Abstract

Abstract. This study presents a new method to derive centerlines for the main branches and major tributaries of a set of glaciers, requiring glacier outlines and a digital elevation model (DEM) as input. The method relies on a "cost grid–least-cost route approach" that comprises three main steps. First, termini and heads are identified for every glacier. Second, centerlines are derived by calculating the least-cost route on a previously established cost grid. Third, the centerlines are split into branches and a branch order is allocated. Application to 21 720 glaciers in Alaska and northwest Canada (Yukon, British Columbia) yields 41 860 centerlines. The algorithm performs robustly, requiring no manual adjustments for 87.8% of the glaciers. Manual adjustments are required primarily to correct the locations of glacier heads (7.0% corrected) and termini (3.5% corrected). With corrected heads and termini, only 1.4% of the derived centerlines need edits. A comparison of the lengths from a hydrological approach to the lengths from our longest centerlines reveals considerable variation. Although the average length ratio is close to unity, only ~ 50% of the 21 720 glaciers have the two lengths within 10% of each other. A second comparison shows that our centerline lengths between lowest and highest glacier elevations compare well to our longest centerline lengths. For > 70% of the 4350 glaciers with two or more branches, the two lengths are within 5% of each other. Our final product can be used for calculating glacier length, conducting length change analyses, topological analyses, or flowline modeling.

Highlights

  • Glacier centerlines are a crucial input for many glaciological applications

  • We present a new algorithm that allows for deriving multiple centerlines per glacier based on a digital elevation model (DEM) and outlines of individual glaciers

  • We have developed a three-step algorithm to calculate glacier centerlines in an automated fashion, requiring glacier outlines and a DEM as input

Read more

Summary

Introduction

Glacier centerlines are a crucial input for many glaciological applications. Glacier length, derived from centerlines, is an important parameter for glacier inventories (Paul et al, 2009). The few studies that derive centerlines fully automatically use a hydrological approach and/or derive only one centerline per glacier (Schiefer et al, 2008; Le Bris and Paul, 2013). We present a new algorithm that allows for deriving multiple centerlines per glacier based on a digital elevation model (DEM) and outlines of individual glaciers. This algorithm splits the centerlines into branches and classifies them according to a geometry order. We carry out a quality analysis by visual inspection of the centerlines, and compare the derived lengths to the lengths obtained from alternative approaches

Objectives
Methods
Results
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call