RECONSTRUCTION OF ATMOSPHERIC SURFACE HUMIDITY OVER THE OCEAN FROM RELEVANT METEOROLOGICAL MEASUREMENTS USING MACHINE LEARNING METHODS

  • S. A. Vostrikova Moscow Institute of Physics and Technology
  • M. A. Krinitsky Moscow Institute of Physics and Technology; Shirshov Institute of Oceanology, Russian Academy of Sciences
  • S. K. Gulev Shirshov Institute of Oceanology, Russian Academy of Sciences
  • M. P. Alexandrova Shirshov Institute of Oceanology, Russian Academy of Sciences
DOI 10.29006/1564-2291.JOR-2024.53(2).7
Keywords relative humidity, data reconstruction, machine learning, regression, linear regression, decision tree, random forest, gradient boosting, multilayer perceptron

Abstract

Air humidity in the near-surface layer of the atmosphere over the ocean is a key climate parameter that has a significant impact on the processes of moisture and heat transfer between the ocean and the atmosphere, as well as on the dynamics of atmospheric processes in general. Analysis of meteorological data collected during the 20th century shows the sparseness of humidity measurement series in space and time. The International Ocean and Atmosphere Data Set (ICOADS) indicates an insufficient density of measurements in the early 20th century compared to later periods, which creates difficulties for adequate analysis of climate trends in relative humidity. Methods for approximating humidity time series presented in the literature often demonstrate limited accuracy, based mainly on statistical and heuristic approaches. Our work is aimed at improving the quality of solving this problem through the use of machine learning methods. As a first, simplest approach, we solved the problem in the formulation of the approximation of relative humidity based on the data of accompanying measurements of atmospheric pressure, air temperature, wind speed and direction, ocean surface temperature, as well as observations of the amount and types of clouds at three tiers. In addition, the accompanying variables include the WMO weather code and the estimated solar altitude. The study used the following types of machine learning models: linear regression, decision tree, random forest, gradient boosting, and fully connected artificial neural network. To improve the territorial and temporal specificity of the developed models, we conducted a study for each 2-degree square and each season separately. The scikit-learn library and the package implementing the CatBoost model were used to train and apply the machine learning models. For each type of model, we optimized the hyperparameters using the Optuna Bayesian optimization library. Based on the results obtained, maps of the spatial distribution of model errors were constructed, which made it possible to identify regions with high and low accuracy of humidity approximation. The study confirmed the effectiveness of machine learning methods for reconstructing climate series, identified the most suitable models for this task, and outlined promising areas for further work.

References


  1. Adele Cutler, D. Richard Cutler, and R. John, 2012: Stevens Random Forests J. Springer. Ensemble Machine Learning. https://doi.org/10.1007/978-1-4419-9326-7_5.

  2. Aleksandrova, M. P. and S. K. Gulyov, 2017: Rekonstrukciya dolgoperiodnoj izmenchivosti oblachnosti i radiacionnyh potokov nad mirovym okeanom (Reconstruction of long-term variability of cloudiness and radiation fluxes over the world ocean). Dis. na kand. fiziko-matematicheskih nauk. 01.06.2017. Moscow, IO RAN.

  3. Barry de Ville, 2013: Decision trees J. WIREs Computational Statistics. 5 (6), https://doi.org/10.1002/wics.1278.

  4. Callum J. Shakespeare and Michael L. Roderick, 2024: What Controls Near-Surface Relative Humidity Over the Ocean J. James. Journal of Advances in Modeling Earth System. https://doi.org/10.1029/2023MS004168.

  5. Carsten Abraham and Colin Goldblatt, 2023: Changes in Relative Humidity Profiles over Earth’s Oceans in a Warming Climate: A Satellite-Data-Based Inference. Journal of the Atmospheric Sciences. https://doi.org/10.1175/JAS-D-22-0119.1.

  6. Efron, B., 1979: Bootstrap Methods: Another Look at the Jackknife. The Annals of Statistics, 1 (7), 1–26.

  7. Freeman, Eric, Scott D. Woodruff, Steven J. Worley, Sandra J. Lubker, Elizabeth C. Kent, William E. Angel, David I. Berry, Philip Brohan, Ryan Eastman, Lydia Gates, Wolfgang Gloeden, Zaihua Ji, Jay Lawrimore, Nick A. Rayner, Gudrun Rosenhagen, and Shawn R. Smith, 2017: ICOADS Release 3.0: a major update to the historical marine climate record. International Journal of Climatology, 5 (37), https://doi.org/10.1002/joc.4775.

  8. Hancock, J. T. and T. M. Khoshgoftaar, 2020: CatBoost for big data: an interdisciplinary review. J. Big Data 7, https://doi.org/10.1186/s40537-020-00369-8.

  9. Hussain Almarzooq, Umair bin Waheed, 2024: Automating hyperparameter optimization in geophysics with Optuna: A comparative study. J. Geophysical Prospecting, https://doi.org/10.1111/1365-2478.13484.

  10. Jiajun Guo, Liang Zhang, and Ruqiang Guo, 2023: Relative humidity prediction with covariates and error correction based on SARIMA-EG-ECM model J. Springer Nature. Modeling Earth Systems and Environment, https://doi.org/10.1007/s40808-023-01738-x.

  11. John, R., 2008: Borchert Regional Differences in the World Atmospheric Circulation J. Annals of the Association of American Geographers, https://doi.org/10.2307/2561080.

  12. Luminda Niroshana Gunawardhana, Ghazi A. Al-Rawas, So Kazama, 2017: An alternative method for predicting relative humidity for climate change studies. J. Meteorological application, https://doi.org/10.1002/met.1641.

  13. Marvin Minsky, Seymour Papert, 1969: Perceptrons: An Introduction to Computational Geometry The MIT Press.

  14. Matthieu Parizy, Norihiro Kakuko, Nozomu Togawa Fast Hyperparameter, 2023: Tuning for Ising Machines 2023 IEEE International Conference on Consumer Electronics (ICCE), https://doi.org/10.1109/ICCE56470.2023.10043382.

  15. Takuya Akiba, Shotaro Sano, Toshihiko Yanase, Takeru Ohta, Masanori Kayama Optuna, 2019.: A Next-generation Hyperparameter Optimization Framework. Conference: the 25th ACM SIGKDD International Conference. https://doi.org/10.48550/arXiv.1907.10902.

Published
2025-06-30
Section
Ocean physics and atmosphere