Gaddam, VK and Ele, SL and Bhandari, S and Raavi, S and Kulkarni, AV and Ranjan, R (2024) Applications of Machine Learning Algorithms via Google Earth Engine Interface to Interpret Snowline Altitudes: A Case Study in Chandra Basin. In: 3rd International Conference on River Corridor Research and Management, RCRM 2023, 15 June 2023through 17 June 2023, Virtual, Online, pp. 243-264.
Full text not available from this repository. (Request a copy)Abstract
Cryospheric components are sensitive to changes in climate and monitoring them is a challenging task in the rugged topography of the Himalaya. Many studies were carried out to assess the changes in Cryosphere like glacial retreat, mass balance, changes in runoff and weathering effect in glaciated valleys. But very few studies have been carried out on the assessment of Snow Line Altitude (SLA), which is an important parameter to understand the sensitivity of glaciers to climate change. Mapping and monitoring the SLAs of glaciers using in situ methods is a crucial task since it is a zonal feature and has high spatio-temporal variability. Remote sensing-based feature tracking is the alternative way for mapping the snowline altitudes of glaciers. With the recent development of various machine learning algorithms, segregation of identical features on the multispectral imagery and tracking the position of SLA�s has become easy. Few of them are Otsu, K-means, cascade K-means, random forest, minimum distance, smile cart and support vector machine algorithms. Thus, this study aims to understand the applicability of machine learning algorithms to map and analyse the variations in SLAs of glaciers in the Chandra River basin, a tributary of the Sutlej River. During the analyses, SLAs are mapped and snowline altitudes are extracted. It indicates that the SLAs of glaciers vary from 4100 to almost 5260�±�30 m. All the glaciers are completely covered with seasonal snow during the accumulation period starting from November to April, and then snowline starts retreating from the lower altitude i.e., 3800 m.a.s.l, and reaches a maximum altitude by the end of the ablation period (in this study it is 5259�±�30 m.a.s.l). Analyses and observations suggest that all the applied methods performed similarly except the support vector machine and minimum distance methods. Estimated uncertainties be within�±�30 m. The regional SLA is observed higher during the hydrological year 2003�04 (estimated approximately to 525 m) and vice versa in the years 2019�2020 (where SLA is estimated to 5206 m). The observations on seasonal snow extents and altitudes of snowlines from this study can be used to reconstruct the mass balance and the hydrological budget of the Chandra river basin. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.
| Item Type: | Conference Paper |
|---|---|
| Publication: | Lecture Notes in Civil Engineering |
| Publisher: | Springer Science and Business Media Deutschland GmbH |
| Additional Information: | The copyright for this article belongs to Springer Science and Business Media Deutschland GmbH. |
| Keywords: | Climate change; Forestry; K-means clustering; Learning algorithms; Mapping; Remote sensing; Surface roughness; Topography; Uncertainty analysis, Cascade K mean; Chandra; Chandrum basin; K-means; Machine learning algorithms; Otsu; Snow line altitude; Snowline; Support vectors machine, Support vector machines |
| Department/Centre: | Division of Mechanical Sciences > Divecha Centre for Climate Change |
| Date Deposited: | 22 Sep 2024 06:16 |
| Last Modified: | 22 Sep 2024 06:16 |
| URI: | http://eprints.iisc.ac.in/id/eprint/85028 |
Actions (login required)
 |
View Item |
