C-band synthetic aperture radar (SAR) imagery for the classification of diverse cropping systems
Bibliographic record
Abstract
Cloudy conditions reduce the utility of optical imagery for crop monitoring. New constellations of satellites – including the RADARSAT Constellation Mission (RCM) and Sentinel-1A/B, both available under free and open data policies – can be used to create stacks of dense seasonal C-band Synthetic Aperture Radar (SAR) data. Yet to date, the contribution of SAR imagery to operational crop mapping is often limited to that of a gap-filler, compensating for optical data obscured by clouds. The Joint Experiment for Crop Assessment and Monitoring (JECAM) SAR Inter-Comparison Experiment is a multi-year, multi-partner project focused on evaluating methods for SAR-based crop classification. Stacks of dense time-series SAR imagery, from RADARSAT-2 and Sentinel-1 satellites, were acquired for 10 sites located in six countries. Decision Tree (DT) and Random Forest (RF) classification methodologies were applied to these SAR data-stacks, as well as to data-stacks of optical only, and optimized SAR/optical data combinations. For the dense time-series SAR stacks, overall classification accuracies above 85% and 80% were obtained for 6 of 10 and 8 of 10 sites, respectively. For maize, the SAR-only data delivered user’s and producer’s accuracies greater than 90% for half the sites. For soya bean, accuracies greater than 80% were reported for 5 of 9 sites and classification accuracies were greater than 80% for wheat on half the sites. Classification results were influenced by the mix and number of agriculture classes present at each site, the available SAR imagery, as well as the training and validation data sets for individual crop types. These results have important operational implications for regions of the world dominated by cloudy conditions and the lack of adequate amounts of optical imagery to support satellite-based crop monitoring.
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 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.002 | 0.003 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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".