On the sensitivity of SAR C- and L-band dual-polarized data for detection of early deforestation in the tropics
Bibliographic record
Abstract
Operational national forest monitoring requires frequent, reliable observations for timely detection of deforestation and other forest changes. However, tropical forests, which encompass all current major deforestation fronts, are often hampered by persistent cloud cover. While radar remote sensing offers a compelling complement to optical data, a significant research gap remains in distinguishing the strengths and limitations of different radar sensors to the multiple stages of forest disturbance. Given the impending rapid expansion of available data outside the C band, it is imperative to assess the ability of these sensors to rapidly and accurately detect changes of high importance to the remote sensing community. To address this gap, we investigated the sensitivity of two freely available radar datasets — JAXA’s ALOS-2 PALSAR-2 (L-band) and ESA’s Sentinel-1 (C-band) — to distinct stages of tropical forest loss. With a particular focus on detecting the earliest stage of deforestation, we compared backscatter values, SAR indices, and statistical metrics against time series data from 92 locations over three years in the Amazon. In these locations, early deforestation, biomass burning, and vegetation regrowth were occurring in different stages of the full conversion process. Our analysis revealed that the L-band-derived Radar Forest Degradation Index (RFDI) is highly sensitive to early deforestation, even when biomass remains on the ground soon after cutting. In contrast, C-band information showed limited ability to sense this critical initial stage of change, but was much stronger at detecting later deforestation stages. Our results point the way toward combining information from the upcoming L-Band NISAR mission with the existing C-band information to produce a multi-component system that can accurately detect deforestation in all its temporal stages.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".