Ackroyd et al. (2026) Spatial assessment of snow grain size from airborne lidar reflectance against coincident imaging spectroscopy retrievals

Identification

• Journal: Remote Sensing of Environment
• Year: 2026
• Date: 2026-03-13
• Authors: C. Ackroyd, Christopher Donahue, Brian Menounos, S. McKenzie Skiles
• DOI: 10.1016/j.rse.2026.115366

Research Groups

• The School of Environment, Society, and Sustainability, University of Utah, Salt Lake City, UT, USA
• Department of Geography, Earth, and Environmental Sciences, University of Northern British Columbia, Prince George, British Columbia, Canada
• Hakai Institute, Cambell River, Quadra Island, British Columbia, Canada
• Bridger Photonics, Inc., Bozeman, MT 59715, USA
• Geological Survey of Canada Pacific, Natural Resources Canada, Sidney, British Columbia, Canada

Short Summary

This study evaluates three methods for deriving snow grain size and albedo from 1064 nm airborne lidar reflectance against coincident imaging spectroscopy retrievals over a snow-covered glacier. It demonstrates that incorporating incidence angle corrections is crucial for accurate lidar-derived snow properties in mountainous terrain, highlighting lidar's potential for high-resolution snow property mapping.

Objective

• To assess the differences among various lidar reflectance products and evaluate their implications for snow property retrievals in complex mountain terrain.

Study Configuration

• Spatial Scale: Place Glacier, British Columbia, Canada (50.42°N, 123.60°W), with elevations ranging from approximately 1830 m to 2500 m. Lidar footprint was approximately 30 cm at nadir, with final gridded products at 1 m resolution.
• Temporal Scale: Data collected on May 15, 2021, between approximately 12:45 PM and 1:15 PM PDT.

Methodology and Data

• Models used:
  • Analytical Asymptotic Radiative Transfer (ART) model (for simulating spectral absorption, albedo, and BRDF for grain size and albedo retrieval).
  • Atmospheric and Topographic Correction for airborne imagery software (ATCOR-4) (for imaging spectroscopy radiance to reflectance correction).
  • Imaging Spectrometer Snow and Ice Algorithm (ISSIA) (for imaging spectroscopy snow property retrieval).
• Data sources:
  • Airborne Coastal Observatory (ACO) platform.
  • Riegl Q780 lidar (1064 nm wavelength).
  • Specim Fenix imaging spectrometer (380–2500 nm, 4.5 nm VIS band spacing, 13.5 nm SWIR band spacing).
  • Global Navigation Satellite System and Inertial Measurement Unit (GNSS-IMU).

Main Results

• Reflectance: Lidar reflectance products incorporating incidence angle corrections (LRIC and LRrieglθ) showed similar median reflectance magnitudes to imaging spectroscopy (0.39 vs. 0.38), with a median difference of less than 0.01 (1%), but displayed greater variability (standard deviation 0.06 vs. 0.03). The vendor-provided product without incidence angle correction (LRriegl) was positively biased (median 0.47 vs. 0.38), with a median difference of 0.09 (24%).
• Grain Size: All lidar products underestimated grain size compared to imaging spectroscopy (median 1030 µm). LRIC and LRrieglθ underestimated grain size by approximately 136 µm (13–14% relative error), while LRriegl showed a greater underestimation of 455 µm (42% relative error).
• Topographic Effects: LRIC and LRrieglθ performed best on flat to moderately sloped terrain (slopes < 30°), with median relative errors below 20%. Errors increased significantly on steeper slopes (>60°), attributed to overcorrection at high incidence angles. LRriegl exhibited the highest errors on flatter terrain.
• Albedo: LRIC and LRrieglθ yielded median albedo values (0.72) closely aligned with imaging spectroscopy (0.71), resulting in a median relative error of 2%. LRriegl showed a higher median albedo (0.75) and a 6% median relative error. These albedo errors correspond to absorbed net solar radiation errors of 10 W m⁻² for LRIC/LRrieglθ and 40 W m⁻² for LRriegl, assuming an incoming solar irradiance of 1000 W m⁻².

Contributions

• This study provides the first comprehensive comparison of different lidar intensity-to-reflectance conversion methods for snow grain size retrieval in mountainous terrain, assessing their accuracy against coincident imaging spectroscopy.
• It demonstrates the significant potential of 1064 nm lidar intensity for high-resolution snow grain size estimation, offering a valuable complement to passive observations, especially in challenging conditions like terrain shadows, mixed pixels, and canopy gaps.
• The research critically highlights the necessity of incorporating incidence angle corrections for accurate lidar reflectance and subsequent snow property retrievals in complex, topographically varied environments.
• It includes a sensitivity analysis quantifying how uncertainty in lidar-derived grain size propagates to clean snow albedo, providing crucial context for interpreting results.

Funding

• NASA Project 80NSSC22K0686
• Tula Foundation
• Natural Sciences and Engineering Research Council of Canada
• Hakai Geospatial Team

Citation

@article{Ackroyd2026Spatial,
  author = {Ackroyd, C. and Donahue, Christopher and Menounos, Brian and Skiles, S. McKenzie},
  title = {Spatial assessment of snow grain size from airborne lidar reflectance against coincident imaging spectroscopy retrievals},
  journal = {Remote Sensing of Environment},
  year = {2026},
  doi = {10.1016/j.rse.2026.115366},
  url = {https://doi.org/10.1016/j.rse.2026.115366}
}