Annabel Conger: Subglacial Lake-Floor Characteriterization in Radar Data: From Archival to Modern

Annabel Conger, Emma (Mickey) MacKie, Dustin Schroeder

In the past fifty years, advances in radio-echo sounding (RES) collection and processing coupled with an  increasing amount of available data have allowed the community to identify hundreds of Antarctic subglacial lakes. Although the bright, smooth, and flat appearance of a lake surface makes their detection by RES fairly easy, lake floors are not typically detectable with radar. As a result, the depth and water composition of subglacial lakes are poorly constrained. Here we investigate historic and modern Antarctic datasets for the presence of long-overlooked reflections from subglacial lake floors. 

To investigate subglacial lakes and potential lake floor reflections, we use recently digitized radargrams recorded on 35 mm film. The data was collected in Antarctica in the late 60s and early 70s in a series of RES flyovers known as the SPRI-TUD-NSF collaboration (Scott Polar Research Institute, Technical University of Denmark, US National Science Foundation). The increased resolution in the digitized data allows us to radiometrically characterize the film’s ‘Z-scopes’ (radargram) and ‘A-scopes’ (power return over time) to analyze the potential penetration of the water bodies. Previous analysis of the dataset identified eight potential lake floor reflectors, however, water conductivity was only calculated for one lake and no temporal conclusions with modern data were drawn. We use echo strength to constrain water conductivity and chemistry for the additional lakes and conduct proof-of-concept analyses to determine which radar system parameters and processing techniques are able to detect lake floors. We conducted a cross-platform analysis with modern CReSIS and HiCARS radar data of spatially-similar lakes to determine which lake conditions allow for floor reflector detection by different systems and investigate lake depth and compositional changes over the decades.

Click HERE for the poster or HERE to view the presentation

9 Comments on “Annabel Conger: Subglacial Lake-Floor Characteriterization in Radar Data: From Archival to Modern

  1. Great presentation Annabel,

    Now that you’ve done done an analysis of modern and archival radar film. If you had to pick one system to go image subglacial lakes, which radar system would you choose to use to collect new data in hopes of seeing the base of the lakes?

    Great work again,
    – Dusty

    • Thank you Dusty!

      And you raise a good question. Since we are not 100% sure that the modern system even saw the reflectors I am hesitant to fully endorse them but I don’t think you can ignore the benefits of the digital system and the advancements we’ve made in the last decades. At this point, I would be very interested in seeing a modern system but at the same frequency (60MHz) as the SPRI radar, especially if you flew the same flight lines and could directly compare the two. Using the same frequency would show you if the difference between the two is frequency dependent or some other intricacy of the system and hopefully the newer digital system would give us additional precision in the measurements. It is certainly an open question though, especially because we haven’t yet isolated frequecy as the most important difference between the systems.

  2. Wow, great presentation and fascinating work, Annabel! You did a really terrific job explaining how useful radar is to understand these features. In particular, I think it’s amazing that you can use it to determine the chemical purity of the subglacial water. Could radar be used to determine the biological purity, i.e., if many microbial cells are in the water or not? Having a way to detect that remotely would be tremendously useful, and could guide further investigations on earth and beyond.

    • Hi Anne! Thank you so much.

      It’s true that being able to detect microbial cells from above the ice would be huge! Unfortunately, at this point I don’t think we can isolate biological purity from the measurments. There are too many unknowns about the water bodies and the data just isn’t precise enough at this point. You can extrapolate a little from the water purity (super-pure, isolated, water, which much of this is, tends to be less likely to host life forms) but as of right now I don’t know how we could seperately calculate biological purity. However, I won’t rule out the possibility that future work with newer, better systems might have a better answer for you!

  3. Annabel – very nice. I learn something new every time I hear or read about your work. I liked how you explained the two different systems and how you can use modern data to understand the film data. Congratulations!

  4. Annabel, thanks for (trying to) address the side-lobe issue. Pls explain the horizontal scale in the half-bandwidth and full bandwidth plots – are these sample numbers? Could you show all three on a plot at constant time-scale? That might help me understand your claim that the “bottom” reflection is uniformly spaced from the “top reflection”.
    Also, how do you distinguish side-wall from bottom reflections?

  5. Dear Annabel,

    I feel like I’ve learned something about what a good presentation is seeing your work – it was very clear and carefully explained. My question is about the particular computer vision techniques you used: I am not familiar with this field, but I’d love to hear more details about your approach and how it works.

  6. Hey Annabel,
    Very cool project! I noticed in the radar sections that ice layering seems to disappear, or at least decrease, the closer you get to the lake floor. Is there any proposed or known explanation for this? Perhaps related to compression or freeze/thaw cycles?

Leave a Reply

Your email address will not be published. Required fields are marked *