ImpDAR:an open-source impulse radar processor

Despite widespread use of radio-echo sounding (RES) in glaciology and broad distribution of processed radar products, the glaciological community has no standard software for processing impulse RES data. Dependable, fast and collection-system/platform-independent processing flows could facilitate co...

Full description

Bibliographic Details
Published in:Annals of Glaciology
Main Authors: Lilien, David A., Hills, Benjamin H., Driscol, Joshua, Jacobel, Robert, Christianson, Knut
Format: Article in Journal/Newspaper
Language:English
Published: 2020
Subjects:
Online Access:https://curis.ku.dk/portal/da/publications/impdar(5708c364-a6b3-400f-82d5-fb88f3a8513d).html
https://doi.org/10.1017/aog.2020.44
https://curis.ku.dk/ws/files/248547856/impdar_an_opensource_impulse_radar_processor.pdf
Description
Summary:Despite widespread use of radio-echo sounding (RES) in glaciology and broad distribution of processed radar products, the glaciological community has no standard software for processing impulse RES data. Dependable, fast and collection-system/platform-independent processing flows could facilitate comparison between datasets and allow full utilization of large impulse RES data archives and new data. Here, we present ImpDAR, an open-source, cross-platform, impulse radar processor and interpreter, written primarily in Python. The utility of this software lies in its collection of established tools into a single, open-source framework. ImpDAR aims to provide a versatile standard that is accessible to radar-processing novices and useful to specialists. It can read data from common commercial ground-penetrating radars (GPRs) and some custom-built RES systems. It performs all the standard processing steps, including bandpass and horizontal filtering, time correction for antenna spacing, geolocation and migration. After processing data, ImpDAR's interpreter includes several plotting functions, digitization of reflecting horizons, calculation of reflector strength and export of interpreted layers. We demonstrate these capabilities on two datasets: deep (similar to 3000 m depth) data collected with a custom (3 MHz) system in northeast Greenland and shallow (