Mapping debris-covered glaciers in the Cordillera Blanca, Peru : an object-based image analysis approach.

Abstract

Biddle,, Donald J. 1982-, "Mapping debris-covered glaciers in the Cordillera Blanca, Peru : an object-based image analysis approach." (2015). ACKNOWLEDGEMENTSThere are so many people without whose guidance and support this thesis would not have come to fruition. To thank them all by name would be major work in its own right! If I must pare the list for this limited space it should begin with my advisors, Dr. Keith Mountain and Dr. Andrea Gaughan. They challenged me to be the best version of myself, even -nay -especially when it was not the easiest way forward. Thank you also to Dr.Michael Croasdaile for putting his geographer's hat back on to see me through this process, and providing an invaluable outside perspective. A very special thank you to Dr.Carol Hanchette, for her patient encouragement over the (many) years it has taken to reach this point. She reminded me that yes, I can do this. Thanks to Forrest Stevens for the crash course in R coding and machine learning that led to a major breakthrough that helped push me over the hump with this project. My brain still hurts. To my lovely wife Nicole, thank you for putting up with me through the deepest moments of frustration, and being a constant companion, confidante, and comedy sidekick…not just now, but always. Accurate remote-sensing based inventories of glacial ice are often hindered by the presence of supraglacial debris cover. Attempts at automated mapping of debris-covered glacier areas from remotely-sensed multispectral data have met with limited success due to the spectral similarity of supraglacial debris to nearby bedrock, moraines, and fluvial deposition features. Data-fusion approaches leveraging terrain and/or thermal data with multispectral data have yielded improved results in certain geographic regions, but remain unproven in others. This research builds on the data-fusion approaches from the literature and explores the efficacy of object-based image analysis (OBIA) and tree-based machine learning classifiers using Landsat OLI imagery and SRTM elevation data, in effort to map debris-covered glaciers in the Cordillera Blanca range of Peru. Results suggest that the OBIA and machine learning methods render advantages over traditional methods given the unique morphological settings associated with debris-covered glaciers.

Glacier Facies Mapping Using a Machine-Learning Algorithm: The Parlung Zangbo Basin Case Study

“…Additionally, an analysis of the results clearly revealed that the proportion of small size glaciers (<1 km 2 ) were 88.3% distributed at lower elevations compared to larger size glaciers (≥1 km 2 ). In addition, the majority of glaciers (both in terms of glacier number and area) were characterized by a mean slope ranging between 20 • and 30 • , and 42.1% of glaciers had a northeast and north orientation in the Parlung Zangbo basin.In the past few decades, much work has been accomplished to map the extent of clean glacial ice and to quantify changes over time using satellite image data [6]. Methods applied range from visual interpretation [7] to segmentation of band ratio or spectral indices (e.g., the Normalized Difference Snow Index) images [8] and different unsupervised (e.g., the Iterative Self Organizing Data Analysis Techniques Algorithm, ISODATA) [9] and supervised (e.g., the Maximum Likelihood algorithm) classification [10] and decision tree methods [5,11].…”

“…In the past few decades, much work has been accomplished to map the extent of clean glacial ice and to quantify changes over time using satellite image data [6]. Methods applied range from visual interpretation [7] to segmentation of band ratio or spectral indices (e.g., the Normalized Difference Snow Index) images [8] and different unsupervised (e.g., the Iterative Self Organizing Data Analysis Techniques Algorithm, ISODATA) [9] and supervised (e.g., the Maximum Likelihood algorithm) classification [10] and decision tree methods [5,11].…”

Development of glacier mapping in Indian Himalaya: a review of approaches

“…Optical, Thermal and DEM derived parameters) while working on near homogeneous objects. Therefore, several studies could successfully demonstrate the delineation of debris cover glaciers within in an object-based environment (Rastner et al 2014;Bajracharya, Maharjan, and Shrestha 2014;Bajracharya and Shrestha 2011;Robson et al 2015;Li, Bao, and Huang 2013;Biddle 2015;Jawak, Jadhav, and Luis 2016). (Rastner et al 2014) presented a comparative study of PBIA and OBIA for debris cover classification in three test regions, namely: Coasta Mountains of Canada, Watkin range of Greenland, and Nepal Himalaya.…”

Novel Machine Learning Method Integrating Ensemble Learning and Deep Learning for Mapping Debris-Covered Glaciers

“…A technique to effectively measure and monitor mountain glaciers is crucial at this stage [16][17][18]. However, in high mountain areas, which experience frequent cloud activity and have a large amount of debris coverage, automatic and semi-automatic glacier mapping based on remote sensing images is challenging [1,11,20], and short-term studies such as these do not necessarily show stable changes over time. Debris-covered glaciers are a special type of glacier [21,22].…”