Multi-temporal Optical and Synthetic Aperture Radar (SAR) Satellite Imagery for Estimating Canola Harvest Dates Using Machine Learning
Bibliographic record
Abstract
Canola, a cornerstone of Canadian agriculture, contributes over $43 billion annually to the national economy and occupies approximately 20 million acres. Traditional monitoring methods, such as crop reporter surveys, are often limited by spatial coverage and consistency. This study explores the integration of multi-temporal Sentinel-1 (SAR) and Sentinel-2 (optical) satellite imagery with deep learning to predict canola harvest dates across Saskatchewan, Canada, over five growing seasons (2020–2024). Twenty-two vegetation and radar-based indices were computed within the Google Earth Engine environment, capturing dynamic canopy, moisture, and phenological changes. Gradient time series (first derivatives) of these indices were used to train univariate Long Short-Term Memory (LSTM) models, with Mc, RBNI, and NGWI achieving the highest predictive performance (RMSE ≈ 10.2 days). Three indices were then combined in a multivariate LSTM model, which significantly improved prediction accuracy, achieving an RMSE of 8.18 days and demonstrating robust alignment with observed harvest dates. The findings underscore the value of combining SAR and optical data for modeling complex phenological events and highlight the potential of scalable deep learning frameworks in advancing precision agriculture. Future work will expand this methodology across Western Canada and incorporate real-time weather and soil data to further enhance prediction generalizability.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".