Keywords: data recovery, groundwater level, artificial neural networks, time series of values, time components of level fluctuations, wavelet analysis, forecasting, cycle

The cessation of regular monitoring of groundwater levels in Ukraine prompts the search for methods for reproducing and predicting the level, which will allow estimating groundwater flow rates, creating models of groundwater resource formation and moisture balance in watersheds. Artificial neural networks (ANN) of various architectures are considered as a data recovery tool for further modeling of water resources. In order to determine the optimal ANN architecture that can simulate the groundwater level (GWL) trend and provide forecasts, the effectiveness of different neural networks (RBF and MLP) in predicting the monthly average GWL was investigated. To select the optimal ANN configuration and assess the effectiveness of each network and its ability to make accurate predictions, the following methods and criteria were used: multiple correlation analysis, spectral analysis of Fourier transforms, wavelet analysis, and component separation by the duration of oscillation cycles. The forecast was made for the average monthly groundwater level from one of the few wells in the Western Bug River basin, for which observations were stopped back in June 2011. The most realistic results using ANN were obtained after isolating short-, medium-, and long-term components in the GWL fluctuations and performing forecasts for the last two components, which is a pioneering step for hydrogeological observations in Ukraine. If for the full (undivided) series of input data it is possible to obtain a forecast/recovery of data with low accuracy up to 4-5 years, then for the medium and long-term components - a more accurate forecast with a sufficiently probable trend up to 11-12 years. Wavelet analysis was used to determine the type of aquifer.

2. Grossmann, A. and Morlet, J. (1984). Decomposition of Hardy functions intosquare integrable wavelets of constant shape. SIAM J. Math. Anal. 15. 723-736.

3. Guo, T., Song, S., Yan, Y. (2022). A time-varying autoregressive model for groundwater depth prediction. J. Hydrol. 613, 128394 https://doi.org/10.1016/j.jhydrol.2022.128394.

4. Javadinejad, S., Dara, R. and Jafary, F. (2020). How groundwater level can predict under the effect of climate change by using artificial neural networks of NARX. Resources Environment and Information Engineering. https://doi.org/10.25082/REIE.2020.01.005

5. Lototska-Dudyk, U. B., & Laboyko, V. V. (2022). Quality of drinking water and the state of water supply of the population of Ukraine under martial law. Current problems of preventive medicine. (24), 5-14. [In Ukrainian]

6. Poonia V., Tiwari H. L., Mishra S. (2018). Hydrological Analysis by Artificial Neural Network: A Review. International Journal of Advance Research, Ideas and Innovations in Technology. 4(3). 265-270.

7. Rudenko, Yu. F., Yurkova, N. A., Gural, O. V., & Saprykin, V. Yu. (2024). Some issues of local drinking water supply of the population during martial law (on the example of Ukrainian cities with a population of over 100 thousand people). Part 1. Mineral resources of Ukraine, (4), 64-70. [In Ukrainian]

8. Shestopalov, V., Rudenko, Y., Koliabina, I., Stetsenko, B., & Yaroshenko, K. (2024). Groundwater for urban water supply in Ukraine: a case study of Mykolaiv (Military challenges and lessons for the future). Acque Sotterranee-Italian Journal of Groundwater,13(3).

9. Shevchenko A.L., Osadchyі V.I., Charnyі D.V. (2019). Changes in the regime, balance and resources of underground waters of Polissya and the Forest-steppe of Ukraine under the influence of global warming. Academic notes of Brest University. 15 (2). 117-128. [In russian]

10. Shevchenko O.L., Skorbun A.D., Osadchyi V.I., Charnyi D.V. (2021). Changing rhythms in the groundwater regime and their relationship with climatic factors. Bulletin of the Taras Shevchenko National University of Kyiv (Geology). 2 (93). 71-82. http://doi.org/10.17721/1728-2713.93.08 [In Ukrainian]

11. Shevchenko O.L., Charnyi D.V., Rudoman M.M. (2022). Forecasting groundwater flow to the Southern Bug River by statistical methods and using artificial neural networks. Meteorology, Hydrology, Environmental Monitoring, 2. 43-53. [In Ukrainian]

12. ?liwi?ska, J., Birylo, M., Rzepecka, Z., & Nastula, J. (2019). Analysis of groundwater and total water storage changes in Poland using GRACE observations, in-situ data, and various assimilation and climate models. Remote Sensing, 11(24), 2949. https://doi.org/10.3390/rs11242949

13. Solovey, T., ?liwi?ska-Bronowicz, J., Janica, R., Brzezi?ska, A. (2025). Assessment of the effectiveness of GRACE observations in monitoring groundwater storage in Poland. Water Resources Research, 61(8), https://doi.org/10.1029/2024WR038888

14. Solovey, T., ?liwi?ska-Bronowicz, J., Janica, R., Brzezi?ska, A., Stradczuk, A. (2025а). Fusing GRACE data into terrestrial water budgets to improve their predictive performance: a case study of the Bug River Transboundary Catchment, Polish-Ukrainian-Belarusian Borderland. Geological Quarterly, 69(16), https://doi.org/10.7306/gq.1789

15. Wagena, M.B., Goering, D., Collick, A.S., Bock, E., Fuka, D.R., Buda, A., Easton, Z.M. (2020). Comparison of short-term streamflow forecasting using stochastic time series, neural networks, process-based and Bayesian models. Environ. Model. Softw. 126, 104669, https://doi.org/10.1016/j.envsoft.2020.104669

16. Yadav, N., Yadav, A., and Kumar, M. (2015). “An introduction to Neural Network methods for Differential equations. Springer briefs in applied science and technology. 17-42.