We’ve been reviewing the output coming from the full Planet Four data reduction pipeline now that we’ve frozen development on the fan and blotch clustering codes. Once we have the fan markings and blotch markings clustered individually, we then have a stage that combines the individual clusters to decide it a source marked by Planet Four volunteers is really a blotch or fan by find clusters where there centers on top of each other and then depending on how many fan markings went into the fan cluster and how many markings went into the blotch cluster we decide it’s a fan or a blotch for the final catalog. What we found in the catalog review that there are nice cases where there are sources that aren’t quite a fan only or not just a blotch. With a citizen science approach we’re able to capture that fuzziness which is fantastic. We highlight a few examples below selected by Michael Aye, who has been hard at work developing this pipeline over the past several years.
In all three figures: the top left is the Planet Four subject image, top middle is all the individual volunteer fan markings, top right is all the individual volunteer blotch markings, the bottom right is the blotch clusters after clustering, the middle bottom is the fan clusters after clustering, and the bottom left is the final sources after combining the fans and blotches (the dots in this panel show the center position of the final fan or blotch in our catalog).
As you can see from above, we’re making great progress on the Planet Four data reduction pipeline. Next steps including handing the fact that the edges of most of our Planet Four subject images overlap with neighboring subject images, and ensuring that we merge overlapping volunteer markings covering the same spot on two different subject images.
I’ve been reviewing output from the blotch clustering with updated parameters, and the good news is the two problematic cases we were trying to have the clustering algorithm identify now are successful:
We’re looking a bunch of parameters, but the bottom left corner configuration is the parameters we think are best for the blotches and and fans.
Here we are showing just the clustering of the fan markings.
Here we are showing just the clustering of the blotch markings
Here we show the clustering of the blotch markings. Now the 2-scale blotch clustering identifies the volunteer draft markings for the large blotch below as well as smaller ones.
I think we’re at the point of locking development of the blotch pipeline. This will be a big step forward towards finishing the first paper and getting science results out of Planet Four. We’ll want to take a look at the next stage of the pipeline after two different types of markings are cluster and the algorithm picks the final shapes based on the number of classifiers who drew fans versus blotches for the same source. If that looks good as we expect it should from previous reviews of the pipeline output, then the next thing we need to do is look at the pipeline results near the edges of each of the cutouts where this is overlap between the different subject images.
A quick update on the Planet Four: Terrains paper. For the past month or so the team has been working on making changes to the manuscript and creating new figures to address the concerns of the two independent referees who reviewed the paper. The referees are experts in the field who assess and critique the paper. Having an independent set of eyes give feedback is useful and makes the paper better. We’ve submitted the revised draft on May 16th with a list of each of the changes we made to address the points raised by the referees. We’re waiting to hear back from the journal. We hope that after this first round of review/edits that there will be only minor changes requested going forward. We’ll have to wait for the referees’ to read the revised manuscript and our report and send their assessment to the editor. Fingers crossed for a speedy review.
As a teaser, below is a new figure we made for the paper. The image is a subframe from a HiRISE observation of one of the regions targeted based on your classifications on Planet Four: Terrains. You can see that this area is like Inca City where we see fans emanating from a top the ice sheet where boulders are embedded/below the ice sheet. Not all the boulders exhibit seasonal fan/carbon dioxide jet activity when this image was taken
The science team is making great progress towards freezing developing of the Planet Four clustering algorithm. I reviewed some of the output from the pipeline Michael Aye has been writing. Basically the task was to check on the few issues we were working on addressing by having a two size regime clustering for blotches drawn by volunteeers and pick the parameters that seemed to work best for the data.The good news is we see an improvement.
I thought I’d share some of then plots so you so you can see how close we are to finalizing the pipeline. These plots are at the stage of clustering all the blotch markings alone and then clustering all the fan markings alone. We combine the fans and blotch markings into one later on in the process. For now we’ve just run the first part of the clustering pipeline and outputted the results to these figures. As you can see we’re doing pretty well at picking up all the fans and blotches marked by the majority of the classifiers who made a marking on the subject image.
We’ve got one or two more tweaks we brainstormed in the last science team call last week, and once we review those I think we’ll be freezing development on this part of the Planet Four analysis pipeline until after the first paper is submitted.
We wanted to give a quick update on the original Planet Four. Michael Aye has been leading the development of the data analysis pipeline. As previously mentioned, we’ve hit a major milestone with completing the fan clustering algorithm for combining your classifications together. We think we’ve now hit that point for finalizing the blotch clustering algorithm.
We think we’ve now got a decent solution for addressing how to cluster very large blotches that take up half the image and very small blotches that are the default blotch circle size. Currently how we’re tackling this is clustering with one linking radius for the center of the blotch markings, and then we run the analysis again using a much larger linking radius. Here’s an example output:
This blotch clustering strategy seems to be a good compromise for our science goals and needs. We’re going to review several more test cases and if all goes well with this step, we will freeze development on the clustering pipeline. That’s one of the last hurdles to applying the pipeline to all of your classifications and dive into what the shapes and sizes and directions of the fans and blotches tell us about the seasonal carbon dioxide jet process and the surface winds in the Martian South Polar region.
We wanted to give a quick update on Planet Four. Our main focus has been to get a data reduction pipeline that robustly clusters all the volunteer drawn markings of each subject image together to identify the seasonal fans and blotches and based on the majority shape select decide if the feature is a fan or a blotch. Michael Aye has been leading this effort. We’re pleased to say that that the main fan identification portion of the analysis pipeline is complete. We still have a few more things Michael has been working on for the blotch identification part. We think we’ve come up with a decent solution for identifying small and very large blotches. We hope to have this part of the analysis pipeline finalized soon. Then we will be able to apply the pipeline to all of your classifications and dive into what the shapes and sizes and directions of the fans and blotches tell us about the seasonal carbon dioxide jet process and the surface winds in the Martian South Polar region.
A quick update on all things Mars or at least all things Planet Four.
We got the referees’ reports back from the Planet Four: Terrains paper. The journal set it to two experts in the field. The read the paper and provided a critique of the paper. The reviewers gave positive feedback and have questions and concerns for us to address as well as other more minor requested changes to the manuscript. These additions and changes will improve the paper. So over the coming weeks, the science team will make modifications additions to the paper draft over the coming weeks and we hope to have it back in to the journal as soon as possible. Then our written response to the referees’ report and the updated manuscript will go back to the referees for their second look. We’ll keep you posted as we make more progress. In the meantime, there are new images we have uploaed on the Planet Four: Terrains website in need of review.
In regards to Planet Four: Ridges, thank to your help we’re completed 100 CTX images of our second search area. We’re currently working on getting new images onto the site. The CTX images are being processed as we speak and cut up into the subimages we need for the website. The images should hopefully be uploaded over the next week or so. Stay tuned to this space for more updates.
For Planet Four, we’re really at the stage of making the last changes and tweaks to the data analysis pipeline and switching gears to working on finishing the paper draft. We’ll have a separate blog post on that in the coming weeks.
Today we have a guest blog by JPL research scientist Laura Kerber, one of our lead researchers on Planet Four: Ridges . Laura studies physical volcanology, aeolian geomorphology, wind over complex surfaces, and the ancient Martian climate,
Hello Ridge-Hunters! We have been finding lots of ridges in Nilosyrtis Mensae, and I wanted to give you a bit of an update on our progress. Here is a map showing the images that we have looked through (blue), the places where I thought there might be ridges before the project started (circled in orange), and the spots where you have actually found ridges (purple dots).
As you can see, most of the ridges were found in the southeast portion of the search area. I took a look at the outliers, and they aren’t the kind of polygonal ridges we are looking for—meaning that all of the polygonal ridges we found have been in a pretty restricted area. Here is a close-up of that area:
There are a couple of important things that we have already learned from what we’ve found so far. First, we can see that the ridges aren’t correlated with craters. One of the early theories about these ridges was that they were breccia dikes—that is, dikes of broken-up material that was forced through surrounding terrain during a violent impact event. The presence of polygonal ridges both in craters and on the inter-crater plains makes this hypothesis seem less likely.
Here are some great ridges that you found on the intercrater plains:
And an even closer close-up:
We also want to know whether or not the ridges are correlated with valley features. At first glance, it looks like valleys and ridges aren’t correlated, because there are plenty of ridges in the inter-valley plains, and valleys like Auqakuh Vallis that don’t have a lot of ridges near them:
Upon closer inspection, we can see that the ridges are correlated with what we call an “etched” terrain—terrain that has been heavily eroded, leaving bits and pieces of the terrain that came before it. The southern part of Auqakuh Vallis is dominated by etched terrain, and we can even see that part of the valley has been inverted by erosion (what was once the valley floor is now standing higher than everything around it). We can also see that the western branch of Auqakuh Vallis has cut this positive feature, meaning that it was active long after the eastern branch stopped flowing. There were a lot of ridges identified both surrounding the river deposits that make up the top of the inverted Auqakuh Vallis channel and around it. This may suggest that ridges are preferentially forming in old river sediments:
But why aren’t there ridges further north along Auqakuh Vallis?
Actually… there are! Here is an image further north along the Vallis. We can see that northern Auqakuh Vallis cuts through a ridge-containing unit, but in most of the surrounding area, the ridge-containing unit is capped by a unit with glacial morphology that hides the ridge unit from view:
Our current hypothesis: The ridge unit formed before or at the same time as the valleys were being cut. Afterwards, glaciers and ice sheets covered the area and deepened and widened the valleys. The glaciers covered the northern Auqakuh Vallis region and most of the terrain north of it, including the western part of the study region.
The next group of CTX images extends our search area to the east. This is the area where this type of polygonal ridges were first mapped, before we had CTX images covering the entire area like we do now. The first mapping project (in 2006) identified ridge lattices inside mostly inside craters, leading to the hypothesis that they were impact-related breccia dikes. The second project (in 2013) mapped ridges along the Nili Fossae trough system, leading those scientists to hypothesize that the original fractures may be related to the trough system. Our study of the ridges to the west has been offering an expanded context for these hypotheses. The other special thing about this region is that we will be covering two of the three remaining potential landing sites for the next NASA Mars rover, called “Mars 2020”. Mars 2020 is carrying a suite of instruments that it will use to search for habitable places on Mars as well as organic material. The new rover will also carry a drill that it will use to take samples of many different rocks and cache them in tubes for a future mission to bring back to Earth. Wouldn’t it be great if they could bring a bit of ridge back for us?
We want to share a quick update on the depth and details we are investigating to close out the last issues for our analysis pipeline for identifying the fans and blotches from the classifications from the original Planet Four.
We recently realized that it might be a good idea to allow different limits for clustering depending on the general marking size. Intuitively this seems to make sense as one automatically takes a bit more care the smaller an object is to mark.
However, more clustering versatility also means more parameters to set which need to be tested for their efficacy.
Below you can see two plots, one for “fan” markings, the other for “blotch” markings, that show different parameter settings for a clustering run and their effects on the final result.
This planet four image tile with the ID ’17a’ is one of the more problematic ones due to its very large but diffusively defined blotch and the markings are, understandably, all over the place.
Each plot title has the values EPS and EPS_LARGE called out. These are the above mentioned distance limits for clustering to happen. Here I leave the EPS value, the one for smaller markings constant over several tests, while I step the one for larger markings, EPS_LARGE, between 50 and 90 in steps of 20.
As one can see the large blotch is “surviving” in all cases (which it wasn’t before we introduced the split-by-size clustering approach), while in the fan case it only survives when the “MS” parameter, the number of minimum markings that a surviving cluster needs to have, is at 5. When requesting 7, it’s just not enough markings to have it survive. But that’s okay, because I’m pretty sure that we will have this survive as a blotch rather than a fan, due to the higher number of markings that voted for that.
We’re now 60% through the third set of CTX images on Planet Four: Terrains. We’ve started to think about where we want to search next. We want to continue to fill in the area searched from -70 N latitude to the Martian South Pole. I’ve been coming up with the CTX image selection since the launch of Planet Four: Terrains. I wrote a code that goes through the list of publicly available CTX images and tries to pull out a well balanced distribution of ice-free CTX observations across specific latitude and longitude bins. I thought I’d share my proposed set of new CTX images to search. I’ve sent this list of images to the rest of the science team, and I’m awaiting their feedback. The new set if accepted by the team, will fill in gaps in our coverage and especially between -70 and -75 N latitude. When we have a final list of CTX image to search after dataset 3, we’ll update you here on the blog.
Color Code for figures below.
Red= first dataset at launched and used in our first paper
Green= second dataset
Magenta = third dataset that expanded out to -70 – currently being reviewed on the site
Gray = 4th proposed set of CTX observations to search
The CTX image outlines are overlaid on an elevation interpolated map. Latitude and longitude lines are in 10 degree intervals for above and below. The colors below represent geologic units, but for this comparison we’re focusing on spatial distribution and coverage. More details can be found here