Deep Learning‐based InSAR Phase Gradient Stacking Method for Mapping Active Geohazards in the Lower Yarlung Tsangpo, China
Bibliographic record
Abstract
Abstract The lower Yarlung Tsangpo River basin of the Qinghai–Tibet Plateau frequently experiences geo‐hazardous occurrences such as landslides, ice/rock avalanches and debris flows, causing loss of human lives and damage to infrastructure. However, a comprehensive inventory map of geohazards is lacking for this region, due to the extreme challenges of the geomorphological and environmental conditions (i.e., steep terrain, dense vegetation cover, and the presence of ice and snow). To this end, we propose a novel approach for mapping active geohazards in complex mountainous regions through InSAR phase gradient measurements based on a deep learning algorithm, which is then applied to the lower Yarlung Tsangpo River basin for the first time, in order to prepare an inventory map of active geohazards using ascending and descending Sentinel‐1 SAR images acquired between March 2017 and July 2023. First, the InSAR phase gradient stacking method was introduced to estimate ground deformation, which offers significant advantages in minimizing the influence of InSAR decorrelation and effectively suppressing topographic residuals and atmospheric delays. InSAR phase gradient rates effectively retrieve patterns of localized ground deformation associated with geohazard activity. Then, a DeepLabv3 deep learning model was established and trained with phase gradient rate maps of manually labeled geohazards, in order to achieve the automatic identification of active geohazards. Our results show that there are 277 active geohazards within the lower Yarlung Tsangpo River basin, encompassing an area of ~25600 km 2 . The DeepLabv3 model achieved good precision, recall rate and F1 scores at 92, 86 and 90%, respectively. The distribution of detected geohazards is closely correlated with the topographic factors, faults and river system. Compared to the results derived from Small Baseline Subset InSAR (SBAS‐InSAR) and optical images, the proposed approach can obtain high density pixels of InSAR measurement in low‐coherence scenarios, thus enabling high‐accuracy mapping of active geohazards in complex mountainous areas.
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 | 0.000 |
| 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 | 0.000 | 0.000 |
| Open science | 0.001 | 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".