Titan Crater Topography Through Stereo

To give a little background, I updated the crater population on Titan by identifying all the craters at the end of Cassini. Then I characterized their morphometry using limited topography in SARTopo, which give thin lines of low resolution topography where overlapping images were taken within a single swath. With this, I define certain parameters to study how the crater was changing. Crater size (diameter) is crucial because the shape of the crater is often presented as a function of diameter. The depth is also important because it gives quantitative constrains on how degraded a crater is from its original (deeper) depths. I took it a step further and measured rim heights as well.

However, the reason I have had to continue studying this is because I have been comparing SARTopo to standard stereo topography. Stereo is similar in that it is acquired where we have entire RADAR swaths overlapping. This provides a 3D look at the crater rather than a single (may 2 or 3) profile through the crater. Furthermore, SARTopo is at ~10km per pixel resolution. Stereo is more like ~2km per pixel. This is relevant because we measured “freshness” of a Titan compared to Ganymede craters. Ganymede is ~same size and density as Titan, so it is expected tat its craters will form in similar manners. The problem is that the topography on Ganymede is ~1.5km per pixel. This is significantly less than Titan, and it becomes a problem when measuring sharp peaks (i.e. low wavelength topography) because lower resolution averages these peaks with surrounding lower topography giving artificially lower rim heights. However, you see that Titan stereo resolution is more on par with Ganymede topography, so we can use it as a test case to see 1) how it compares to SARTopo and 2) if degradation still remains.

Unfortunately, we have come across yet another problem. The Titan stereo results (presented in various Neish et al. papers) was measured using a method that takes statistics of the entire rim and floor values rather than a single profile. I would argue that this is probably the most representative measurement of the crater parameters. However, our comparison is with Ganymede using previously published results from Bray et al. (2012). Brey et al. (2012) averages up to 8 profiles through Ganymede craters; we need Titan stereo measurements that mirror the Bray et al. (2012) methods for it to be comparable.

So, without getting to into the details, I found the 6 craters with stereo (Ksa, Soi, Santorini Facula, Shikoku, Forseti, and Hano). I uploaded the DEM png’s into matlab and manually assigned lat and long values (that was annoying). That may not have been necessary, but I wanted to be able to go back and see where the data was taken. Then I decided to take 8 precise profiles from the crater. I did this by creating a vector (list) of angles from the y-axis then found the right x and y values for each profile. This took a lot of troubleshooting, and I am fairly impressed with my ability to get this done one step at a time. It was particularly difficult where we had partial coverage. I almost just went with the two best craters we had and let it be, but I really wanted to include all the data we had.

The first thing to notice is that not all craters have full coverage, so results are limited to where we do have coverage. I tried to extend up to 5r out (as with Bray et al) where possible. However, I never ended up needing the extra topography because I didn’t go so far as to back out topography. Each black x represents the rim point I found. Then a show 3 of the best examples we have data for with the 8 profiles for Ksa, Soi, and Shikoku where you can see how these values where found. The rims and floor positions are marked with blue marks; the red marks are from previous literature for reference.

This slideshow requires JavaScript.

This slideshow requires JavaScript.


The Final Result

What I have done is re-purposed my previous plot of rim to floor depths (the traditional depth measurement) to include H18 stereo measurements. Ganymede craters are in black. Titan craters are color coded for the 6 stereo craters to compare Neish stereo, H18 stereo, and SARTopo. SARTopo is shown in red if there is no stereo to compare to. What we notice is that there is a 3:2 ratio of craters that have SARTopo depths larger than stereo. Forseti (blue) is lower than SARTopo, well outside the margin out error, but ~same as Neish measurements. Hano (light blue) is lower than SARTopo but more on par than the ~0km depth from Neish et al. Soi (orange) and Ksa (cyan), the ones with the best coverage, show SARTopo slightly below stereo but well within the margin of error for each. While Shikoku (yellow) is below SARTopo, well outside the margin out error. H18 stereo is about the same as Neish et al. for Santorini (green) but slightly above.


Rim to floor depths of Titan (circle, triangle, square) and Ganymede Craters (diamond, line). SARTopo is in circles. Neish stereo in squares. H18 stereo in triangles. The 6 craters in stereo are color coded in both stereo and SARTopo (if available). SARTopo without stereo is shown in red.



Published by

Josh Hedgepeth

PhD student in Geophysics with CPSX at Western University in Ontario, Canada.

One thought on “Titan Crater Topography Through Stereo”

  1. It seems like either the stereo data is within error of the SARTopo data or way lower. I wouldn’t trust those latter results; there is the concern that surrounding topography is being averaged into the rim measurements. In the other three cases, though, they seem to agree reasonably well. I’m not sure if it means something that two of the “problem” craters are in the plains.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s