Evaluating simulated compact polarimetry for Emergency Geomatics Service flood mapping
Bibliographic record
Abstract
When severe flooding occurs in Canada, the Emergency Geomatics Service (EGS) is tasked with creating and disseminating maps that depict the flood extent in near real time. These maps delineate open water and flooded vegetation and are created using a combination of raster image processing, machine learning classification, and threshold-based region growing. The predominant data source used to create these maps is synthetic aperture radar (SAR) imagery from RADARSAT-2 (R2). With the commissioning phase of the RADARSAT Constellation Mission (RCM) nearing completion, the EGS must adapt its methods for use with what is expected to be the EGS's new default source of SAR data. The introduction of RCM's circular-transmit linear-receive (CTLR) beam mode provides the option to exploit compact polarimetric (CP) information not previously available through R2. The aim of this study is to determine the most effective CP parameters for use in mapping open water and flooded vegetation by applying current EGS methodologies, as well as to assess the quality of these products in comparison to products created using R2 data. Nineteen quad-polarization R2 scenes selected from three regions prone to springtime flooding were used to create reference flood maps using current EGS methodologies. These scenes were also used to simulate RCM CP data that included 22 parameters at three different noise floors and spatial resolutions representative of three RCM beam modes. Using a multi-criteria ranking procedure, CP parameters were ranked in order of importance and entered into a stepwise classification procedure for evaluation against reference R2 products. Results suggest the top four CP parameters - m-chi-volume or m-delta-volume, RR intensity, Shannon Entropy intensity (SEi), and RV intensity - achieved a minimum omission and commission, and maximum agreement with baseline R2 products when evaluated across all 19 scenes and three beam modes. Separability analyses between manually delineated flooded vegetation polygons and other land cover classes identified four candidate CP parameters - RH intensity, RR intensity, SEi, and the first Stokes parameter (SV0) - suitable for threshold-based flooded vegetation region growing. Region growing thresholds based on CP values beneath flooded vegetation polygons were found to be dependent on incidence angle for each of these four parameters. After region growing flooded vegetation using established threshold values for each of the four candidate CP parameters, results were evaluated against flooded vegetation polygons to assess omission error, and against upland land cover to assess commission error, wherein RH intensity was deemed best to map flooded vegetation. The results of the study are a set of suitable CP parameters to generate flood maps from RCM data using current EGS methodologies that must be verified and validated further once real RCM data become available.
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 | 0.002 | 0.004 |
| 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 | 0.001 |
| 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".