Automated Polar Ice Thickness Estimation From Radar Imagery

Abstract

This work focuses on automating the task of estimating Polar ice thickness from airborne radar data acquired over Greenland and Antarctica. This process involves the identification and accurate selection of the ice sheet's surface location and interface between the ice sheet and the underlying bedrock for each measurement. Identifying the surface and bedrock locations in the radar imagery enables the computation of ice sheet thickness, which is important for the study of ice sheets, their volume, and how they may contribute to global climate change. The timeconsuming manual approach requires sparse handselection of surface and bedrock interfaces by several human experts, and interpolating between the selections to save time.

Accurate, Dynamic, and Distributed Localization of Phenomena for Mobile Sensor Networks

“…Through these radars, the robots identify layers and areas of potential water at the interface between the ice sheet and the underlying bedrock. Together, these remote sensing techniques allow us to study and further understand how polar regions are changing in response to climate change [Gifford et al 2010]. Finally, there are systems that combine robotics with remote sensing systems to autonomously image the subsurface for polar and planetary applications [Gifford et al 2009].…”

“…Many strategic MSN applications are used to improve centralized or decentralized coverage of ROIs [Gaojun and Shiyao 2010], [Gifford et al 2010], through dynamic deployment algorithms [Dhillon and Chakrabarty 2003, Wang et al 2009]. Most of them focus on optimal deployment of the nodes in order to maximize the coverage performance.…”

Estimating bedrock and surface layer boundaries and confidence intervals in ice sheet radar imagery using MCMC

“…In other work, Ferro and Bruzzone used echograms of the Martian subsurface to detect basal returns. The subglacial identification problem was studied by Gifford et al [2], who compared two primary approaches, namely an active contour ('snake') model and an edge-based technique. Ilisei et al [5] developed a two-phase technique to exploit the properties of a radar signal for generating a statistical map and applying a segmentation algorithm.…”

“…The model defined by equations (1), (2), and (4) is a firstorder Markov Random Field. Unfortunately, finding the values of L that maximizes equation (1) is NP-hard in the general case [14].…”

“…A few recent papers have studied how to use image processing and computer vision techniques to determine layer boundaries automatically or semi-automatically from echograms [2,3,4,5,6,7,8,9,10], but this is a hard problem because of the high degree of noise, the often faint layer boundaries, and confusing linear structures caused by signal reflections and clutter. In fact, even human annotators produce diverging estimates of the boundaries in many cases.…”