Cite this DOI

Uma Kannan, & Rajendran Swamidurai (2025). Attention-Enhanced Sequential GAN for Reliable Groundwater Recharge Time-Series Augmentation. *Journal of Science & Technology*, *10*(10), 01–09. https://doi.org/10.46243/jst.2025.v10.i10.pp01-09

@article{umakannan2025attentionenhanced,
  author    = {Uma Kannan and Rajendran Swamidurai},
  title     = {{Attention-Enhanced Sequential GAN for Reliable Groundwater Recharge Time-Series Augmentation}},
  journal   = {Journal of Science \& Technology},
  year      = {2025},
  month     = {oct},
  volume    = {10},
  number    = {10},
  pages     = {01--09},
  publisher = {Longman Publishers},
  issn      = {2456-5660},
  doi       = {10.46243/jst.2025.v10.i10.pp01-09},
  url       = {https://doi.org/10.46243/jst.2025.v10.i10.pp01-09},
  language  = {en},
  abstract  = {Groundwater recharge modeling is critically hindered by the scarcity of long-term, high-resolution time-series data, limiting the robustness and generalization capability of predictive models. We propose the Attention-enhanced Sequential Generative Adversarial Network () to synthesize high-fidelity, multivariatehydrological records, explicitly addressing the complex temporal dependencies required for groundwater dynamics. The architecture incorporates three key innovations: stabilization via the WGAN-GP objective for continuous learning; utilization of a pre-trained LSTM autoencoder to establish a meaningful latent space; and integration of a Self-Attention mechanism within the generative networks to effectively capture critical long-range dependencies, such as multi-year climatic cycles. A three-pronged evaluation demonstrated exceptional data quality: Statistical Fidelity confirmed the preservation of feature relationships, and Temporal Coherence validated the realism of sequential patterns. Crucially, the Predictive Utility was confirmed, with an auxiliaryforecasting model trained on synthetic data achieving a Mean Absolute Error (MAE only 4.4\% higher) than a model trained on real data. This provides a stable and effective generative approach for time-series augmentation, offering a viable path to developing reliable forecasting tools in data-scarce hydrological contexts.}
}

TY  - JOUR
TI  - Attention-Enhanced Sequential GAN for Reliable Groundwater Recharge Time-Series Augmentation
AU  - Uma Kannan
AU  - Rajendran Swamidurai
JO  - Journal of Science & Technology
PY  - 2025
DA  - 2025/10/27/
VL  - 10
IS  - 10
SP  - 01
EP  - 09
PB  - Longman Publishers
SN  - 2456-5660
LA  - en
AB  - Groundwater recharge modeling is critically hindered by the scarcity of long-term, high-resolution time-series data, limiting the robustness and generalization capability of predictive models. We propose the Attention-enhanced Sequential Generative Adversarial Network () to synthesize high-fidelity, multivariatehydrological records, explicitly addressing the complex temporal dependencies required for groundwater dynamics. The architecture incorporates three key innovations: stabilization via the WGAN-GP objective for continuous learning; utilization of a pre-trained LSTM autoencoder to establish a meaningful latent space; and integration of a Self-Attention mechanism within the generative networks to effectively capture critical long-range dependencies, such as multi-year climatic cycles. A three-pronged evaluation demonstrated exceptional data quality: Statistical Fidelity confirmed the preservation of feature relationships, and Temporal Coherence validated the realism of sequential patterns. Crucially, the Predictive Utility was confirmed, with an auxiliaryforecasting model trained on synthetic data achieving a Mean Absolute Error (MAE only 4.4% higher) than a model trained on real data. This provides a stable and effective generative approach for time-series augmentation, offering a viable path to developing reliable forecasting tools in data-scarce hydrological contexts.
DO  - 10.46243/jst.2025.v10.i10.pp01-09
UR  - https://doi.org/10.46243/jst.2025.v10.i10.pp01-09
ER  -

{
    "type": "article-journal",
    "id": "10.46243/jst.2025.v10.i10.pp01-09",
    "DOI": "10.46243/jst.2025.v10.i10.pp01-09",
    "URL": "https://doi.org/10.46243/jst.2025.v10.i10.pp01-09",
    "title": "Attention-Enhanced Sequential GAN for Reliable Groundwater Recharge Time-Series Augmentation",
    "source": "Smart Scholars DOI Registry",
    "container-title": "Journal of Science & Technology",
    "author": [
        {
            "family": "Uma Kannan"
        },
        {
            "family": "Rajendran Swamidurai"
        }
    ],
    "issued": {
        "date-parts": [
            [
                2025,
                10,
                27
            ]
        ]
    },
    "volume": "10",
    "issue": "10",
    "page": "01-09",
    "publisher": "Longman Publishers",
    "language": "en",
    "abstract": "Groundwater recharge modeling is critically hindered by the scarcity of long-term, high-resolution time-series data, limiting the robustness and generalization capability of predictive models. We propose the Attention-enhanced Sequential Generative Adversarial Network () to synthesize high-fidelity, multivariatehydrological records, explicitly addressing the complex temporal dependencies required for groundwater dynamics. The architecture incorporates three key innovations: stabilization via the WGAN-GP objective for continuous learning; utilization of a pre-trained LSTM autoencoder to establish a meaningful latent space; and integration of a Self-Attention mechanism within the generative networks to effectively capture critical long-range dependencies, such as multi-year climatic cycles. A three-pronged evaluation demonstrated exceptional data quality: Statistical Fidelity confirmed the preservation of feature relationships, and Temporal Coherence validated the realism of sequential patterns. Crucially, the Predictive Utility was confirmed, with an auxiliaryforecasting model trained on synthetic data achieving a Mean Absolute Error (MAE only 4.4% higher) than a model trained on real data. This provides a stable and effective generative approach for time-series augmentation, offering a viable path to developing reliable forecasting tools in data-scarce hydrological contexts.",
    "ISSN": "2456-5660"
}

From the record as registered (version 2) — the record and its history. Programs: https://registry.smartscholars.in/api.php?action=cite&doi=10.46243%2Fjst.2025.v10.i10.pp01-09 gives all four in one JSON answer.