensive review of deep learning applications in hydrology and water resources. Water Science and Technology, 2020, 82: 2635-2670
[33] Kang L. Wave monitoring based on improved convolution neural network. Journal of Coastal Research, 2019, 94: 186-190
[34] Duan S, Ullrich P, Shu L. Using convolutional neural networks for streamflow projection in California. Frontiers in Water, 2020, 2: 28
[35] Laloy E, Hérault R, Jacques D, et al. Training-image based geostatistical inversion using a spatial generative adversarial neural network. Water Resources Research, 2018, 54: 381-406
[36] Fan H, Jiang M, Xu L, et al. Comparison of long short term memory networks and the hydrological model in runoff simulation. Water, 2020, 12: 175
[37] Mouatadid S, Adamowski JF, Tiwari MK, et al. Coupling the maximum overlap discrete wavelet transform and long short-term memory networks for irrigation flow forecasting. Agricultural Water Management, 2019, 219: 72-85
[38] Kratzert F, Klotz D, Herrnegger M, et al. Toward improved predictions in ungauged basins: Exploiting the power of machine learning. Water Resources Research, 2019, 55: 11344-11354
[39] Eisenstein M. Big data: The power of petabytes. Nature, 2015, 527: S2-S4
[40] Lazer D, Kennedy R, King G, et al. The parable of Google Flu: Traps in big data analysis. Science, 2014, 343: 1203-1205
[41] Li ZJ, Kan GY, Yao C, et al. Improved neural network model and its application in hydrological simulation. Journal of Hydrological Engineering, 2014, 19: 04014019
[42] Yang T, Sun FB, Gentine P, et al. Evaluation and machine learning improvement of global hydrological model-based flood simulations. Environmental Research Letters, 2019, 14: 114027
[43] Noori N, Kalin L. Coupling SWAT and ANN models for enhanced daily streamflow prediction. Journal of Hydro-logy, 2016, 533: 141-151
[44] Rahmani F, Lawson K, Appling A, et al. Process lear-ning of stream temperature modelling using deep learning and big data. AGU 2021 Fall Meeting, New Orleans, LA, USA, 2021: 1
[45] Roweis S, Ghahramani Z. A unifying review of linear Gaussian models. Neural Computation, 1999, 11: 305-345
[46] Pelissier C, Frame J, Nearing G. Combining Parametric Land Surface Models with Machine Learning [EB/OL]. (2020-12-21) [2022-01-01]. https://arxiv.org/ftp/arxiv/papers/2002/2002.06141.pdf
[47] Karpatne A, Atluri G, Faghmous JH, et al. Theory-guided data science: A new paradigm for scientific discovery from data. IEEE Transactions on Knowledge and Data Engineering, 2017, 29: 2318-2331
[48] Ma J, Ding YX, Chen JCP, et al. A temporal-spatial interpolation and extrapolation method based on geographic long short-term memory neural network for PM_2.5. Journal of Cleaner Production, 2019, 237: 117729
[49] Daw A, Thomas RQ, Carley CC, et al. Physics-guided architecture (PGA) of neural networks for quantifying uncertainty in lake temperature modeling. Proceedings of the 2020 SIAM International Conference on Data Mining, Cincinnati, OH, USA, 2020: 532-540
[50] Hoedt PJ, Kratzert F, Klotz D, et al. MC-LSTM: Mass-conserving LSTM. International Conference on Machine Learning, PMLR, Chengdu, China, 2021: 4275-4286
[51] Zhao WL, Gentine P, Reichstein M, et al. Physics-constrained machine learning of evapotranspiration. Geophysical Research Letters, 2019, 46: 14496-14507
[52] Raissi M, Perdikaris P, Karniadakis GE. Physics Informed Deep Learning (part Ⅰ): Data-driven Solutions of Nonlinear Partial Differential Equations [EB/OL]. (2017-12-12) [2021-12-12]. https://doi.org/10.48550/arXiv.1711.10561
[53] Mo W, Luo X, Zhong Y, et al. Image recognition using convolutional neural network combined with ensemble learning algorithm. Journal of Physics: Conference Series, 2019, 1237: 022026
[54] Iten R, Metger T, Wilming H, et al. Discovering physical concepts with neural networks. Physical Review Letters, 2020, 124: 010508
[55] Chauvet JM. The 30-year Cycle in the AI Debate [EB/OL]. (2018-12-12) [2021-12-12]. https://doi.org/10.48550/arXiv.1810.04053
[56] Camps-Valls G, Tuia D, Zhu XX, et al. Deep Learning for the Earth Sciences [EB/OL]. (2021-12-12) [2022-01-01]. https://doi.org/10.1002/9781119646181.ch19
[57] Hornik K, Stinchcombe M, White H. Multilayer feed forward networks are universal approximators. Neural Networks, 1989, 2: 359-366