Yeah, that's the one I was referring to. It's within the dark plain, but I suspect it may be different in kind from what we might call the 'dudes proper' - the bright arrowhead shapes and dark elliptical lobes around and to the east of the landing site itself.
From your correlation I agree the latter seem to be totally invisible to RADAR (does that make them stealthy as well as spooky?) whereas the 'enclosure' thingy can, I think, just be seen in the SAR images, suggesting that it's ridges are either higher or made of different material - or both.

I poked and prodded to get the T41 RADAR Swath to line up perfectly over the T8 Swath (stretching, warping, slewing, etc.).

And it came out really, really good.

With the overlay, I was able to combine both images to get some noise to average out. AFAIK, this makes this the highest resolution RADAR image yet of the Huygens Channel:

Click to view attachment

But it gets much, much better.....

Look closely at the bigger island to the SW. Right at the top of the island, at dead center is a cute little crater. The higher resolution portion comes from the T41 Swath. (look angle I'll wager is from the SW) This could be the "bright standing island in the Equatorial Dune Sea with a volcanic caldera on top". (There is another suspect cryovolcanic island to the ESE, better seen in the T8 Swath.)

This image is at the full 256 pixel / degree of both swaths, so on a good day it would be 200 m per pixel resolution.

-Mike

Zoom of the Third Island:

Click to view attachment

-Mike

Those are brilliant Mike. All that's needed now is a really accurate overlay of the Huygens mosaics. One problem with ESA's final colour versions of these is that they are (correct me if I'm wrong) stereographic projections. There were earlier greyscale versions which were (I think) horizontal plane projections. I saved these, but without unfortunately keeping note of where I got them. I'll attach them to this post in a few minutes, once I've dug them out.

BTW those blinking gifs - is there a way to slow them down, freeze them, toggle back and forth etc.? (Excuse the elementary question.)

Nice work Mike!

Is it my eyes seeing a fortuitous arrangement of pixels, or is there another, more funnel-shaped vent on the left flank of the Third Island? The crater at the crest looks like it may have a flat, filled-in floor, while the smaller flank vent may be, or have been more recently, active.

I think I've got it pretty close now, I'll be playing a bit more this weekend.

The closest level DISR Mosaic I've coordinated is PIA06438 (I think this is similar to your mid level one)
The upper level Huygens DISR Mosaic I've coordinated is PIA06437 (I think this is similar to your high level one)

When putting PIA06348 on top of PIA06437 there is a slight jump between the two. The Spooky Dude flares out, and the Pockety channel islands to the SW shift farther away. I'm not sure whether to try and correct this or leave it alone. There is a similar shift going up to the RADAR image, so I think it is a projection thing.


For the toggling the blink animations, I'm not sure how to do it. IIRC, it is possible to import an animated GIF into Photoshop and render frames to layers. (You need to type in the .GIF file extension, it is not one of the default file import types).

Is there a particular sequence you'd like to see? [The first GIF sequence was pretty busy].

-Mike

I suspect the problems arise from the fact that you have two DISR image products which are stereographic projections with centers at different heights. They won't match each other and they won't match the RADAR everywhere either. I think it would be very difficult to compensate for these distortions just with ad hoc stretching and bending. I think they need reprojecting to a plane first. Then the two DISR mosaics should match exactly. After that any stretching and bending to match the RADAR will be more manageable. I know I'm no practical help here, though other members might be (e.g. Phil Stooke). I just have an obsessive desire to see this done as exactly as possible. Only then will we be able to state with certainty which of the Huygens features can be correlated, however weakly, with RADAR data.

I agree. Although I'm kinda scared to mess with an image from the Planetary Photojournal, I think I might be able to do a few gentle spherical corrections to manually line everything up. I'll be treating the T8 RADAR Swath as "Gospel" and map everything to that. (I already warped the T41 RADAR Swath to match the T8 RADAR Swath.) And the T8 RADAR Swath was already sized and tilted to match the ISS Oct. 2007 Titan Map (PIA08399) (The T8 Swath did not need distortion to match the ISS image.)

Ultimately, everything should match up to the plane projected Oct 2007 map. Since this is a pretty equatorial location, there should be minimal distortion.


For manually toggling of images back and forth, I think I've got a solution. I'll post a series of individual coordinated JPEGS all at the same scaling. Then it will be possible to download these to the same folder, and move between them in a viewer like Irfanview. Alternatively, they could all be cut/copy/pasted into individual slides in Powerpoint, and you can rearrange and toggle between them as you wish.

-Mike

Here is a graphic with an approximate scale (I used Nigel's posted high DISRmosaic for the scale).

Click to view attachment

With this scale, the Third Island summit caldera is about 100 km to the ESE of the Huygens landing site.

The summit caldera is less than 5 km wide. IIRC, Titan's current thick atmosphere would shield the surface from impact craters less than 10 km.

-Mike

Mike, that combined radar image of the Huygens landing site looks great!

Let me provide some background information on the DISR mosaics.


These are from the Tomasko et al. 2005 Nature paper.


If I remember correctly, these images (and the ones from the paper) are gnomonic projections. The two mosaics were made by manually combining DISR images by matching features in regions of overlap, allowing the spacecraft roll, pitch, and yaw to vary (the altitude was more or less known). Images taken at the same time were fixed with respect to each other. Due to the sheer complexity of the process of building a 360 degrees mosaic from tiny images it was inevitable that systematic errors crept in, which may explain most of the mismatches (the two mosaics were probably made independent of each other, using different sets of images).

The color mosaics that were released later were created by Erich Karkoschka, and are based on his superb Huygens trajectory and attitude reconstructions. They are probably the most accurate to date with respect to feature placement. But keep in mind that while these mosaics come close to perfection (IMHO), it is inevitable that they contain artefacts too.

Stefan.

This is a simple and naive mosaic of the three Photojournal mosaics, blended a bit. Simply for your viewing pleasure, no science value warranted nor included.

Click to view attachment

Wow great stuff everyone.

What a tantalizing mosaic, JK, thanks! That small peek by the Huygens DISR in the context of the lower-res imagery really illustrates how much more there is of Titan to know. God, I hope I live long enough to see global imagery of Mariner 9 quality of better.

Going back a few posts to the question about stereographic projections... the 'filter-distort-spherize' filter in Photoshop can fix the problem, but a bit of messing around would be needed to get it right.

Phil

Here's an animated view of the Huygen's landing site using T41 SAR and the stereographic color view of the DISR at 20 km from Titan's surface. The dunes and bright edge of the northern-most island line up fairly well. The V-shaped ridges (pointing right) may be there but are displaced a bit to the left of the actual location; may have to tweak it some.

Click to view attachment

Interesting, but from the curvature of the cat-scratch dunes it looks as if you have left the DISR mosaic insufficiently corrected for the effects of stereographic projection. This would leave the centre of the DISR magnified and the perimeter compressed relative to the 'flat' SAR, by quite a significant factor I think. It is intriguing though that there are some pointy shapes in that SAR image that just might be the 'spooky dudes'. That's what I thought I saw on peter59's version of the T41 SAR way back in post 44 of the T41 thread:
http://www.unmannedspaceflight.com/index.p...st&id=16882

Here is yet another complicated GIF animation showing my final lineup of the Huygens Landing site.

I treated the T8 RADAR Swath as "fixed" (thus the ISS as well) as well as the upper level DISR/RADAR mosaic PIA06437 as "fixed".
The T41 RADAR swath was warped.
The mid-altitude DISR mosaic (PIA06438) was manually warped (inverse spheroid filter didn't work so well) to fit over PIA06437.

Here is my FINAL result:
Click to view attachment
(click to animate)

animation sequence:
1) pseudocolorized ISS + T8+T41+upper DISR mosaic + middle DISR mosaic
2) T8+T41+upper DISR mosaic + middle DISR mosaic
3) T8 + T41 + middle DISR mosaic
4) T8 + T41
5) T8 + T41 + upper DISR mosaic
6) T8 + T41 + upper DISR mosaic + middle DISR mosaic

I'm pretty happy with how some of the upper DISR mosaic pieces line up with RADAR blips underneath in the WSW end of Hugyens Channel.

I'll post the individual frames next...

-Mike

Here are the individual frames for line up. (Download - Cut/Copy/Paste into Powerpoint and toggle away!)

Click to view attachmentClick to view attachment
Click to view attachmentClick to view attachment
Click to view attachmentClick to view attachment

Have Fun!

-Mike

Here is a three-way blink comparison of a zoom closer to the landing site:

Click to view attachment
(click to animate)

mid-level DISR mosaic + T8 +T41
T8 + T41
mid-level DISR mosaic + upper level DISR mosaic +T8+T41

(Note how the long stubby channel channel in the DISR might be visible in the T8+T41 combo image. The dendritic drainage area corresponds to a slightly darker RADAR zone on Huygens Island.)

Here's a graphic showing how the stubby dark channel is visible in both DISR and in the combined T8+T41 SAR RADAR images:
Click to view attachment

-Mike

Zoom with only the T8 + T41 RADAR with the mid-level modified (manually warped) DISR mosaic of PIA06438:

Click to view attachment
(click to animate)

This is at maximal RADAR resolution (256 pixels/deg) or 170 m/pixel.

For the next images, I'll use the DISR mosaics at full resolution (they were shrunk onto the RADAR basemap).

But not tonight.....

-Mike

This is looking pretty good now. It seems to fit almost perfectly at the left hand edge (in the region I was calling the enclosure). But when I look at that animation I keep wanting to rotate the DISR just slightly anticlockwise about that point, raising the spooky dudes a little. I'm sure you're going to tell me that can't be right though . .

I'm really happy that this is getting a thorough examination now and very grateful for all the image work so far.

Here is a blink based on a lineup presented in Soderblom et al. Planet and Space Sci., 2007. The DISR-RADAR lineup presented in the figure of the text was somewhat squished to correct for the DISR projection, but the island outline should line up with it's RADAR trace.

The Huygens Island lineup was painstakingly reconstructed based on that publication. (Actually shifted a tad, the red plus signs used to show Huygen's landing point are a bit off in the publication.) The blink is between PIA06437 and the T8+T41 combo image.

Click to view attachment
(click to animate)

Nigel will be happy, the island is rotated counterclockwise a tad.

In this image, the distance between the inland corner of "Long Island" and the pointy tip of the Spooky Dude formation is 47 pixels = 8 km. This matches up exactly with what is expected according to the bar scale presented in PIA06437 at the T8 RADAR scale (4 km = 24 pixels: 170 m/pixel).

[The previous line up above had a distance of 8.6 km, still pretty close.]

-Mike

And here is another blink based on a faint outline observed in the T8 RADAR trace. Assuming this is the full island, and that the bright RADAR part is only a section of Huygens island, the scale changes.

Click to view attachment
(click to animate)

Doing this overlay, we get a much bigger distance for the Long Island - Spooky Dude tip distance, = 81 pixels, which is 14 km.
This is twice what we'd expect based on the PIA06438 image scale (if that is correct).

Arguments for this image:
Nice lineup of outer faint RADAR trace of Huygens Island's arrowhead shape
Line up of terrain feature on Huygens island (the stubby channel marsh terrain)

Problems with this image:
Distance scale 2x what we'd expect based on lower DISR mosaic PIA06438
Poor line up with nearest dark dunes (could be lineup with further set, but where are the near ones?)
Poor lineup with South Island RADAR bright area
Hard to explain why a marshy area would be more RADAR bright

Of these three lineups of PIA06437 and the T8+T41 combo, I'm voting for the Soderblom-based lineup.
That also puts the RADAR-bright area of Huygens Island in the middle of the mountain tops. (There also might be a subtle lineup of the Spooky dude with a line of sl. brighter pixels....

Any advice?

-Mike

OK I played with the Soderblom-based image a bit more. a slight twist of about 5 degrees clockwise and a little shift to the SE, and a tiny bit of shrinkage gets the little bright splot in the SE to line up with the Soderblom figure and the T8+T41 RADAR. I'm very happy with this.

It puts PIA06437 at a 351 degree orientation (top is to the NNW).

Here is the final blink:

Click to view attachment
(click to animate)

Note in this version, I removed the RADAR portion of PIA06347 and left the DISR mosaic. Note the overlap with several features (dunes/W end of the island, South Island, etc.)

-Mike

Actualy a tad too much for what I had in mind! I'm hopeless with images but I will try to attach a couple with annotations to show what I'm proposing. The yellow circle indicates the approximate centre of rotation in your zoomed gif (post 520). The green line indicates the point (literally) I was intending to match up, and the red line indicates another feature I think may also line up slightly better as a result of that same small rotation.

Do mission scientists know what the "Spooky Dude" formation actually is? Is it a series of linear "islands", a set of clouds or something else?

They are definitely geologic structures.

Photoclinometry gives a maximum height of 200 m (+/- 50 m) above the "bright/dark boundary line". This was published in Soderblom et al Planetary and Space Science, 2007.

They were referred to as "bright islands" in both Soderblom et al. and Keller et al..

It is not clear if the spectroscopic measurements in Keller et al. cleanly hit the islands or not. So spectroscopic determination might not be possible. (Not sure if VIMS has sufficient resolution to see this pocketed structure).

Both authors described them as having formed from large laminar currents in the channel. (Soderblom et al. had the currents running from W to E.)

Examination of one of the "pockets" of the structure reveals a drainage channel breaching the edge. So these pockets must have overtopped at some point.

My personal pet theory is that these were formed during impact-generated tsunamis during a much earlier period when the basins were filled with a nitrogen sea. The pockets being formed by solid methane icebergs which later melted out.
(I'm finding a hard time imagining other process that could build large tapered structure with large cavities and drainage breaches from the inside pockets.)

Both authors do propose laminar flow in the channel as the major process that explains the shapes of things in the channel. (So channel flow rearranged anything the streams dumped into the channel)

-Mike

What kind of flow would the migration of Titan's tidal bulge create in these seas? Would it "slosh" in both directions?
Ahh.. good description here. Asked and answered. Guess it doesn't migrate per se.

And I'm finding a hard time NOT imagining your theory with a triumphant musical score. Nitrogen Tsunamis - Methanobergs - Giant Impacts - that scenario sounds perfect for dramatic computer graphics on a NOVA special.... Titanically Titan!

I thought 'flow features' as soon as I saw the first Huygens mosaics, but the other word that came to mind was 'fresh'. I have difficulty believing they are geologically ancient. I think the tsunami idea is helpful in understanding the manner of their emplacement but at the moment I lean towards the idea that occasional catastrophic rain-generated floods, possibly driven horizontally across the ground by gale force winds, could be responsible for periodically reworking this material in the way that we see. I also suspect that the most mobile debris is also the least dense, maybe actually buoyant on liquid methane, and that this is what will be found uppermost on the ridge deposits. This would make greater ridge heights easier to attain for a given severity of flood. I'm not sure how much of the formation we have elevations for. It is also possible that a few of the highest ridges (though not I think the ones immediately surrounding the landing site) may be underlying topographic features that are merely camouflaged by the surrounding flow materials. (Oops -have to go now - more in another post).

To conclude . .

I also note that the formation I have been calling 'the enclosure' to the WSW of the landing site (for which we have a terrain model - posted earlier) has an interior which is darker than the surrounding plains. This suggests to me that it is relatively protected from the to and fro of debris-laden floods.

Silly question, what is the "Spooky Dude"? Are you talking about the streamlined forms south of the main island?

Yup. The feature as a whole has been referred to as "bright islands" in the literature.

They could be an eroded chain of ice-volcanos --see picture (well, surely it has been suggested a gadzillion times before).

That site shows some mechanical ice volcanos --water is pushed up by wave pressure, not by heat (vapor). When they wake up they can throw away some big boulders.

I think this very early amateur mosaic illustrates well how the nickname arose (Mike can confirm or correct this):

A very interesting phenomenon I had never heard of before. But don't you need some neighbouring open water for the waves to form in the first place? That would indeed be a very different Titan from the one we see today.

...and you can buy the T-shirt here: http://www.amazon.com/Scream-Mask-Shirt-co...0408&sr=8-8

I'm not a geologist. Maybe Titan's subsurface ocean has strong tidal currents, or the drag of the crust as it moves over the ocean creates turbulences. The currents could hit a tectonical feature and flow through the cracks. Also, the final shape of a chain of ice volcanos could be dictated by prevailing winds and not by erosion.

Thanks for responding to my enquiry. When I saw the "Spooky Dude" formation at first from the Huygens DISR miages when it landed (4 years ago now...hard to believe) I thought they looked like low, lineear clouds. Maybe they are "islands" of lighter icy material formed in a flared pattern by a directional force of liquid and gas, like the "teardrop" shaped formations in the dry channels on Mars are believed to have been formed by a catastrophic flood?

[quote name='ngunn' date='Jan 12 2009, 05:14 AM' post='134029']
Interesting, but from the curvature of the cat-scratch dunes it looks as if you have left the DISR mosaic insufficiently corrected for the effects of stereographic projection. This would leave the centre of the DISR magnified and the perimeter compressed relative to the 'flat' SAR, by quite a significant factor I think.

You are so right! I have tried to 'amend' the DISR stereographic at 20 K by de-spherizing the image so the cat scratches are flatter. I have then combined that image with Juramikes' superb composite image from T8 and T41 matching landmarks (dunes, S island bright points, island edges etc) and overlaid the two images in 10% gradient steps. Looks better now. Can't say the spooky dudes stand out but there's some matching of points in the w end of the channel.
(Note: image layers take a little while to load)

Click to view attachment

Gosh, that's really different from the version of Mike's that I liked best. I like the slow, gentle fade though. It gives you time to think about what's disappearing and whether anything's reappearing in the same spot. I've looked at it for quite a while and to be honest I'm not sure what to think now, except that this is still unfinished business.

Here is a blink animation of my final alignment of the DISR mosaic over an averaged combination of the ISS image and the T8/T41 RADAR combination image (previous blinks used RADAR only):

Click to view attachment
(click to animate)

The lineup is almost perfectly matched with Figure 1 of Soderblom et al, Planetary and Space Sci., 2007 (doi: 10.1016/j.pss.2007.04.014). I used this figure as a guide to align the Huygens Island orientation and scale.

Here is a graphic showing the 14 features I used to coordinate up the two images:
Click to view attachment


-Mike

Here are the two images in the blink animation for cut/copy/pasting into Powerpoint:
Click to view attachmentClick to view attachment

Enjoy!

-Mike

Pseudocolorized [T8/T41+ISS RADAR] mosaic (4 pixel Gaussian blur) of the Huygens Channel with the DISR high and medium altitude mosaics overlaid (PIA06437 and PIA06438):

Click to view attachment

The pseudocoloring of the [T8/T41+ISS] basemap has been adjusted to roughly conform to altitude for Huygens Island and Huygens channel as presented in Figures 4-7 in Soderblom et al. Planetary and Space Sci., 2007 (doi: 10.1016/j.pss.2007.04.015).

Both DISR mosaics have been warped slightly to line up the features.

The view corresponds to 170m/pixel (the medium altitude DISR image was reduced) which is approximately the full RADAR resolution (256 pixels/degree). This view is approximately 137 km x 160 km.

-Mike

For an interesting side-by-side comparison here's a reminder of the earlier official version:
http://www.lpl.arizona.edu/~kholso/landing_site_poster_2.jpg

Pseudocolorized view of Huygens Landing Site
Detail of Huygens Island and Huygens Channel from overlay of T8/T41+ISS image + DISR mosaics

Click to view attachment

A much, much larger full-resolution version (14 m/pixel!!) is available here: http://www.flickr.com/photos/31678681@N07/3206221056/
(This is at the level of the larger released version of PIA06438)

-Mike

As a check for any residual distortion of the stereographic image that was de-spherized in post # 540, the following animation now incorporates the DISR image PIA06438 of the Huygens landing site (HLS). There is very little change in the central portion of the image with this and so some confidence in assigning DISR objects to the T8 + T41 composite SAR of Juramike overlay image. The animation ends with higher resolution DISR images of the landing site rotated for a better view.

Click to view attachment

I don't like to sound negative about such impressive and painstaking image work, but here's one small example of why I'm still not sure I'm looking at the definitive fit between DISR and SAR. I measured the external angle of the very prominent V, arrowhead, spooky dude thing on Mike's latest version and the TR one above. I found about 5 degrees difference between the two measurements. Indeed the V seems to stretch and become narrower as the TR animation progresses. I know that lots of ad hoc adjustments have been made to arrive at these results, but are they the right ones? For example has rotation been applied when a shear distortion would have been more appropriate in some cases? Is the conversion between stereographic and plane projections mathematically rigorous or is that too achieved by ad hoc means and therefore dependant on assumptions about how the images should fit?

All in all I'm still not absolutely sure which, if any, version has the Huygens channel DISR features mapped exactly right relative to SAR. Therefore I still don't know if 'spooky dudes' are marginally visible or totally invisible to RADAR. Therefore I still don't know if they could be all over the place on Titan, or if the Huygens channel is a highly atypical location. However I do still have hopes that this can be teased out of the data eventually, perhaps with the aid of VIMS. (Anybody heard anything more about the VIMS?)

Another unanswered question: why is the T41 SAR at this location so much noisier than the T8?

I've been using the "Karkoschka Mosaic" (PIA08114 - a very big file) to try to line up angles for the lower lelel DISRs.

(It is a fisheye projection, I've tried to despherize it and warp it manually to make it "flat" but haven't been able to do it without destroying all the useful information. For the lowest-level DISR, I'll be using PIA06439.)

Despite the projection issues, the vectors from the HSL touchdown point should be the most accurate.

Here are the vectors I get to some of the key features (Karkoschka mosaic N = 0 degrees):
HSL-->mouth of stubby creek drainage = 280 degrees
HSL-->mouth of dendritic creek drainage = 10 degrees
HSL-->pointy peak of Spooky Dude formation = 76 degrees
HSL-->W most point of dark inner basin of oval hollow located to SE (the feature that had it's topography measured) = 120 degrees
HSL-->"brightest 'x' of S-most bright cup shaped feature to SW of HSL = 212 degrees
HSL-->mouth of Huygens Island W-bounding channel into Huygens Channel = 251 degrees

-Mike

Pseudocolorized zoom of Bright Islands in Huygens Channel
Detail of Huygens Channel, Huygens Island drainage patterns and the Bright Islands from overlay of T8/T41+ISS image + DISR mosaics

Click to view attachment

A much, much larger full-resolution version (again, 14 m/pixel!!) is available here: http://www.flickr.com/photos/31678681@N07/...22734/?edited=1

-Mike