Quantification and Analysis of Icebergs in a Tidewater Glacier Fjord Using an Object-Based Approach

Area of Science:

• Glaciology
• Marine Ecology
• Remote Sensing

Background:

• Tidewater glaciers calve icebergs into the ocean, influencing oceanography and providing habitat for marine animals like harbor seals.
• Quantifying glacier ice availability and characteristics in fjords is essential for understanding marine mammal distribution but has not been systematically done.
• Predicted changes in glacier habitats necessitate methods to quantify ice availability in tidewater glacier fjords.

Purpose of the Study:

• To develop and present a novel method using object-based image analysis (OBIA) to classify floating glacier ice from high-resolution aerial imagery.
• To quantify the amount and fine-scale characteristics of floating glacier ice in a tidewater glacier fjord.
• To automate the analysis of floating glacier ice for large image datasets.

Main Methods:

• Employed object-based image analysis (OBIA) to classify floating glacier ice in Johns Hopkins Inlet (JHI) using high-resolution aerial digital imagery.
• Developed specific workflows and rule sets for accurate ice classification.
• Applied the method to a representative dataset from June 2007 to quantify ice cover and characteristics.

Main Results:

• Classified JHI as 45.2% brash ice, 52.7% water, and 2.1% icebergs.
• Calculated an average iceberg size of 5.7 m² and estimated total iceberg habitat area as 455,400 ± 123,000 m².
• Achieved a classification accuracy of 75.6% for icebergs, with errors primarily in densely packed or low-contrast ice areas.

Conclusions:

• Object-based image analysis (OBIA) is an effective tool for classifying floating glacier ice and quantifying ice habitat characteristics.
• The developed method provides a robust approach for assessing changes in glacier ice habitat, applicable to other ice environments like sea ice.
• Accurate quantification of ice habitat is vital for understanding its influence on marine mammal populations and predicting ecological responses to climate change.