Mapping surface hoar from near-infrared texture in a laboratory
Bibliographic record
Abstract
Abstract. Surface hoar crystals are snow grains that form when water vapor deposits on the snow surface. Once buried, surface hoar creates a weak layer in the snowpack that can later cause large avalanches to occur. The formation and persistence of surface hoar are highly spatiotemporally variable, making its detection difficult. Remote-sensing technology capable of detecting the presence and spatial distribution of surface hoar would be beneficial for avalanche forecasting, but this capability has yet to be developed. Here, we hypothesize that near-infrared (NIR) texture, defined as the spatial variability of reflectance magnitude, may produce an optical signature unique to surface hoar due to the distinct shape and orientation of the grains. We tested this hypothesis by performing reflectance experiments in a controlled cold laboratory environment to evaluate the potential and accuracy of surface hoar mapping from NIR texture using a near-infrared hyperspectral imager (NIR-HSI) and a lidar operating at 1064 nm. We analyzed 41 snow samples, three of which were surface hoar and 38 of which consisted of other grain morphologies. When using NIR-HSI under direct and diffuse illumination, we found that surface hoar displayed higher NIR texture relative to all other grain shapes across numerous spectral bands and a wide range of spatial resolutions (0.5–50 mm). Due to the large number of spectral- and spatial-resolution combinations, we conducted a detailed samplewise case study at 1324 nm spectral and 10 mm spatial resolution. The case study resulted in the median texture of surface hoar being 1.3 to 8.6 times greater than that of the 38 other samples under direct and diffuse illumination (p < 0.05 in all cases). Using lidar, surface hoar also exhibited significantly increased NIR texture in 30 out of 38 samples, but only at select (5–25 mm) spatial resolutions. Leveraging these results, we propose a simple binary classification algorithm to map the extent of surface hoar on a pixelwise basis using both the NIR-HSI and lidar instruments. The NIR-HSI under direct and diffuse illumination performed best, with a median accuracy of 96.91 % and 97.37 %, respectively. Conversely, the median classification accuracy achieved with lidar was only 66.99 %. Further, to assess the repeatability of our method and demonstrate its mapping capacity, we ran the algorithm on a new sample with mixed microstructures, with an accuracy of 99.61 % and 96.15 % achieved using NIR-HSI under direct and diffuse illumination, respectively. As NIR-HSI detectors become increasingly available, our findings demonstrate the potential of a new tool for avalanche forecasters to remotely assess the spatiotemporal variability of surface hoar, which would improve avalanche forecasts and potentially save lives.
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 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 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".