OK I've been poring over these horizon images along with my retrojected orbital HRSC image and other bit and pieces from this thread + Google Mars for ages! I am now pretty confident that I have everything identified correctly (or pretty close).

Click to view attachment

James

That is a fantastic piece of work James. I did wonder about that part of the skyline beyond Endeavour too. Are we sure it's not part of the Iazu apron?

PS - I wish they'd pan around to the north rim again.

Would projected CTX images be of use? The PDS released images are uncalibrated and unprojected, but I've got the process sorted in ISIS3, it's not too bad actualy, so if someone gives me the ob-ID, I'll get them map projected.

That's a brilliant piece of detective work, James. Really brings the whole scene to life.

I wish we had some names - even just working names - for those distant hills and peaks. Maps are so much better with names. Soon tho, I'm sure.

ngunn,

I don't think the distant horizon can be part of Iazu, it just doesn't stretch that far east. Also, it looks more 'hazy' than the rest of Iazu. Looking at Google Mars it does suggest that we should have an horizon about 10km (ish) beyond Endeavour in that direction.

Doug,

I'd love to have a CTX base map of the area. Not sure how much time I'll have to do any further analysis, but here are some ID's of CTX images that would cover all of Endeavour and Iazu.

P13_006135_1789_XN_01S005W
P15_006847_1770_XN_03S005W
P17_007849_1793_XN_00S005W

Stu,

Rui has named everything around here hasn't he?

James

Anything Rui has not named, will just call it Ultreya A through Z.

I've got those three CTX images downloading (140-250 meg each - and I generate about 4 times that in the CTX processing pipeline) - I'll squirt them out tomorrow...and SEE how hard it is to mosaic them within ISIS...but I wont spend to much time on that. I'll zip up PNGs or something and send you a link.

For the P13 ob - ISIS is happy. For the other two, I don't think the kernels have hit the ISIS Rsync server, so I can't project them reliably. I'm going to fudge it assuming nadir-pointing.

Had you tried searching for their correlative HiRISE's? The pointing would be the same.

It would error the same way- tge kernels are external to the images, and HiRISE data would simply be looking for the same missing data. The kernels are quite laggy getting into ISIS, way behind the imaging data.

There may be others that cover the same area, I just grabbed a few ID's off of Google Mars.

It would be good to have as accurate a map as possible to help with alignment of horizon features - fudging it may just confuse me more than I already am!

Well - anything starting 13 or less is great. I'm not sure they started imaging that way much before it was decided to head there thigh. CRX is 30km wide though - so we might be in luck.

I'm used to search at this site: Mars Global Data.
Start with the "webmap" option, zoom to the area of interest, click on the arrow button (select) and then on the map again.

WARNING! If you've never used Mars Global Data before, be very careful. It's dangerous. VERY dangerous. It's the Mars website equivalent of a packet of Pringles or a box of Maltesers - once you start, you can't stop. Oh yes, you go on there thinking "I can handle it, I'll just have a quick look", then you start clicking... and clicking... You tell yourself "Just one more"... click... "Ok, just ONE more..." click... and before you know it it's dark outside, the rest of the family are in bed, and there's a cold cup of coffee on the desk beside you...

Fantastic work James! I can't believe we have come this far, able to see the hills REALLY unfolding before our eyes...
About the names...we all know they've been chosen for quite some time don't we? Ah!

I wasn't discounting your names Rui I meant more 'official' names so we can all follow the journey. Your names are great. You're going to need some more tho, looks like lots of smaller peaks on top of the main hills. Better start researching if any of those sailors had pet dogs or cats...

My personal favorite is P21_009141_1780_XI_02S005W_080708, because it was taken with a very small emission angle (and is thus nearly perfectly Nadir), and so high-frequency topography is least distorted in this image, compared to others.

I'm working on a new CTX/HiRISE mosaic for my route mapping for the project, and put this CTX mosaic together some time ago (so it doesn't include much of Iazu). At least it's not so big I can't paste it here (I *think* >> edited to add, oops! It's an order of magnitude too big!). Scaled up to match the overlying HiRISE images in something like Photoshop, it's about 100GB, though! Fortunately, Arcmap and Global Mapper don't resample to match the scales.

-Tim.

I've been using most of these images on my route map version.
Here's the top left part of it --the so called "context map"-- where I have labelled all background pictures.
Click to view attachment

There are three CTX images (from left to right: P01_001414_1780_XI_02S005W, P17_007849_1793_XN_00S005W, and P13_006135_1789_XN_01S005W) and basically two HiRISE images (PSP_009141_1780_RED and PSP_010341_1775_RED) plus a tiny bit of HiRISE ESP_011765_1780_RED to cover the area we're right now.

Edit: Now I'm wondering that I should change the old "001414" image. It's the only one with an off-nadir pointing so it doesn't match so well with the other ones.

Just making some new slides for my next talk (thanks for the help, James!), thought some of you might find this useful in making sense of Oppy's horizon view...

Click to view attachment

Close up of those particular hills here...

http://twitpic.com/16xi9m/full

Crying out for the full 3D view treatment, I'd say...

Bopolu on the horizon!

Amazing, since it's so far away. (about 65 km to center of the crater).

-Tim.

Where, where! Oh, that's unfair Tim.
You have privileged access to the raw images. We here will have to wait for the next exploratorium update in about two hours.

Tim, that's just cruel.

Here it is!
Click to view attachment
(Click to enlarge)

Ahh! I got my daily fix.

nice but that is one noisy image
quick cleaning

Click to view attachment

I kinda like the noisy image better. Seems the cleaned image has lost a lot of detail.

Amazing! So many things to name so few words...guess I'll have to invent a new lexicon...

FWIW - Calibrated images almost always have less noise than the raw JPG's. That 'detail' you're talking about is most likely noise that shouldn't even be there.

Here's a different version... worst noise removed and a tweak to the brightness.

Phil

Click to view attachment

And a stretched version of the topography. Is there something very faint near the right edge?

Phil

Click to view attachment

there is always a trade off between noise and detail

gmic dose a good job and so dose a fft

and some of the noise is the jpg artifacts
after fft showing the jpg
Click to view attachment

My personal pet peev is the jpg artifacts .the jpg format should be illegal

Where's fredk?
I think we already see that (and some tiny bits in between) in his post 104.

I'm over here!

Here's my take on the latest view of Bopolu. L6 and R1 registered and averaged to reduce jpeg artifacts. Linear contrast stretch and 3x vertical stretch (bottom). No smoothing/noise reduction apart from what happened during rotation.
Click to view attachment
Much better than our views on previous sols! I wonder if the improvement is just a more fortuitous auto stretching of the jpl jpegs, or if the sky is clearing too.

I think it's a case of increasing atmospheric transparency-- remember the crisp views of Endeavour/Iazu just a few days ago at Sol 2170. And it's only going to get better...

--Bill

I've taken the image fredk made and tried to do some feature mapping on a reprojected HRSC image of Bopolu. I think this should be pretty close, the Google Mars 3D matches pretty well with what we see (also included) which gives me some confidence I'm in the right ballpark.

Click to view attachment

James

If it's down to atmospheric opacity, it's almost certain to get worse and better and worse and better..

Nice job, James! That GM view is really convincing.

True, atmospheric transparency will have it's ups and downs, but starting the trek downhill, the view will be getting better...

I'm puzzled by that comment; although we will be getting closer, usually the view is best from the top of a hill.
Steve M

What about the view of the hill?

Phil

The views should improve in the coming months because we're heading into winter, when the transparency's best. I doubt the tiny changes in elevation we'll do will make much difference.

We should also take into account that the route will turn due east and head directly towards Endeavour once we go past the twin craters. After that point, the view should also start to become better in terms of resolution.

Slight downgrade and smaller ripples should also help.

I was delighted to find a nice 4-frame horizon pan among today's images and I've assembled a paper copy. Bopulu is on the right and each of the leftmost two frames contains one prominant dark horizon feature. The one in the second frame looks like the feature Phil identified here:

http://www.unmannedspaceflight.com/index.p...st&id=20656.

From our current position I reckon the twin craters should be about 7 degrees to the left of that, but the two dark horizon features are about 15 degrees apart. In short, I'm not really sure of the identification of either of them, or where (or if) the twin craters actually appear in this view.

Help!!

I get these values from the position on sol 2181. The feature at 181° should be a part of the twin craters.
But it seems to me it's just the left rim of the left crater (from the viewpoint of Oppy on sol 2181) and because of the twin craters cover about 5 degrees in this image your 7 degrees could roughly match nevertheless...

Click to view attachmentClick to view attachment

OK, but that's a little unexpected since from fredk's earlier post below it would seem that the twins' most prominent feature is the mound smack in between them and the rest of the leftmost crater doesn't show at all:

http://www.unmannedspaceflight.com/index.p...st&id=20655

Of course if the feature that Phil identified (tentatively) is actually something rather farther away then a 15 degree angle between it and the twins midpoint could make sense. But Phil's identification was quite convincing . . .

I'm not sure what features are twin craters or if we can even see them yet, but the features I ID'd in the image you posted above can't be the twins. From our current viewpoint, they should've parallax shifted to the left compared to the distant feature much more then they have. So it seems all of these features are quite a bit farther than the twins. I guess it's not too surprizing the twins are hard to see, if they don't have much in the way of raised rims.

Edit - I agree with Tman that the feature at azimuth 181 degrees is about right for Twins.

OK so we have two horizon features. Tim's '181' and the one about 15 degrees to the right of it as seen from here. Probably both are further away than the twins, which may not break the horizon although they should lie close to the 181 azimuth of the leftmost horizon feature. Would that be a fair assessment? Phil's earlier identification of the rightmost of the pair probably depended on your earlier ID of the twins, fredk, so it seems to me that both of the horizon features now still await secure identification on the map. Here are links to the two images in which they appear, a little right of centre in both cases:
http://qt.exploratorium.edu/mars/opportuni...00P2391L6M1.JPG
http://qt.exploratorium.edu/mars/opportuni...00P2391L6M1.JPG

Click to view attachment

No Dougal, those hills are far, far away...

Wow, just noticed today marks the 6th anniversary of me joining the forum... Thanks to everyone who makes it such a great place to hang out and, of course, thanks to Doug for starting it in the first place.

Given the fact that the rover positions are quite accurately known, if the same features turn up in images taken at different sols (different positions), wouldn't it be possible to triangulate the features on the map?
The distance between the L/R pancam/navcam is way too short for accurate measurements over such distances, but if you have images/measurements taken at different positions, you might have a baseline which is long enough to get a good enough fix on the positions of these features.

I don't know whether there is accurate azimuth bearing information for the images (relative to true north, iso relative to rover) or if you can derive it from the relative camera bearing and the rover bearing itself, but otherwise, once you have clearly identified one feature you should be able to get the bearing of every other feature by counting pixels (given that the camera field of view is known).

Yeah, that's more or less what I did when I concluded the original ID's were wrong. We should've seen much more parallax shift between the features. But that was a relative argument. I guess what you're saying is if we have absolute azimuths for each image, we can work out distances for features, if they're not too much farther away than our baseline, and if errors in the absolute azimuths don't dominate.

Yep, if you have absolute azimuths for each feature, you can work out the exact position (lat/lon) of each feature. Accuracy will depend on the length of your baseline (distance between positions/images) and the accuracy of your azimuth. The more different bearings (different images/azimuths) you get, the more accurate your positionplot also.

But even if you do not have absolute azimuths you could probably work out locations if you have two known and one unknown feature, just by measuring the number of pixels (along the horizon) between the features. Given a known baseline (distance between the two positions where the images were taken) and known positions, you can work out the true position of object number 3 (compared to known objects 1 and 2) if you measure the angle (count pixels) between objects 1-2, 2-3, and 1-3 in both images. I would have to work through some of the books again, but it should be possible.