Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM69061
Volume:
10 |
11 |
Nov 2026
Pages:
1-18
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.

Department of computer science engineering (Artificial Intelligence & Data Science)
Vivekananda Global University (VGU), Jaipur

B.Sc. (Hons.) Agriculture
Vivekananda Global University (VGU), Jaipur

Department of computer science engineering
Vivekananda Global University (VGU), Jaipur

Department of Management
Vivekananda Global University (VGU), Jaipur

Department of Bsc
Vivekananda Global University (VGU), Jaipur

Department of Computer Science & Engineering
Vivekananda Global University (VGU), Jaipur
Satellite imagery may aid in the repeat mapping of snow/glacier surfaces, though the presence of clouds, shadowed areas, pixel mixing and unclear ground reference labels presents challenges in the context of the Himalayan landscape. In this paper, we propose a reproducible deep learning approach for our intended regional investigation at Chhota Shigri Glacier, Himachal Pradesh, India. Our approach incorporates Sentinel-2 L2A reflectance data, glacier inventory context and topographic height data on an aligned 20 m resolution grid. We specify a minimal U-Net architecture that segments visible snow/glaciers and other surface types, while uncertain or obscured pixels are ignored in the process of supervised assessment. We use the Normalized Difference Snow Index as a conventional baseline for the binary task derived from the union of snow and clean-ice surface labels. Geographic blocks and time-series held out satellite acquisitions are employed in order to control for information leakage, in addition to normalization solely based on the training samples. Class-specific segmentation accuracy metrics and area calculation accounting for temporality establish the link between prediction quality and interpretation. The contribution here is in methodological design, while empirical results, satellite-based area measurements and comparisons remain pending until imagery and trained network labels become.
KEYWORDS
This manuscript describes a replicable methodology to study visible snow cover dynamics and clean ice exposure through Deep Learning using Sentinel-2 images and geospatial information. This connects a small U-Net architecture with an explicit NDSI benchmark, stand-alone image classification and withheld area calculation based only on feasible observations. Contribution is a precisely described regional analysis including distinct demarcation between surface properties, glacier extent and mass balance. The framework is still not proven through empirical testing. Superiority of the models, accuracy of segmentation and changes in the temporal area are thus open questions, and there is no inference made about environmental trends. What is left to do is real image acquisition, mask verification, splitting for evaluation and storage of the model predictions. Future work could include more areas, debris covered glaciers, additional terrain features or field observations independent of the first test.
[1] S. Gascoin, M. Grizonnet, M. Bouchet, G. Salgues, and O. Hagolle, “Theia Snow collection: high-resolution operational snow cover maps from Sentinel-2 and Landsat-8 data,” Earth System Science Data, vol. 11, pp. 493-514, 2019. doi: 10.5194/essd-11-493-2019. [2] R. Aberle et al., “Automated snow cover detection on mountain glaciers using spaceborne imagery and machine learning,” The Cryosphere, vol. 19, pp. 1675-1693, 2025. doi: 10.5194/tc-19-1675-2025. [3] K. A. Maslov, C. Persello, T. Schellenberger, and A. Stein, “Globally scalable glacier mapping by deep learning matches expert delineation accuracy,” Nature Communications, vol. 16, Art. no. 43, 2025. doi: 10.1038/s41467-024-54956-x. [4] M. Drusch et al., “Sentinel-2: ESA’s Optical High-Resolution Mission for GMES Operational Services,” Remote Sensing of Environment, vol. 120, pp. 25-36, 2012. doi: 10.1016/j.rse.2011.11.026. [5] O. Ronneberger, P. Fischer, and T. Brox, “U-Net: Convolutional Networks for Biomedical Image Segmentation,” in Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015, LNCS, vol. 9351, pp. 234-241, 2015. doi: 10.1007/978-3-319-24574-4_28. [6] D. K. Hall, G. A. Riggs, and V. V. Salomonson, “Development of methods for mapping global snow cover using moderate resolution imaging spectroradiometer data,” Remote Sensing of Environment, vol. 54, no. 2, pp. 127-140, 1995. doi: 10.1016/0034-4257(95)00137-P. [7] X. Chu, X. Yao, H. Duan, C. Chen, J. Li, and W. Pang, “Glacier extraction based on high-spatial-resolution remote-sensing images using a deep-learning approach with attention mechanism,” The Cryosphere, vol. 16, pp. 4273-4289, 2022. doi: 10.5194/tc-16-4273-2022. [8] D. R. Roberts et al., “Cross-validation strategies for data with temporal, spatial, hierarchical, or phylogenetic structure,” Ecography, vol. 40, no. 8, pp. 913-929, 2017. doi: 10.1111/ecog.02881. [9] F. Paul et al., “On the accuracy of glacier outlines derived from remote-sensing data,” Annals of Glaciology, vol. 54, no. 63, pp. 171-182, 2013. doi: 10.3189/2013AoG63A296. [10] M. F. Azam et al., “Reanalysis of the longest mass balance series in Himalaya using a nonlinear model: Chhota Shigri Glacier (India),” The Cryosphere, vol. 18, pp. 5653-5672, 2024. doi: 10.5194/tc-18-5653-2024.
Publication Date
11/16/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
1
Total views since publication
PDF Downloads
3
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...