Kevin Hammonds
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Kevin Hammonds has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Christopher Paul Donahue
Bridger Photonics, Inc., Bozeman, MT, 59715, USA
17 shared publications - Evan Schehrer
Department of Civil Engineering, Montana State University, Bozeman, MT, USA
12 shared publications - James W. Dillon
Department of Civil Engineering, Montana State University, Bozeman, MT, USA
12 shared publications James Dillon
Department of Civil Engineering, Montana State University, Bozeman, MT 59717, USA
7 shared publicationsKarl Birkeland
USDA Forest Service National Avalanche Center, Bozeman, MT, USA
5 shared publications- S. McKenzie Skiles
University of Utah
5 shared publications
Latest Publications
- Open Access
- 6 August 2025
- Published by Copernicus GmbH in The Cryosphere
- Vol. 19 (8) , 2913-2933
Evaluating sensitivity of optical snow grain size retrievals to radiative transfer models, shape parameters, and inversion techniques
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability and Earth's energy budget toward future climates. While numerous studies have demonstrated estimates of ropt by means of optical instruments that span scales from in situ to satellites, they leverage different retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors causes tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Toward enhanced ropt retrieval accuracy, we sought to determine (1) the optimal modeled snow grain shape; (2) the best-performing radiative transfer model; and (3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation, with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6 %–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Toward a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9 %–26.2 %, depending on model and technique, with most pairings resulting in values ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improves, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- 6 August 2025
- Published by Copernicus GmbH in The Cryosphere
- Vol. 19 (8) , 2913-2933
Open AccessEvaluating sensitivity of optical snow grain size retrievals to radiative transfer models, shape parameters, and inversion techniques
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability and Earth's energy budget toward future climates. While numerous studies have demonstrated estimates of ropt by means of optical instruments that span scales from in situ to satellites, they leverage different retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors causes tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Toward enhanced ropt retrieval accuracy, we sought to determine (1) the optimal modeled snow grain shape; (2) the best-performing radiative transfer model; and (3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation, with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6 %–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Toward a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9 %–26.2 %, depending on model and technique, with most pairings resulting in values ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improves, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- Open Access
- 8 February 2025
- Published by Copernicus GmbH
- Vol. 2024, 1-31
Reply on RC1
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability, and Earth’s energy budget towards future climates. While numerous studies have demonstrated estimates of ropt via optical instruments that span scales from in situ to satellites, they leverage differing retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors cause tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Towards enhanced ropt retrieval accuracy, we sought to determine 1) the optimal modeled snow grain shape, 2) the best-performing radiative transfer model, and 3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape-optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Towards a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9–26.2 % depending on model and technique, with most pairings resulting in values < 15 %. Our results demonstrate that the accuracy of ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improve, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- 8 February 2025
- Published by Copernicus GmbH
- Vol. 2024, 1-31
Open AccessReply on RC1
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability, and Earth’s energy budget towards future climates. While numerous studies have demonstrated estimates of ropt via optical instruments that span scales from in situ to satellites, they leverage differing retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors cause tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Towards enhanced ropt retrieval accuracy, we sought to determine 1) the optimal modeled snow grain shape, 2) the best-performing radiative transfer model, and 3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape-optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Towards a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9–26.2 % depending on model and technique, with most pairings resulting in values < 15 %. Our results demonstrate that the accuracy of ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improve, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- Open Access
- 8 February 2025
- Published by Copernicus GmbH
- Vol. 2024, 1-31
Reply on RC2
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability, and Earth’s energy budget towards future climates. While numerous studies have demonstrated estimates of ropt via optical instruments that span scales from in situ to satellites, they leverage differing retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors cause tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Towards enhanced ropt retrieval accuracy, we sought to determine 1) the optimal modeled snow grain shape, 2) the best-performing radiative transfer model, and 3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape-optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Towards a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9–26.2 % depending on model and technique, with most pairings resulting in values < 15 %. Our results demonstrate that the accuracy of ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improve, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- 8 February 2025
- Published by Copernicus GmbH
- Vol. 2024, 1-31
Open AccessReply on RC2
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability, and Earth’s energy budget towards future climates. While numerous studies have demonstrated estimates of ropt via optical instruments that span scales from in situ to satellites, they leverage differing retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors cause tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Towards enhanced ropt retrieval accuracy, we sought to determine 1) the optimal modeled snow grain shape, 2) the best-performing radiative transfer model, and 3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape-optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Towards a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9–26.2 % depending on model and technique, with most pairings resulting in values < 15 %. Our results demonstrate that the accuracy of ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improve, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- Open Access
- 14 December 2024
- Published by Copernicus GmbH
- Vol. 2024, 1-31
Comment on egusphere-2024-3141
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability, and Earth’s energy budget towards future climates. While numerous studies have demonstrated estimates of ropt via optical instruments that span scales from in situ to satellites, they leverage differing retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors cause tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Towards enhanced ropt retrieval accuracy, we sought to determine 1) the optimal modeled snow grain shape, 2) the best-performing radiative transfer model, and 3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape-optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Towards a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9–26.2 % depending on model and technique, with most pairings resulting in values < 15 %. Our results demonstrate that the accuracy of ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improve, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- 14 December 2024
- Published by Copernicus GmbH
- Vol. 2024, 1-31
Open AccessComment on egusphere-2024-3141
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability, and Earth’s energy budget towards future climates. While numerous studies have demonstrated estimates of ropt via optical instruments that span scales from in situ to satellites, they leverage differing retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors cause tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Towards enhanced ropt retrieval accuracy, we sought to determine 1) the optimal modeled snow grain shape, 2) the best-performing radiative transfer model, and 3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape-optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Towards a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9–26.2 % depending on model and technique, with most pairings resulting in values < 15 %. Our results demonstrate that the accuracy of ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improve, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- Open Access
- 2 December 2024
- Published by Copernicus GmbH
- Vol. 2024, 1-31
Comment on egusphere-2024-3141
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability, and Earth’s energy budget towards future climates. While numerous studies have demonstrated estimates of ropt via optical instruments that span scales from in situ to satellites, they leverage differing retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors cause tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Towards enhanced ropt retrieval accuracy, we sought to determine 1) the optimal modeled snow grain shape, 2) the best-performing radiative transfer model, and 3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape-optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Towards a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9–26.2 % depending on model and technique, with most pairings resulting in values < 15 %. Our results demonstrate that the accuracy of ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improve, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- 2 December 2024
- Published by Copernicus GmbH
- Vol. 2024, 1-31
Open AccessComment on egusphere-2024-3141
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability, and Earth’s energy budget towards future climates. While numerous studies have demonstrated estimates of ropt via optical instruments that span scales from in situ to satellites, they leverage differing retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors cause tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Towards enhanced ropt retrieval accuracy, we sought to determine 1) the optimal modeled snow grain shape, 2) the best-performing radiative transfer model, and 3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape-optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Towards a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9–26.2 % depending on model and technique, with most pairings resulting in values < 15 %. Our results demonstrate that the accuracy of ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improve, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- Open Access
- 7 November 2024
- Published by Copernicus GmbH in EGUsphere
- Vol. 2024, 1-31
Evaluating sensitivity of optical snow grain size retrievals to radiative transfer models, shape parameters, and inversion techniques
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability, and Earth’s energy budget towards future climates. While numerous studies have demonstrated estimates of ropt via optical instruments that span scales from in situ to satellites, they leverage differing retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors cause tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Towards enhanced ropt retrieval accuracy, we sought to determine 1) the optimal modeled snow grain shape, 2) the best-performing radiative transfer model, and 3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape-optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Towards a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9–26.2 % depending on model and technique, with most pairings resulting in values < 15 %. Our results demonstrate that the accuracy of ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improve, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- 7 November 2024
- Published by Copernicus GmbH in EGUsphere
- Vol. 2024, 1-31
Open AccessEvaluating sensitivity of optical snow grain size retrievals to radiative transfer models, shape parameters, and inversion techniques
The near-infrared (NIR) albedo of snow is controlled by optical snow grain size (ropt). Therefore, characterizing the spatial and temporal variability in ropt at the snow surface is critical for understanding melt timing and magnitude for water availability, and Earth’s energy budget towards future climates. While numerous studies have demonstrated estimates of ropt via optical instruments that span scales from in situ to satellites, they leverage differing retrieval techniques, radiative transfer models, and modeled snow grain shapes. Variation in these factors cause tremendous uncertainty in ropt retrievals, yet a thorough evaluation has yet to be conducted. To address this knowledge gap we conducted a laboratory bidirectional reflectance study, using NIR hyperspectral imaging (NIR-HSI) to retrieve grain size metrics for a wide variety of snow microstructures and evaluate them against micro-CT benchmarks. Towards enhanced ropt retrieval accuracy, we sought to determine 1) the optimal modeled snow grain shape, 2) the best-performing radiative transfer model, and 3) variability associated with retrieval techniques, spanning broadband, narrowband, multispectral, and hyperspectral approaches. Our results for optimizing grain shape parameters align with existing studies for the TARTES model, and we offer first recommendations for the SNICAR model. The retrieval technique also displayed considerable variation with the hyperspectral residual method performing best. Multispectral and single-band techniques were comparable to their hyperspectral counterparts at times, but this was attributed to the idealized laboratory conditions and high instrument signal-to-noise ratio. Following shape-optimization, the SNICAR and TARTES models produced the best results (median absolute error of 15.6–17.4 %, depending on technique), outperforming the AART model and the Random Mixture model. Towards a more direct comparison with albedo estimate error, we also evaluated the square root of ropt retrievals; median absolute error values ranged from 7.9–26.2 % depending on model and technique, with most pairings resulting in values < 15 %. Our results demonstrate that the accuracy of ropt retrievals is highly sensitive to the choice of retrieval technique, radiative transfer model, and grain shape parameters. To minimize error, each of these factors should be carefully selected in the context of the specific measurement. As NIR-HSI instruments and other NIR detectors become increasingly affordable and their resolution improve, the findings presented here provide guidance for improved ropt and snow albedo mapping across ground-based, aerial, and satellite platforms.
- Open Access
- 24 May 2024
- Published by Copernicus GmbH in The Cryosphere
- Vol. 18 (5) , 2557-2582
Mapping surface hoar from near-infrared texture in a laboratory
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.
- 24 May 2024
- Published by Copernicus GmbH in The Cryosphere
- Vol. 18 (5) , 2557-2582
Open AccessMapping surface hoar from near-infrared texture in a laboratory
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.
- Open Access
- 29 March 2024
- Published by Copernicus GmbH
- Vol. 2024, 1-36
Reply on RC2
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, however 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 grains distinct shape and orientation. 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 forty-one snow samples; three of which were surface hoar and 38 that 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 resolutions. The case study resulted in the median texture of surface hoar being 1.3 to 8.6 times greater than 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, median classification accuracy with lidar was only 66.99 %. Further, to assess the repeatability of our method and demonstrate mapping capacity, we ran the algorithm on a new sample with mixed microstructures, with accuracy of 99.61 % and 96.15 % for direct and diffuse, 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.
- 29 March 2024
- Published by Copernicus GmbH
- Vol. 2024, 1-36
Open AccessReply on RC2
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, however 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 grains distinct shape and orientation. 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 forty-one snow samples; three of which were surface hoar and 38 that 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 resolutions. The case study resulted in the median texture of surface hoar being 1.3 to 8.6 times greater than 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, median classification accuracy with lidar was only 66.99 %. Further, to assess the repeatability of our method and demonstrate mapping capacity, we ran the algorithm on a new sample with mixed microstructures, with accuracy of 99.61 % and 96.15 % for direct and diffuse, 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.
- Open Access
- 29 March 2024
- Published by Copernicus GmbH
- Vol. 2024, 1-36
Reply on RC1
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, however 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 grains distinct shape and orientation. 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 forty-one snow samples; three of which were surface hoar and 38 that 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 resolutions. The case study resulted in the median texture of surface hoar being 1.3 to 8.6 times greater than 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, median classification accuracy with lidar was only 66.99 %. Further, to assess the repeatability of our method and demonstrate mapping capacity, we ran the algorithm on a new sample with mixed microstructures, with accuracy of 99.61 % and 96.15 % for direct and diffuse, 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.
- 29 March 2024
- Published by Copernicus GmbH
- Vol. 2024, 1-36
Open AccessReply on RC1
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, however 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 grains distinct shape and orientation. 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 forty-one snow samples; three of which were surface hoar and 38 that 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 resolutions. The case study resulted in the median texture of surface hoar being 1.3 to 8.6 times greater than 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, median classification accuracy with lidar was only 66.99 %. Further, to assess the repeatability of our method and demonstrate mapping capacity, we ran the algorithm on a new sample with mixed microstructures, with accuracy of 99.61 % and 96.15 % for direct and diffuse, 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.
- Open Access
- 18 March 2024
- Published by Copernicus GmbH
- Vol. 2024, 1-36
Comment on egusphere-2023-3133
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, however 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 grains distinct shape and orientation. 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 forty-one snow samples; three of which were surface hoar and 38 that 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 resolutions. The case study resulted in the median texture of surface hoar being 1.3 to 8.6 times greater than 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, median classification accuracy with lidar was only 66.99 %. Further, to assess the repeatability of our method and demonstrate mapping capacity, we ran the algorithm on a new sample with mixed microstructures, with accuracy of 99.61 % and 96.15 % for direct and diffuse, 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.
- 18 March 2024
- Published by Copernicus GmbH
- Vol. 2024, 1-36
Open AccessComment on egusphere-2023-3133
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, however 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 grains distinct shape and orientation. 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 forty-one snow samples; three of which were surface hoar and 38 that 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 resolutions. The case study resulted in the median texture of surface hoar being 1.3 to 8.6 times greater than 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, median classification accuracy with lidar was only 66.99 %. Further, to assess the repeatability of our method and demonstrate mapping capacity, we ran the algorithm on a new sample with mixed microstructures, with accuracy of 99.61 % and 96.15 % for direct and diffuse, 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.