The BBC News article Isolated lakes found beneath Canadian ice sheet links to the open access Science Advances article Discovery of a hypersaline subglacial lake complex beneath Devon Ice Cap, Canadian Arctic.

The discovery is based on radar "soundings" of the Devon Ice Cap in Nunavut, Canada.

An earlier question What is likely to have happened with these newly discovered antarctic volcanos below the ice? shows sub-surface radar measurements of antarctic ice, in this case to a depth of 1.5 kilometers, and therefore with a total path length of 3 kilometers in ice.

The water molecule is strongly polar and interacts with microwave radiation by absorbing and thermalizing it. This effect is obvious in microwave ovens. Microwave ovens can be used to "defrost" frozen food even a few centimeters thick, using reduced power and longer time. This is necessary because frozen water molecules have a lower absorption than those in unfrozen areas, but they are still quite absorbing.

Radar measurements travel through kilometers in ice round trip, substantially larger than the thickness of a slab of frozen peas.

How is ice-penetrating radar able to work through such long path lengths of ice? Is there a special frequency thats used, or other tricks allowing for extremely weak reflections after bulk attenuation to be detected and measured?

  • $\begingroup$ Related question Help understanding radar data in subglacial lake complex beneath Devon Ice Cap $\endgroup$
    – uhoh
    Apr 15 '18 at 8:35
  • 2
    $\begingroup$ your questions are probably the ones I actually look forward to considering you bring in "groundbreaking" research questions to ES SE. In this case would the tag remote sensing be valid? $\endgroup$
    – gansub
    Apr 15 '18 at 9:08
  • $\begingroup$ lindzey.github.io/blog/2015/07/27/… radar can penetrate through because ice is transparent to electromagnetic energy at radar frequencies $\endgroup$
    – gansub
    Apr 15 '18 at 9:14
  • $\begingroup$ @gansub Thanks, and thanks for the heads-up on the tag as well. If you can find a source better than one sentence in a blog that actually uses the term "mostly transparent" (which you've slightly misquoted), that would be great. Also, it may be prudent to review what the term "radar frequencies" really means in this context. $\endgroup$
    – uhoh
    Apr 15 '18 at 9:35
  • 1
    $\begingroup$ I am not the one answering this ground breaking ground penetrating radar question. That was just some sunday afternoon musings :) $\endgroup$
    – gansub
    Apr 15 '18 at 9:38

Water molecules do interact with microwaves and in microwave ovens that operate at 900MHZ- 2.5GHz. This Radar instrument operates using Very High Freq Radio Waves at 193MHz [1].

At lower-than-microwave frequencies, the loss factor is much lower, and as a result the wave can penetrate deeper into the ice sheet with minimal absorption. The difference in permittivities between liquid water and solid ice is what allows them to visualise boundaries and identify the subglacial lakes

From Nasa [2]

Ice, on the other hand reacts differently depending on the radar's frequency. It reflects high-frequency radio waves, but despite being solid, lower frequency radar can pass through ice to some degree. This is why MCoRDS uses a relatively low frequency—between 120 and 240 MHz. This allows the instrument to detect the ice surface, internal layers of the ice and the bedrock below. "To sound the bottom of ice you have to use a lower frequency," said John Paden, CReSIS scientist. "Too high a frequency and signal will be lost in the ice."


When electromagnetic waves travel through a medium, it can interact with the medium in question. But that interaction depends on the electric properties of the medium. Ice, being a dielectric medium, do interact with the incident e.m. wave from the sensor. The penetration depth is directly proportional to the wavelength of the radiation used (there are different forms in which this depth can be expressed but the easiest for me to remember is Eq 2.14 of the Fundamentals of Remote Sensing by George Joseph; available in Google Book as well):

$l_a=\frac{\lambda_0 \sqrt{\varepsilon'}}{2 \pi \varepsilon''}$

Where $l_a$ is the penetration depth in ice, defined as the depth at wich the signal power falls to $1/e$, $\varepsilon'$ and $\varepsilon''$ are respectively the real and imaginary parts of the dielectric constant, and $\lambda_0$ is the free space wavelength of the electromagnetic radiation.

And so, the microwave radiation (higher wavelength radiations in the electromagnetic spectrum) used in ground penetrating radar can penetrate through the ice surface, the depth thru which it can penetrate being given by the same equation mentioned above. Cheers.

  • $\begingroup$ Good answer! Welcome to Earth Sciences SE. I'v added the equation you mention to your answer, please feel free to edit what I've added or delete it altogether. But I think including the equation does improve the answer. $\endgroup$ Mar 3 '19 at 18:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.