Evaluating the capacity of single photon lidar for terrain characterization under a range of forest conditions
Bibliographic record
Abstract
Accurate digital elevation models are key data products used to inform forest management. Light detection and ranging (lidar) technologies have emerged as a useful tool for acquiring detailed terrain information, although the accuracy of this data is known to vary with topographic complexity and the density and characteristics of overlying vegetation. Single Photon Lidar (SPL) provides a high-density point cloud that can be acquired from a much higher altitude than discrete return, small-footprint lidar (hereafter, linear-mode lidar or LML), providing efficiencies and potential cost savings for operational mapping programs. Herein, we assess the absolute and relative accuracies of leaf-on and leaf-off SPL data acquired at different altitudes for characterizing terrain under varying vegetation types and densities and compare to results for LML data. Our assessment was forest-focused and primarily point based, using 299 Real-Time Kinematic survey checkpoints to quantify elevation errors (Δh); however, we also investigated and reported accuracy for linear transects, and conducted a wall-to-wall comparison of the SPL-derived 1-m digital elevation models (DEMs) against an LML-derived DEM. Point cloud characteristics for the leaf-on 2018 SPL data were markedly different, with 88% of returns as first returns, compared to 17% for the LML, and 59% and 46% for the leaf-off SPL data acquired at 3800 m and 2000 m, respectively. Of the datasets considered herein, the SPL data acquired under leaf-on conditions in 2018 had the lowest accuracy and precision for characterizing terrain underneath vegetation cover, with an RMSE of 10.97 cm and a 95th quantile of 24.03 cm; however these values are within commonly accepted error limits for elevation products. The leaf-off SPL data were most accurate overall; however, the differences between the leaf-off SPL data acquired at 3800 m versus 2000 m were often minor (< 1 cm on average), with similar patterns in Δh between the two datasets (r = 0.8). In terms of the relative performance of the lidar datasets examined, results from the analyses of linear transects were similar to those of the checkpoints, but highlighted the variability in elevation accuracy within similar cover types. Wall-to-wall comparisons of the SPL-derived DEMs to the 2012 LML DEM also corroborated the results of the checkpoint assessment, with the 2018 SPL leaf-on DEM having the largest differences (mean difference = 7.44 cm; RMSD = 18.07 cm). Differences between DEMs did not trend consistently with increasing canopy cover or with the percentage of returns that were within ±15 cm of the ground surface. We found that it was not only the density of the vegetation, but also the composition and configuration of both the overstory and understory vegetation that influenced the accuracy with which the lidar characterized the terrain surface. Overall, our results indicated that leaf-on SPL is capable of capturing terrain information under a wide variety of forest and vegetation conditions, albeit at a lower accuracy than what is possible with leaf-on LML or leaf-off SPL, but at a level of accuracy that is within acceptable limits for most forest applications.
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".