Using Planet data to classify mangrove subtypes in Northwestern Madagascar in 2020 and 2022
Bibliographic record
Abstract
In response to the growing concerns about mangrove deforestation, recent studies have used various remote sensing technology like satellite imagery to measure the mangrove extent. In this work, we investigated the mangrove distribution in Northwestern Madagascar by using fine spatial imagery with a pixel size as small as 3m and compared it with the result of the traditional method based on relatively coarser Landsat data. Mangroves are an essential biodiverse ecosystem found along tropical and subtropical intertidal beaches, providing critical goods and services to coastal communities, and supporting diverse organisms. However, anthropogenic activities have caused the loss of mangroves in Madagascar, necessitating a new mapping approach utilizing the fine spatial resolution map from Planet data to create a map with advanced detail. The quantitative result central to this work is the new multi-date map of the Tsimipaika- Ampasindava-Ambaro Bays (TAB) from 2020 to 2022, which provides advanced detail and direct comparison with the shift in local mangrove species. The classification maps are based on Random Forest and Maximum Likelihood algorithms, and all of them have an overall accuracy of over 85%. The dynamics of mangrove forests from 2020 to 2022 are quantified, with a 12.6% loss in closed-canopy mangroves, and a 24.1% loss in open-canopy mangroves I am overestimated. Limitations regarding the classification model are also found in this study, including the overestimation of open canopy mangroves caused by the shadow and the seamline in the base map. This result shows the potential of using fine-resolution satellite imagery in supervised land cover classification, and the corresponding challenges raised by the smaller pixel size.
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) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| 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".