Way back in March I wrote about my upcoming work at LPSC. I recommend you check that out for a full introduction to this, but I will give a quick run down before I get into the major details of this post. Pluto’s crust is thought to be entirely water ice. However, the extremely low temperatures lead to other ices being stable on the surface. The primary ices are carbon monoxide (CO), methane (CH4), and nitrogen (N2). We know ice can be very malleable on geologic timescales (e.g. glaciers on earth, craters on Europa), but the rates of medication vary based on the ice parameters and the conditions of the surface. Water ice on Pluto is extremely rigid and strong, but the same cannot be said for the other ices. They are both more prone to viscous relaxation (i.e. flattening due to gravity), and at least nitrogen is known to cycle from the surface and into the atmosphere (and losing some of its supply in the process). My work posits impact craters as a means to constrain the ways in which these ices would contribute to the modification of craters. As I discuss in my previous post, the two processes primarily effecting impact craters is escape erosion (like nitrogen) and relaxation. The extent to which this modifies the craters will depend on what ice the crater is made of. My work using the degradational state (how shallow is the crater) to constrain what type of ice the crater must be formed in, and we use the surficial compositional data to test (or more aptly, constrain) what the ice is made of. This gives us information about the volatile content in the region of the crater and the history of these volatiles.
In my previous post, we conclude that H2O, N2 and CH4 ices are the most likely to have craters form within them. In my figure I show a serious of possible scenarios. 1) a crater formed in a very thick layer of N2. Over time this would easily relax (flatten) and the N2 is almost certainly lost as well, leaving no trace today. 2) The same situation occurs, but in CH4 ice. This is both stronger and less likely to escape away. Nevertheless, it will still relax on the timescales of the solar system (i.e. Pluto), so we would expect to see a crater formed in CH4 ice that is shallower than expected. 3) Imagine scenario 1 but the crater dips into the bedrock H2O ice, creating a crater formed in an upper layer of N2 ice and a base of water ice. The water ice is too strong to relax, and the N2 will likely be loss. Therefore, we are likely left with a crater that appears to be formed in H2O but is significantly degraded. 4) we reimagine scenario 3 with CH4 instead of N2, and the upper layer doesn’t escape. In fact, we are left with a pristine crater that appears to be formed in CH4 ice. Then, 5) (not shown) we have the standard scenario of a crater formed in pure water ice which would be unlikely to modify at all.
As the title suggests, I am working on finalizing this work and compile it into a publisher manuscript. I had completed most of the work by LPSC. There were two major steps left for me to complete it. 1) I needed to figure out how small to go in the craters I measured; I only measured down to 15 km sized craters because of time constraints. 2) I needed to add a step in the code to remove the terrain slope (large scale topographic variations that craters likely impacted into). This would essentially put the terrain at ~0 km and leave only the crater topography. I did this when I measured Titan impact crater depths, but I didn’t for Pluto do to time constraints. This is a major step because it requires me to redo all the measurements I’ve done. This isn’t hard, just tedious. At 8 profiles for each crater with over 300 hundred craters, that’s thousands of crater profile measurements. Alternatively, I could use the profile positions (assuming I saved them) that I measured for each profile and automate it to find the height at those same positions. I would love some feedback on that idea, but right now I am planning to do it all manually. I have added the step in the code to remove the terrain slope which leaves the step of processing the craters again (one way, or another).
Sadly, we are not done yet. After presenting at LPSC, my conclusions prompted me to consider a major assumption of mine. That is, are the surficial composition measurements reflective of the underlying ice? The surface is covered with material, including surficial ices. I posit in this work that there are only so many possible formations a crater can form, and it speaks to the type of ice it needs for that crater shape to be viable. The compositional data is intended to act as further confirmation, with limited reach. Nevertheless, I wanted a way to demonstrate this is a fair assumption, so I started to consider what type of measurements I could take to test this claim. Let’s take another look at my figure above.
My figure demonstrates where we would expect to measure the highest amounts of the ice the crater is formed in with orange highlighter. In the areas of orange strips, this is expected to be covered, at least in part, but surficial deposits. My work at LPSC considered the crater composition of the rim to rim. That is, we would expect to include some of the purist and most biased regions of the crater. If I want to test whether these measurements are reflective of the crater ice layer, I should be able to measure the composition in each region and show the rims are richer in the predicted ice than the floor of the crater. With 300+ craters, the big question is how do I do that? I could map precise regions in ArcGIS, but that would take so much time. Still, it is likely the most accurate approach. The alternative, that I am currently working on, is to import the data to MATLAB. I can automate the process and take measurements of set sizes around the rims (10%, 20%, etc). Except, I don’t know how wide to make this. Nevertheless, I am currently in the process of doing this. The other option is to process the eight topographic profiles and mark where I want the rim and flow to be measured. This would take about as much time as in ArcGIS. I wish I could just do that with the points I use to take the depth measurements, but the region where I tell it to look for the peak rim is not necessarily what I would constitute the entire rim. Why? Simply put, profiles get really weird.
Now, here I am. The main purpose of this post is to think through what I am doing and request feedback. Although, after writing this I am beginning to think the best option is to map these regions in ArcGIS but only do it for the largest 100 craters (or some sufficiently large sample size). I don’t necessarily have to test all the craters, but in this scenarios I might want to focus on the craters richest in N2 and CH4, seeing as these are the ones where the assumption applies. Now I am very disheartened because I’ve spent days working to do this process in MATLAB, and I am seriously considering switching to ArcGIS because I literally talked myself into it. Let me know what you think!