Mars research - help required

[Jumble]

I'm looking for help from IT, computational or image analysis experts who would like to do some genuine, published ground-breaking research (and potentially further the search for life on Mars). I'll try to keep this short(ish) and snappy(er), but pm me for details and published articles.  It goes like this:
- I'm an astrophysicist. I did my doctorate in the field of galaxy alignment, analyzing deep-field galaxy survey images looking for alignment patterns in the orientation of galaxies then modeling the tidal effects that could cause these alignments. (I'll give more details if you are really interested)
- I have a friend whose a geophysicist, and analyzes earthquakes to predict volcanic eruptions, or future, larger earthquakes
One day, in our local corner shop/convenience store (no, really) we were discussing a paper my friend had recently published in which he had been looking at surface images of Mars, and had found signs that would, on Earth, indicate active venting of gases. The results were not conclusive and even my friend accepted that the effect could easily have been produced by venting of carbon dioxide during the Martian summer. He had real trouble getting the paper published.
- We were talking about the possibility of active volcanism on Mars, and the signature it would leave, such as seismic activity. My friend explained that seismic activity was impossible to demonstrate as no one had landed a working seismometer on Mars (really! The one on the Vicking landers was next to the anometer and therefore simply recorded the dodgy anometer bearing.)
- I asked if there were particular geological features you could look for that would indicate seismic activity, such as certain valley shapes, rock falls etc. but my friend said there were non that could not have been produced by erosion, wind, etc. So I asked if the shaking action sifted the bolder and debris by size (like a prospecto panning for gold). My friend said he didn't thi so but he'd look into it. 
- A week later I had a very excited call from my friend (not normally an excitable man) to say that this 'sifting' effect was actually backed by research and he'd looked at some plates from the Mars Global Surveyor and has suspicions he could see this exact effect in a valley that he had suspected of being from seismic origin. 

So, we set to measuring boulder sizes and wrote up our results. We've now been peer reviewed and published in the Journal of Geophysical Research  (planets) (possibly not a mass-readership publication). We've now been picked up by Nature in the Research Highlights section (which is certainly more mainstream). I'm sure I don't need to spell this out, but the impact is basically:
- If there's active, current seismic activity (which we believe we have demonstrated) then it is probably the result of a volcanic heat source, I,e, active volcanism.
- Given the volume of frozen water we know exists on mars, if there is active volcanism then there should be liquid water at some depth.
- If there is liquid water then my exobiology colleagues believe there are more likely to be the conditions for life.
So, where do I need help? Well, I didn't do as much of the computational science as I should have done because a) I'm kind if busy and I have a serious day job which is not science-related and b) I'm a lazy git. I'd like to have digitized the boulder images so that we could computationally plotted boulder size against location. This would have had 2 advantages:
1) it would have been analytical and uninfluenced by human bias (i.e. us picking the boulders that looked right for the effect we wanted).
2) we could cover much more ground. We only looked at 20 plates, but there are 1000s from the Mars Global Surveyor. There is the possibility of producing a seismic activity map for a significant percentage of the surface of mars.  

So, I'm asking for help. I need people who can computationally analyze images and identify and size objects. I used gimp and source extractor, which worked pretty well but needed more work. I need someone to make sure we are really measuring boulders (not shadows, wind streaks in the sand, passing martians, etc. (actually if you see the last let me know!)). Then I need people who can produce a heat map based on the 2-point correlation of boulder sizes in a given range.

Frankly, from what I know of people on this forum this sounds right up your street. Pm me if you are interested and I'll provide a PDF of our article, examples of my attempts to analyze boulder sizes computationally and sample Mars Global Surveyor plates. Our intention is to publish further papers and I promise your efforts will be credited.  Are you in? Please PM and I'll be in touch.

[Tass]

I would love too, but I have less than two months to finish my phd-thesis, and I don't really have the expertise.

But I wish you the best of luck, and will just remark how cool I think it is that it is possible to find actual scientific collaboration on this forum.

[Jumble]

Thanks. I remember that stage and you don't need any other distractions at the moment. Good luck with the thesis.

However, I should have said something about timescales. My thoughts were to aim to get a result in around 6 months, with a couple of months getting the team together, sharing research & data etc. and agreeing scope and approach. Then we will need a couple of months to write code to process plates, identify images and calculate the correlation. Finally a couple of months for errors, refining and presenting so that I can talk my research partner (the proper, full time scientist with all the academic links) to write another paper. So, if you want to keep momentum going when you've finished your thesis let me know (although I wouldn't blame you if you want to take some time off - I did).

I'm guessing I need around 4 or more working together for this to work. So far I've got 2 and counting.

[SlyReaper]

I have some background in image processing, albeit in an engineering environment rather than a scientific one. I don't know if that would be any help at all; there are almost certainly people here with more experience in the area, and who are familiar with tools other than Matlab.

[Jumble]

Matlab sounds perfect. My supervisor kept telling me I should learn matlab, but as he announced that two-thirds of the way through my research project (and he'd talked me into learning IDL in the first place (which is rubbish)) I ignored him and stuck with what I knew. It would be good to have you on board.

[Azrael001]

I've got an undergrad in Physical Geography, and took every available GIS and remote sensing course. I have access to ArcGIS at home and experience with ILWIS. I would like to be involved with this.

[jmorgan3]

I have nothing to add to your project, but I think it's really interesting. I'd really appreciate it if you could let us know when you've published.

[Jumble]

Thanks to everyone for the good wishes, and especially to those that have volunteered. I'm on leave for a week from tonight, so I hope to put a bit of time into this at the weekend. At the very least I'll set up an email box that we can use and a web link with a drop box and the files etc (that I need to pull off my server - note to self).

Given the volunteers to date, we definetely have a project team (although more would be welcome). My research partner is considering if there is anything else we can do beyond the approach I outlined above, but if we can start by analysing the currently available plates that will be a significant step forward. We also have the possibility that, should we find anything particularly anomalous, we can retask the Mars Global Surveyor to take a closer look. Not kidding - we did this for the first paper and I have to admit I found it remarkably cool to be telling NASA how to drive a Martian satellite. However, I'm sure you guys are far more chilled about this stuff than I.

[SlyReaper]

I'm pretty sure the Mars Global Surveyor broke a few years ago, and they've not been able to contact it since. Do you mean one of the other satellites?

[Meem1029]

Wow, that is awesome research! It sounds like something awesome to be involved in, but sadly I cannot be of much assistance as I don't have any experience in those areas (yet). If this were a few years in the future after my graduation, I would be totally for it. As it is, I wish you the best of luck in your research! And be sure to let us know when you publish.

[Jumble]

I'm pretty sure the Mars Global Surveyor broke a few years ago, and they've not been able to contact it since. Do you mean one of the other satellites?

Err, hang on. I thought that was the one but I was on a train to London when I wrote all that. I've got to go and run a birthday party for my 8 year old daughter and her friends, but in the unlikely event that I survive I'll check and get back to you...

Ok. Checked. You're right. It was the HiRISE (High Resolution Imaging Science Experiment) and we pointed it at the Cerburus Fossae.

[eran_rathan]

I'm a licensed land surveyor/photogrammetrist with a moderate background in geology, if you are still looking for help with this.

Also, has your friend looked into whether or not this sorting would be equivalent to eskers caused by glaciation and not seismic activity? I ask because a good portion of my geology background is from glaciology, and eskers are a common feature around here.

[Azrael001]

You and your Eskers. We've got the worlds best Drumlin field here.

[Jumble]

It would be good to look at this to account for potential errors in our interpretation. Admittedly, given the location (bottom of a rift valley) and the potentially very recent dating of the rockfalls (dust deposition rates would indicate 100s rather than millions of years) I'm not sure this will work. However, what do I know: I'm an astrophysicist!

It would be great to have you involved and I reckon we can handle another 5 volunteers before my team management skills start to fall apart. Can I suggest you follow the web site and email address I set up this weekend and I'll send you a copy of the paper.

[Jumble]

OK, I'm making some progress.

Some of you have already given me an email address. For those that hAven't but want to join in (or want to read the paper) can you email me at mars@cosmos.uk.com and I'll send you the PDFs of the research paper.

Meanwhile I've uploaded some relevant stuff to a free file store and have pointed http://www.cosmos.uk.com at it. Let me know if this works as a way of sharing files. If it's too clunky I'll think of something else.

I've written a quick description of the approach I used to endeavour to automate this last time (but didn't complete) as it may be helpful starting point, or at least illustrate what I'm trying to do. Its in the drop box called 'first try'.

Thanks again, everyone.

[Jumble]

Another thought, for anyone interested in this project, I don't think this
Nasa budget slashes Martian fundsis good news.

[SlyReaper]

Not great but at least their budget in other areas goes up. Reading between the lines, it sounds like they've been told to concentrate on the SLS. A massive rocket that still isn't as powerful as a Saturn V.

[Jumble]

I guess we are back to the old debate about space research I. E. is it there to push back the frontiers of knowledge (& does that really require the cost and risk of human-occupied boots on the ground) or is it state-level willy-waving competition (my rockets bigger than yours, etc.). I guess space science works best when it satisfies both camps. My suspicion is that we are presently swinging (!) towards the latter objective.

Anyway, sorry about the delay in comms- I was on holiday. Back now and need to get a plan out to all of you who have volunteered to help. Hopefully later this week.  Meanwhile  have a read of  this! That's us.

[edit] and this. Oh, and someone called from Time magazine today who wants right a piece on this. No, really.

[Ptolom]

I saw the New Scientist article and thought it sounded similar to what I'd read here. Congratulations on drawing such interest.

[Azrael001]

If you know where there is a DEM (Digital Elevation Map) of the selected section that would be super. With that I can get slope, and hopefully crunch some useful stuff. Assuming that thisis the proper site, I'm downloading what I hope to be additional bands of colour, which may or may not be useful.

[Jumble]

I think you need to look at the image in PSP_008779_1905_RED.JP2 (HiRISE data in the drop box) but I'll confirm that over the weekend. Good question about DEM: don't know. Will find out and get back to you.

Anyone tried googling marsquakes recently? It's going a bit mad!

[Azrael001]

I missed page two while downloading everything. This explains my difficulty finding the "First try" and other stuff.

I've been looking at PSP_008779_1905_RED.JP2 (which is georeferenced nicely by the way) but it seems to only be the red band. The three band image that I downloaded is the same area, but is a narrower strip.

[Jumble]

Thanks for letting me know. As I mentioned, I struggled to work with a jp2 of that size, so you're already ahead of me. I'll find out about the other colour bands. 

By the way, have a look at  this. It's gone nuts.

[Azrael001]

So it would seem.

My first try at classifying things was a horrible failure as I didn't train the classifier nearly enough. More bands makes things easier, but I understand that we don't have nearly the satellite coverage of Mars as we do on Earth.

The canyons here seem to be more or less parallel to the prevailing winds. This might get you a lot of false positives...

[Sockmonkey]

How do you tell the difference between a crack formed by gological activity and one formed by a meteor impact that's been eroded for centuries anyhow?

[Azrael001]

Well it depends on how much data you can gather. Meteor impacts will leave evidence that lasts longer than just the crater. I forget the terminology, as it's been a few years since I've done any reading about them, but they can hit hard enough to transform the rock, creating something akin to metamorphic rock, impacted quartz I think. Additionally, the pattern of fractures is usually quite identifiable.

[Jumble]

@Azrael: I suspect you are right.  I seem to remember that you find shock quartz, etc. in the rocks around an impact crater, and I would have thought the fracture pattern would be radial to the impact.  Still, I'm not an expert.  I remember discussing with Gerald Roberts (lead author) the possibility that the seismic activity had been produced by aftershocks meteor impact, rather than geological activity internal to mars. Gerald checked and concluded that the impact record was not consistent with this explanation.  Furthermore, we believe that this is not one event, associated with an impact, but a series of separate quakes over an extended period.  You can see this in some of the boulder tracks that cross over earlie boulder tracks now partially obscured by dust.

It's been an interesting couple of days but it seems to be calming down again. I've not be (partially) responsible for something going viral on the net that way before. It's interesting to watch.  It seems to have all kicked off when the Christian Science Monitor (!) picked up on the original JGR press release that came out through the AGU.  They contacted Gerald and wrote quite a balanced piece. Then it went nuts.  At some point after NASA, Nature and Time magazine, Fox news ran it.  Then people starting extrapolating. I think two of my favorites are the write up in High Pants(odd name, and what's with the guy with the M16), where we are starting an initiative to terraform mars, and in earth changes  where we have apparently found an indicator that the progenitor of the apocalypse, Planet X, is heading our way and we are in the biblical end times. As an agnostic Jew it brings a tear to my eye.  

I'm surprised (and to be honest a bit disappointed) that the bbc hasn't picked up on it. I'd have been tempted to push it their way myself, but it appears the only way for a pleb like me to reach one of their science correspondents is via twitter, and I d'nae tweet, ye ken?

I'm going to write out an action plan today and discuss it with Gerald. The I'll put it in the drop box and we can work out who wants to try what. I'll get back to you all as soon as I can....

[SlyReaper]

Sounds like this is getting to be quite the Big Deal.