Articles | Volume 28, issue 4
https://doi.org/10.5194/hess-28-917-2024
https://doi.org/10.5194/hess-28-917-2024
Research article
 | 
27 Feb 2024
Research article |  | 27 Feb 2024

A comprehensive study of deep learning for soil moisture prediction

Yanling Wang, Liangsheng Shi, Yaan Hu, Xiaolong Hu, Wenxiang Song, and Lijun Wang

Related authors

Interpretable Soil Moisture Prediction with a Physics-guided Deep Learning Approach
Yanling Wang, Xiaolong Hu, Yaan Hu, Leilei He, Lijun Wang, Wenxiang Song, and Liangsheng Shi
EGUsphere, https://doi.org/10.5194/egusphere-2025-4440,https://doi.org/10.5194/egusphere-2025-4440, 2025
This preprint is open for discussion and under review for Hydrology and Earth System Sciences (HESS).
Short summary

Cited articles

Abbaszadeh, P., Moradkhani, H., and Zhan, X.: Downscaling SMAP radiometer soil moisture over the CONUS using an ensemble learning method, Water Resour. Res., 55, 324–344, 2019. 
Abdel-Hamid, O., Mohamed, A. R., Jiang, H., Deng, L., Penn, G., and Yu, D.: Convolutional neural networks for speech recognition, IEEE T. Audio, Speech, 22, 1533–1545, https://doi.org/10.1109/TASLP.2014.2339736, 2014. 
Ahmed, A. A. M., Deo, R. C., Ghahramani, A., Raj, N., Feng, Q., Yin, Z., and Yang, L.: LSTM integrated with Boruta-random forest optimiser for soil moisture estimation under RCP4.5 and RCP8.5 global warming scenarios, Springer Berlin Heidelberg, 1851–1881 pp., https://doi.org/10.1007/s00477-021-01969-3, 2021. 
Ajit, A., Acharya, K., and Samanta, A.: A Review of Convolutional Neural Networks, Int. Conf. Emerg. Tr., 1–5, https://doi.org/10.1109/ic-ETITE47903.2020.049, 2020. 
Albawi, S., Mohammed, T. A., and Al-Zawi, S.: Understanding of a convolutional neural network, Proc. 2017 Int. Conf. Eng. Technol., ICET 2017, Antalya, Turkey, 21–23 August 2017, IEEE: Piscataway, NJ, USA, 2017, 1–6 https://doi.org/10.1109/ICEngTechnol.2017.8308186, 2018. 
Download
Short summary
LSTM temporal modeling suits soil moisture prediction; attention mechanisms enhance feature learning efficiently, as their feature selection capabilities are proven through Transformer and attention–LSTM hybrids. Adversarial training strategies help extract additional information from time series’ data. SHAP analysis and t-SNE visualization reveal differences in encoded features across models. This work serves as a reference for time series’ data processing in hydrology problems.
Share