Using Machine Learning to Explore Changes in Atmospheric Water Vapor Stable Isotopes: A Case Study at Muztagh Ata, Western Tibetan Plateau

Abstract

Atmospheric water vapor stable isotopes (δ¹⁸O and δD) are an important proxy to track the water cycle, changes of historical temperature and humidity, and characterize the monsoon intensity. Isotope general circulation models (iGCMs) are the main method for water vapor stable isotopes (Vapor_iso) simulation, but the simulations from iGCMs are difficult to investigate Vapor_iso dynamics at meso and sub-mesoscales due to the low spatial resolution and uncertainties of parameterization. Here, we present how machine learning can be a powerful tool in representing changes of daily Vapor_iso, combined with the high-resolution Vapor_iso observed at Muztagh Ata, western Tibetan Plateau (TP), from January 2022 to February 2023, where is dominated by the Westerlies. Different machine learning methods (random forest, BP neural network, support vector machine) and different parameter optimization schemes (genetic algorithm, particle swarm optimization) are used to represent the Vapor_iso and corresponding meteorological conditions. We also compared simulations from machine learning and IsoGSM with the observed Vapor_iso. The root-mean-square error (RMSE) of δ¹⁸O and δD simulated by machine learning is smaller than IsoGSM, and the correlation coefficient (R) is much larger than IsoGSM. We suggest that the daily and monthly changes of Vapor_iso are captured when the machine learning is trained on the meteorological elements and isotopic dataset. It is also potentially useful for the accurate simulation from iGCMs by using machine learning parameterization.

Keywords

water vapor stable isotopes / machine learning / isotope general circulation models / genetic algorithm / particle swarm optimization / water cycle

References

[1] Bao H M, Lyons J R, Zhou C M. Triple Oxygen Isotope Evidence for Elevated CO₂ Levels after a Neoproterozoic Glaciation. Nature, 2008, 453(7194): 504-506

[2] Breiman L. Random Forests. Machine Learning, 2001, 45(1): 5-32

[3] Cao X B, Bao H M. Dynamic Model Constraints on Oxygen-17 Depletion in Atmospheric O₂ after a Snowball Earth. Proceedings of the National Academy of Sciences of the United States of America, 2013, 110(36): 14546-14550

[4] Chen X B. A Study on Pairs Trading Based on Support Vector Machine and Convolution Neural Network, 2019, Wuhan, Wuhan University(in Chinese with English Abstract)

[5] Cheng H, Zhang H W, Cai Y Jet al.. Orbital-Scale Asian Summer Monsoon Variations: Paradox and Exploration. Science China Earth Sciences, 2021, 64(4): 529-544

[6] Conroy J L, Cobb K M, Noone D. Comparison of Precipitation Isotope Variability across the Tropical Pacific in Observations and SWING2 Model Simulations. Journal of Geophysical Research: Atmospheres, 2013, 118(11): 5867-5892

[7] Cortes C, Vapnik V. Support-Vector Networks. Machine Learning, 1995, 20: 273-297

[8] Crockford P W, Hayles J A, Bao H Met al.. Triple Oxygen Isotope Evidence for Limited Mid-Proterozoic Primary Productivity. Nature, 2018, 559(7715): 613-616

[9] Dai D, Gao J, Steen-Larsen H Cet al.. Continuous Monitoring of the Isotopic Composition of Surface Water Vapor at Lhasa, Southern Tibetan Plateau. Atmospheric Research, 2021, 264: 105827

[10] Delattre H, Vallet-Coulomb C, Sonzogni C. Deuterium Excess in the Atmospheric Water Vapour of a Mediterranean Coastal Wetland: Regional vs. Local Signatures. Atmospheric Chemistry and Physics, 2015, 15: 10167-10181

[11] Ding G Q, Chen S Q, Sun Y Het al.. Holocene Hydroclimatic Variations in the Asian Drylands: Current Understanding and Future Perspectives. Journal of Earth Science, 2024, 35(1): 292-295

[12] Evans R E, Bhaskar M K, Sukachev D Det al.. Photon-Mediated Interactions between Quantum Emitters in a Diamond Nanocavity. Science, 2018, 362(6415): 662-665

[13] Fiorella R P, Bares R, Lin J Cet al.. Wintertime Decoupling of Urban Valley and Rural Ridge Hydrological Processes Revealed through Stable Water Isotopes. Atmospheric Environment, 2019, 213: 337-348

[14] Frankenberg C, Yoshimura K, Warneke Tet al.. Dynamic Processes Governing Lower-Tropospheric HDO/H₂O Ratios as Observed from Space and Ground. Science, 2009, 325(5946): 1374-1377

[15] Graf P, Wernli H, Pfahl Set al.. A New Interpretative Framework for Below-Cloud Effects on Stable Water Isotopes in Vapour and Rain. Atmospheric Chemistry and Physics, 2019, 19(2): 747-765

[16] Han L F, Gröning M, Aggarwal Pet al.. Reliable Determination of Oxygen and Hydrogen Isotope Ratios in Atmospheric Water Vapour Adsorbed on 3A Molecular Sieve. Rapid Communications in Mass Spectrometry, 2006, 20(23): 3612-3618

[17] Holland J H. Genetic Algorithms. Scientific American, 1992, 267(1): 66-72

[18] Jhansi G, Sujatha K. HRFSVM: Identification of Fish Disease Using Hybrid Random Forest and Support Vector Machine. Environmental Monitoring and Assessment, 2023, 195(8): 918

[19] Joussaume S, Sadourny R, Jouzel J. A General Circulation Model of Water Isotope Cycles in the Atmosphere. Nature, 1984, 311598124-29

[20] Killingsworth B A, Hayles J A, Zhou C Met al.. Sedimentary Constraints on the Duration of the Marinoan Oxygen-17 Depletion (MOSD) Event. Proceedings of the National Academy of Sciences of the United States of America, 2013, 110(44): 17686-17690

[21] Kennedy J, Eberhart R. Particle Swarm Optimization. Proceedings of ICNN’95. International Conference on Neural Networks, Perth, WA, Australia, 199519421948

[22] Lacour J-L, Risi C, Worden Jet al.. Importance of Depth and Intensity of Convection on the Isotopic Composition of Water Vapor as Seen from IASI and TES δD Observations. Earth and Planetary Science Letters, 2018, 481: 387-394

[23] Lei S J. Stable Hydrogen and Oxygen Isotopic Compositions in Near-Surface Water Vapor and Their Simulations in the Lanzhou City, 2022, Lanzhou, Northwest Normal University(in Chinese with English Abstract)

[24] Li Y, Yu W S, Tian L Det al.. Diagnosis of Atmospheric Circulation Shifts in the Central Tibetan Plateau: Evidence from Stable Isotopes. Atmospheric Research, 2023, 282: 106536

[25] Lin J, Wang Z, Zhang Y. Spatial Distribution Characteristics of Stable Isotopes in River Water in Hainan Island: Implication for Water Vapor Delivery Path. Earth Science, 2024, 49(12): 4622-4634(in Chinese with English Abstract)

[26] Liu J F, Ding M H, Xiao C D. Review on Atmospheric Water Vapor Isotopic Observation and Research: Theory, Method and Modeling. Progress in Geography, 2015, 34: 340-353(in Chinese with English Abstract)

[27] Liu L, Weng C S, Li S Let al.. Passive Remote Sensing of Ice Cloud Properties at Terahertz Wavelengths Based on Genetic Algorithm. Remote Sensing, 2021, 13(4): 735

[28] Liu X T. Research on the Forecast of Railway Freight Volume Based on BP Neural Network Optimized by Genetic Algorithm and the Influence Factors, 2020, Beijing, Beijing Jiaotong University(in Chinese with English Abstract)

[29] Liu Y J, Man W M, Zhou T Jet al.. A Tracing Study on Influence Factors of East Asian Stable Isotopes in Atmospheric Water Vapor. Chinese Journal of Atmospheric Sciences, 2023, 473616-630(in Chinese with English Abstract)

[30] Liu Z F, Yoshimura K, Kennedy C Det al.. Water Vapor δD Dynamics over China Derived from SCIAMACHY Satellite Measurements. Science China Earth Sciences, 2014, 57(4): 813-823

[31] Luo S Y, Xu T J, Wei S J. Prediction Method and Application of Shale Reservoirs Core Gas Content Based on Machine Learning. Journal of Applied Geophysics, 2022, 204: 104741

[32] Midhun M, Ramesh R. Validation of δ¹⁸O as a Proxy for Past Monsoon Rain by Multi-GCM Simulations. Climate Dynamics, 2016, 46(5): 1371-1385

[33] Nan Y, He Z H, Tian F Qet al.. Can we Use Precipitation Isotope Outputs of Isotopic General Circulation Models to Improve Hydrological Modeling in Large Mountainous Catchments on the Tibetan Plateau. Hydrology and Earth System Sciences, 2021, 25(12): 6151-6172

[34] Nguyen D H, Le X H, Anh D Tet al.. Hourly Streamflow Forecasting Using a Bayesian Additive Regression Tree Model Hybridized with a Genetic Algorithm. Journal of Hydrology, 2022, 606: 127445

[35] Nimya S S, Sengupta S, Parekh Aet al.. Region-Specific Performances of Isotope Enabled General Circulation Models for Indian Summer Monsoon and the Factors Controlling Isotope Biases. Climate Dynamics, 2022, 59(11): 3599-3619

[36] Norouzi H, Bazargan J. Calculation of Water Depth during Flood in Rivers Using Linear Muskingum Method and Particle Swarm Optimization (PSO) Algorithm. Water Resources Management, 2022, 36(11): 4343-4361

[37] Oza H, Ganguly A, Padhya Vet al.. Hydrometeorological Processes and Evaporation from Falling Rain in Indian Sub-Continent: Insights from Stable Isotopes and Meteorological Parameters. Journal of Hydrology, 2020, 591: 125601

[38] Pereira F F, Mendes T S G, Simões S J Cet al.. Comparison of LiDAR- and UAV-Derived Data for Landslide Susceptibility Mapping Using Random Forest Algorithm. Landslides, 2023, 20(3): 579-600

[39] Rangarajan R, Laskar A H, Bhattacharya S Ket al.. An Insight into the Western Pacific Wintertime Moisture Sources Using Dual Water Vapor Isotopes. Journal of Hydrology, 2017, 547: 111-123

[40] Risi C, Noone D, Frankenberg Cet al.. Role of Continental Recycling in Intraseasonal Variations of Continental Moisture as Deduced from Model Simulations and Water Vapor Isotopic Measurements. Water Resources Research, 2013, 49(7): 4136-4156

[41] Risi C, Noone D, Worden J. Process-Evaluation of Tropospheric Humidity Simulated by General Circulation Models Using Water Vapor Isotopologues: 1. Comparison between Models and Observations. Journal of Geophysical Research: Atmospheres, 2012, 117(D5): D05303

[42] Rumelhart D E, Hinton G E, Williams R J. Learning Representations by Back-Propagating Errors. Nature, 1986, 3236088533-536

[43] Saha S, Saha A, Roy Bet al.. Integrating the Particle Swarm Optimization (PSO) with Machine Learning Methods for Improving the Accuracy of the Landslide Susceptibility Model. Earth Science Informatics, 2022, 15(4): 2637-2662

[44] Samuels-Crow K E, Galewsky J, Sharp Z Det al.. Deuterium Excess in Subtropical Free Troposphere Water Vapor: Continuous Measurements from the Chajnantor Plateau, Northern Chile. Geophysical Research Letters, 2014, 41(23): 8652-8659

[45] Sansjofre P, Ader M, Trindade R I Fet al.. A Carbon Isotope Challenge to the Snowball Earth. Nature, 2011, 478(7367): 93-96

[46] Santos C L, Masson-Delmotte V, Casado Met al.. A 4.5 Year-Long Record of Svalbard Water Vapor Isotopic Composition Documents Winter Air Mass Origin. Journal of Geophysical Research: Atmospheres, 2020, 125(23): e2020JD032681

[47] Schmidt G A, Hoffmann G, Shindell D Tet al.. Modeling Atmospheric Stable Water Isotopes and the Potential for Constraining Cloud Processes and Stratosphere-Troposphere Water Exchange. Journal of Geophysical Research: Atmospheres, 2005, 110D212005JD005790

[48] Shang B J, Gao J, Chen Get al.. Stable Isotopes in Atmospheric Water Vapour: Patterns, Mechanisms and Perspectives. Science China Earth Sciences, 2024, 67123789-3813

[49] Shang B J, Zhu G F, Li Tet al.. Global Precipitation D-Excess Dataset: A Critical Resource for Geographical Science Research. Geoscience Data Journal, 2026, 13: e70043

[50] Shi X Y, Pu T, He Y Q. Spatial Distribution of Stable Isotope from the Lakes in Typical Temperate Glacier Region. Environmental Science, 2016, 37(5): 1685-1691(in Chinese with English Abstract)

[51] Shi X Y, Risi C, Pu Tet al.. Variability of Isotope Composition of Precipitation in the Southeastern Tibetan Plateau from the Synoptic to Seasonal Time Scale. Journal of Geophysical Research: Atmospheres, 2020, 125(6): e2019JD031751

[52] Sjolte J, Hoffmann G, Johnsen S Jet al.. Modeling the Water Isotopes in Greenland Precipitation 1959–2001 with the Meso-Scale Model REMO-Iso. Journal of Geophysical Research, 2011, 116D18D18105

[53] Song M Y, Li Z Q, Wang F Tet al.. Influence of Sub-Cloud Secondary Evaporation Effects on the Stable Isotopes in Precipitation of Urumqi Glacier No. 1, Eastern Tianshan. Journal of Earth Science, 2024, 351177-189

[54] Steen-Larsen H C, Johnsen S J, Masson-Delmotte Vet al.. Continuous Monitoring of Summer Surface Water Vapor Isotopic Composition above the Greenland Ice Sheet. Atmospheric Chemistry and Physics, 2013, 13(9): 4815-4828

[55] Steen-Larsen H C, Masson-Delmotte V, Sjolte Jet al.. Understanding the Climatic Signal in the Water Stable Isotope Records from the NEEM Shallow Firn/Ice Cores in Northwest Greenland. Journal of Geophysical Research, 2011, 116(D6): D06108

[56] Steen-Larsen H C, Risi C, Werner Met al.. Evaluating the Skills of Isotope-Enabled General Circulation Models Against in situ Atmospheric Water Vapor Isotope Observations. Journal of Geophysical Research: Atmospheres, 2017, 1221246-263

[57] Steen-Larsen H C, Sveinbjörnsdottir A E, Peters A Jet al.. Climatic Controls on Water Vapor Deuterium Excess in the Marine Boundary Layer of the North Atlantic Based on 500 Days of in situ, Continuous Measurements. Atmospheric Chemistry and Physics, 2014, 14(15): 7741-7756

[58] Strong M, Sharp Z D, Gutzler D S. Diagnosing Moisture Transport Using D/H Ratios of Water Vapor. Geophysical Research Letters, 2007, 34(3): 2006GL028307

[59] Tada M, Yoshimura K, Toride K. Improving Weather Forecasting by Assimilation of Water Vapor Isotopes. Scientific Reports, 2021, 11: 18067

[60] Thompson L G, Yao T D, Davis M Eet al.. Ice Core Records of Climate Variability on the Third Pole with Emphasis on the Guliya Ice Cap, Western Kunlun Mountains. Quaternary Science Reviews, 2018, 188: 1-14

[61] Tian J W, Liu Y, Zheng W Fet al.. Smog Prediction Based on the Deep Belief: BP Neural Network Model (DBN-BP). Urban Climate, 2022, 41: 101078

[62] Tindall J C, Valdes P J, Sime L C. Stable Water Isotopes in HadCM3: Isotopic Signature of El Niño-Southern Oscillation and the Tropical Amount Effect. Journal of Geophysical Research: Atmospheres, 2009, 114(D4): 2008JD010825

[63] Wang L, Yu Z W, Zhang Y Jet al.. Review of Machine Learning Methods Applied to Enhanced Geothermal Systems. Environmental Earth Sciences, 2023, 82369

[64] Wang S J, Jiao R, Zhang M Jet al.. Changes in Below-Cloud Evaporation Affect Precipitation Isotopes during Five Decades of Warming across China. Journal of Geophysical Research: Atmospheres, 2021, 1267e2020JD033075

[65] Wang X J. Simulation and Comparison of Stable Isotopic GCM in Global Precipitation, 2018, Changsha, Hunan Normal University(in Chinese with English Abstract)

[66] Wen X F, Zhang S C, Sun X Met al.. Water Vapor and Precipitation Isotope Ratios in Beijing, China. Journal of Geophysical Research: Atmospheres, 2010, 115D12009JD012408

[67] Wei Z W, Yoshimura K, Okazaki Aet al.. Understanding the Variability of Water Isotopologues in Near-Surface Atmospheric Moisture over a Humid Subtropical Rice Paddy in Tsukuba, Japan. Journal of Hydrology, 2016, 533: 91-102

[68] Wolf A, Roberts W H G, Ersek Vet al.. Rainwater Isotopes in Central Vietnam Controlled by Two Oceanic Moisture Sources and Rainout Effects. Scientific Reports, 2020, 10: 16482

[69] Worden J, Bowman K, Noone D. Tropospheric Emission Spectrometer Observations of the Tropospheric HDO/H₂O Ratio: Estimation Approach and Characterization. Journal of Geophysical Research: Atmospheres, 2006, 111: D16309

[70] Xie F, Tian W S, Zhou Xet al.. Increase in Lower Stratospheric Water Vapor in the Past 100 Years Related to Tropical Atlantic Warming. Geophysical Research Letters, 2020, 47(22): e2020GL090539

[71] Xie L C. Research on Data Mining Technology Based on Improved BP Neural Network and Its Application in the Law Analysis of Coal Thermal Conversion Data, 2021, Xi’an, Northwest University(in Chinese with English Abstract)

[72] Yang Y, Yang R W, Cao Jet al.. Relationship between the Asian Summer Monsoon Circulation and Speleothem δ¹⁸O of Xiaobailong Cave. Climate Dynamics, 2019, 5396351-6362

[73] Yao T D, Guo X J, Thompson Let al.. δ¹⁸O Record and Temperature Change over the Past 100 Years in Ice Cores on the Tibetan Plateau. Science in China Series D, 2006, 49(1): 1-9

[74] Yoshimura K, Frankenberg C, Lee Jet al.. Comparison of an Isotopic Atmospheric General Circulation Model with New Quasi-Global Satellite Measurements of Water Vapor Isotopologues. Journal of Geophysical Research, 2011, 116(D19): D19118

[75] Yoshimura K, Kanamitsu M, Noone Det al.. Historical Isotope Simulation Using Reanalysis Atmospheric Data. Journal of Geophysical Research: Atmospheres, 2008, 113D192008JD010074

[76] Yu X X. Spatio-Temporal Characteristics and Influencing Factors of the Proportion of Recycled Moisture during Summer Half Year in the Arid Region of Northwest China Based on LMDZiso, 2019, Lanzhou, Northwest Normal University(in Chinese with English Abstract)

[77] Yu W D, Dessler A E, Park Met al.. Influence of Convection on Stratospheric Water Vapor in the North American Monsoon Region. Atmospheric Chemistry and Physics, 2020, 20: 12153-12161

[78] Yuan X, Wang Y M, Ji Pet al.. A global Transition to Flash Droughts under Climate Change. Science, 2023, 380: 187-191

[79] Zakharov V I, Imasu R, Gribanov K Get al.. Latitudinal Distribution of the Deuterium to Hydrogen Ratio in the Atmospheric Water Vapor Retrieved from IMG/ADEOS Data. Geophysical Research Letters, 2004, 31(12): 2004GL019433

[80] Zhang H W, Zhang X, Cai Y Jet al.. A Data-Model Comparison Pinpoints Holocene Spatiotemporal Pattern of East Asian Summer Monsoon. Quaternary Science Reviews, 2021, 261: 106911

[81] Zhang X P, Guan H D, Zhang X Z. Comparisons and Assessment on Stable Isotopic Effects in Precipitation Simulated by Different Models. Quaternary Sciences, 2016, 36: 1343-1357(in Chinese with English Abstract)

[82] Zhang X P, Sun Z A, Guan H Det al.. GCM Simulations of Stable Isotopes in the Water Cycle in Comparison with GNIP Observations over East Asia. Acta Meteorologica Sinica, 2012, 26(4): 420-437

[83] Zhang Z C, Zeng X M, Li Get al.. Summer Precipitation Forecast Using an Optimized Artificial Neural Network with a Genetic Algorithm for Yangtze-Huaihe River Basin, China. Atmosphere, 2022, 13(6): 929

[84] Zhou J, Zhu S L, Qiu Y Get al.. Predicting Tunnel Squeezing Using Support Vector Machine Optimized by Whale Optimization Algorithm. Acta Geotechnica, 2022, 1741343-1366

[85] Zhu J J, Yang M Q, Ren Z J. Machine Learning in Environmental Research: Common Pitfalls and Best Practices. Environmental Science & Technology, 2023, 57(46): 17671-17689

RIGHTS & PERMISSIONS

China University of Geosciences (Wuhan) and Springer-Verlag GmbH Germany, Part of Springer Nature