Knowledge Distillation for Deep Learning-Based Hydrological Prediction and Forecasting
Bibliographic record
Abstract
This study introduces knowledge distillation (KD) as a method for improving deep learning (DL)-based hydrological prediction. The performance of DL models, including long short-term memory networks (LSTMs), is highly dependent on model structure and training data quality, posing challenges for transfer learning. In the KD framework, a ”better” teacher model, based on an ensemble of models or trained with higher-quality input data, is employed to improve the training of a weaker ”student” model with simpler structure or lower quality inputs. This enables improved performance in situations where computational complexity or input data quality are limiting. We demonstrate the effectiveness of KD in two experiments conducted across 421 catchments throughout the contiguous United States. In the first experiment, KD is applied for model compression, distilling the knowledge of an ensemble of five LSTM models (teacher) into a single model (student) that outperforms single-model LSTMs trained without KD. In the second experiment, KD is applied to improve hydrologic prediction when using lower-quality reanalysis precipitation data instead of higher-quality, gauged based observations. This experiment is relevant to forecasting applications, where lower quality (e.g., lower resolution and model-based) precipitation forecasts are used to force a DL hydrologic model trained with high-resolution, observed precipitation. Results demonstrate that KD leads to substantial performance improvements, especially for catchments not used in training (> 25% improvement in median Nash-Sutcliffe efficiency). These findings underscore the value of KD in optimizing DL models for hydrological prediction, offering a model agnostic, and scalable approach that facilitates efficiency and model transferability.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".