I made a list of 42 putative impact craters on Titan to try to estimate erosion rates of craterform features.

This list has the “typical” craterform features all characterized by an ISS or RADAR circular bright rim with either a RADAR or ISS dark central portion or a RADAR or ISS bright inner portion (dome or peak) and a darker inner circle. Several of these have been previously reported in the literature. Due to the extensive rim erosion, some of these are broken or incomplete circles.

These (broken)circular features have been classified into five groups depending on the level of erosion evident (from most pristine in appearance to the most used and abused):

- Fresh craters - with little to no erosion evident on the crater rim or debris apron.

- Recent craters – with some evidence of fluvial erosion on the rim, some crenellation being present.

- Eroded craters – with a single breach of the rim or with severe erosion of the rim wall.

- Multiple breached craters – with several complete breaks in the rim structure.

- Degraded craters – with collapse or removal of large sections of the rim structure with extensive invasion of new materials (e.g. dune sands).

Here is a map showing the approximate locations of these features on Titan:

Click to view attachment

-Mike

Here is a gallery of Fresh Craterform features from the identified list on Titan.

[In each of the sets of images that follow, the least eroded set is first in the series, the most eroded set is last]

Sinlap is a typical example of a fresh crater, it has a nice bright debris apron, a clear complete circular rim with little crenellation. Alluvial fans or erosion in the rim is not evident. There is a slight hint of streams beginning to form in the debris apron.

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

-Mike

Here is a gallery of images of Recent Craterform features from the identified list on Titan.

Ksa is the least eroded example of a recent crater. There is some initial incision of the rim present. Some alluvial structure with bright streambeds (rapid erosion) is visible beyond the rim either on the outside or on the inside portion of the rim. More typical examples include the T23 crater and Menrva. In addition, there can be some burying of the structure evident in higher latitudes (more evidence they've been around a while).

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

-Mike

Here is a gallery of images of Eroded Craterform features from the identified list on Titan.

The Shikoku crater is a good example. The crater rim is now showing definite signs of degradation. There is at least one crack or breach of the rim. Definite RADAR dark streambeds cut into the apron.

Click to view attachment
Click to view attachment


-Mike

Here is a gallery of images of Multiple-Breached craterform structures from the identified list on Titan.

Guabonito is a typical example. It shows multiple cracks or narrow breaches around the crater rim. Severe crenellation is evident in the remaining sections of the rim. In some features, streambeds or dune sands have breached the wall and penetrated the interior section of the craterform.

The pattern of erosion around appears to be uniform. Crack formation does not appear dependant on local tectonics or wind direction as evidenced by the random vectors of the cracks around the rim. This supports that crenellation and erosion due to rainfall-created streambeds drives the initial crack formation around the crater rim.

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


-Mike

Here is a gallery of images of Degraded craterform structures from the identified list on Titan.

These are the most difficult to identify. Several of these are only partially complete rims and are thus more open to speculation. Some identifications are based on a combination of ISS and RADAR images. ISS imaging is assumed to indicate elevation. In extreme cases, (e.g. entry 42), the ghost image of the rim is evident from a change in the dune sand pattern.

A good example is the crater to the West of the Huygens landing site (entry 29). In this example, the cracks have widened and some of the remaining crater sections have collapsed and been eroded away. Only a few rim section remain surrounded by an ice mantle or eroded deposits presumably from the rim. Dune sands have invaded and pass through the central section.

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


-Mike

Here is an EXCEL file showing the entry number, latitude/longitude coordinates, and diameter of the crater, it’s identified level of erosion, the source where the image was plucked from, and the literature reference, if applicable of the list of 42 craterform features I identified on Titan.

Click to view attachment
Here is another map showing the numbered locations of the craters above on a combined ISS/RADAR map of Titan. (numbers on map correspond to entry number in EXCEL file and image numbers above):


Click to view attachment

-Mike

Wow...I did not realise there were so many craters on Titan.

How does Juramike manages to do all this extraordinary work?! Amazing!

It is actually remarkably few, compared to most other bodies with solid surfaces in the solar system. It seems that a 50-100 km scale impact crater doesn't last more than a few hundred million years on Titan.

With 42 craters sprinkled among different erosion states, it should be possible to calculate the erosion rate of the crater rims. (That's kinda where I'm going with all this: using emprical data to calculate the observed erosion rate].


The overall erosion sequence for the observed craters is:

Fresh --> Recent --> Eroded --> Multiple Breached --> Degraded


Each step should have it's own k value (rate constant).

Since there are not so many eroded craters in the list, it seems likely that k(Eroded --> Multiple breached) is faster than the other rates.

Qualitatively, once one crack happens in a crater rim, other breaches happen rapidly.

I'm not sure of the exact mathematical treatment to use to calculate erosion rate. Unlike radionuclei or molecules bouncing into each other, degradation is not a random event, so a T1/2 logrhythmic decay doesn't seem the right way to go.

[For example, if a bunch of craters of y km rim thickness are all being eroded at x km rim/yr, then they should all start to decay after y/x yr. That is, they should all start to go at once. So it would look more like a step function rather than a smooth logrhythmic decay]

-Mike

For grins and giggles, I made a combo/hybrid of the T23 crater and Menrva. They are both at the same level of erosive degradation.

Here it is:

Click to view attachment

(I think they fit pretty good. What do you think?)

-Mike

Binning the total list in log 2 sizes and making a log 10/log 2 Hartmann Plot, the observed data (blue dots and line) seem pretty consistent with a 4.5 Gyr old surface:

Click to view attachment

Craters > 100 km fit the projected rates for 4.5 Gyr.

Craters < 100 km seem to fit a lower rate.
(Some Possible reasons for a paucity of smaller craters: some atmospheric filtering of smaller impactors, as well as smaller craters being difficult to observe by only ISS. RADAR seem better at identifying smaller craters, and RADAR coverage is still not complete.)

-Mike

[EDIT: These rates were calculated based on the ratio of binned observed craters over the ENTIRE surface of Titan (8E7 km2)]

To try to figure the erosion rate I’ll focus on one question:

On average, how long does it take to breach a crater rim on Titan?

Using a Hartmann Plot, I plotted combined crater rates of craters that had intact rims (“fresh” + “recent” craters as defined in the list) in the same graphic with total crater rates on Titan.

In the graphic, the blue line is total crater rates on Titan, the purple line is the rate for craters with intact rims (fresh and recent) calculated over the entire surface of Titan:

Click to view attachment

Two things can be gleaned from the graphic: - Even “fresh” + “recent” craters are enough imply an older surface (>1 Gyr).

- Craters with intact rims are approximately 50% of the total crater count.

Here is the breakdown by size:

Crater bin size (diameter in km, %intact craters of total):
16 (70%)
32 (33%)
64 (0%)
128 (20%)
256 (50%)
512 (50%)
1024 (0%) [The degraded sliced carrot feature is the only crater in this bin]


Conclusion: Even with a limited sample, assuming the observed total crater count implies an 4.5 Gyr surface, it seems to take about 2 billion years to degrade the rim of a crater on Titan. This rate appears independent of crater size.

-Mike

I don't quite agree with your interpretation. While the relative shortage of small craters could to some extent be due to the thick atmosphere and the difficulties of detection it seems to be quite likely that it is also due to many smaller craters having eroded until they are undetectable. After all SAP radar is pretty good at detecting heavily eroded craters here on Earth. Your # 37 is a beautiful example of a (probable) crater that has eroded almost into invisiblity. On Earth (with a vastly higher erosion rate) I would guess that it is from tens to hundreds of million years old, depending on geological context. Where there are craters that have eroded to that point it seems almost certain that others have disappeared completely.

The other issue is that a number of the crateriforms that Mike has identified, while intriguing, might not be impact features. For example, I don't believe that the "Sliced carrot" feature is caused by an impact feature. Impact structures of that size, if you look at Callisto and Ganymede, are typically multi-ring structures. The sliced carrot feature appears to only be a single, partial arc. I don't think we can rule out a tectonic origin with that feature.

As structures become more degraded, they are much more difficult to positively identify as impact craters.

I have tried to pair up the images so that similar shapes and erosion levels are in each graphic. For example, the "Sliced Carrot" (entry 32) looks similar to a craterform feature in S Senkyo basin (entry 31). [In retrospect, these should have paired up in the same graphic, as they are done below:]

Click to view attachment


For both features, I would agree that tectonics has played a very important part in the evolution of the structure. (Either impact-->tectonic modification-->erosion or tectonic modification-->impact-->erosion)


Looking at all the images listed in this thread, it seems that there is a diverse variety of crater shapes. Final crater shape should be function of impactor energy, crustal structure, and viscosity of target material. A study of impacts into slurries (Bingham materials) shows that even small changes in viscosity (or impact energy) can have a dramatic effect on crater shape, particularly on spacing and appearance of secondary ring structures.

I strongly suspect that the crustal structure and viscosity of target material is not uniform around Titan. The different shapes of lakes (fractal w/rivers vs. pothole/sinkholes) in the two regions of the north polar area demonstrate that at least the near surface geology varies regionally. So it would seem to reason that the shapes of craters resulting from a similar impactor energy may also vary regionally.

-Mike

It is very likely that some craters have been eroded beyond recognition. Here is a graph that shows the distribution of identified craters by erosion state:

Click to view attachment


How many craters have been eroded beyond “degraded”? There is no easy way to tell. But since the observed distribution for larger craterforms fits a 4.5 Gyr age profile, it seems likely that “not that many” craters (at least for larger craters) have been eroded or buried into obscurity.


[I take "not that many" to be within an an order of magnitude of the observed numbers - enough to still fit within the errors bars of a 4.5 Gyr crater distribution range]

Here is some evidence that craters have been obscured on Titan:


Here is a graph showing the crater distribution according to latitude zone:

Click to view attachment
It is very striking that most of the identified craters are in the Equatorial Zone, despite that fact that Polar Zone SAR RADAR coverage is pretty complete.

Assuming a uniform distribution of impactors, we would expect more craters in the polar and temperate region. I’ll leave it to a math goo-roo to calculate the surface areas of Polar Zones, Temperate Zones and Equatorial Zones of Titan. But taking the areas as equal (they’re not), we’d expect 30 more craters in the combined Temperate Zones and 30 more craters in the combined Polar Zones. Thus, about 60 extra craters or so have been “disappeared” (buried or eroded) in the non-Equatorial Zones of Titan.

This provides some observational evidence as well that burial (or erosion) is latitude dependant. Craters are more visible at low latitudes. The rates calculated from this data are weighted heavily by Equatorial Zone craters and are probably slower than those calculated for higher latitudes.

I would expect that as RADAR coverage in the Equatorial Zone increases that the number of craters will also increase. I would also expect that the identification of smaller and more degraded craters would especially increase as RADAR seems to be better at picking out the smaller and more subtle structures.

-Mike

Mike, you also might expect some equatorial enhancement just because of the existence of the rings. We certainly don't know their history with any certainty, but there sure as hell is a nearby very abundant population of potential impactors...

Of course, the flip side of that postulate is whether the inner large icy moons show any evidence of equatorial impact enhancement. Doesn't look like it to me as a casual observer, but I defer to the professional crater-counters...

Very even-handed of you to shoot down your own hypothesis in the same post ;-)

But lack of evidence aside, note that confinement in the ring plane does not imply concentration
of impacts at low latitude anyway - Titan's orbit is inclined at 0.3 deg or so to the ring plane, so
Titan would bob up and down through this alleged disc of impactors. (How you increase the orbital
energy of ring particles (which are all too small to make craters on Titan anyway?) without also
giving them some inclination too is another challenge.)

So the latitude effect is principally an issue of latitude-dependent visibility or latitude-dependent
removal.

By examining the alluvial fans surrounding a crater that is right on the cusp of becoming eroded, it might be possible to obtain an estimation of the erosion rate. This is possible since it is estimated (from above) that a fresh crater will transition to a breached (eroded) crater in 2 Gyr.

Click to view attachment

Taking the Shangri-La crater as a typical "fresh" crater, we can use the measured values from the ISS image for the rim extension. The measured value is 16 km for the slightly darker area presumed to be the rim and the pedestal beyond the rim. The rim height is estimated to be 1.3 km in analogy with the floor depth of Sinlap. It is assumed that the rim height is 1.3 km above the surface outside the crater as well. A 40 km rim section would thus have a volume of approximately 832 km3.

Click to view attachmentClick to view attachment

The T23 crater is an example of a recent crater very close to being breached. The T23 crater rim can be discerned in the image as a semicircular section displaying peak foldover. (This is well beyond the inner dark - bright dark boundary on the left side). Measurement of the width of this rim gives a value of approximately 13.5 km. Again, the rim height is assumed to be the same as for Sinlap, or 1.3 km. A 40 km rim section would thus have a volume of approximately 350 km3.

The second graphic shows the red circular outline of the RADAR blocky crater rim. On the right side of the crater, a nice alluvial fan structure can be seen just beyond the rim, highlighted in light blue. The volume of this alluvial fan can be used to estimate the erosion rate for this crater.

The volume can be calculated based on the 40 km x 40 km approximate dimensions of the fan spread. A deposition slope of 1 degree is used for the fan structure. (This is approximately half the value of 2 degrees described in Radebaugh et al Icarus 2007 Mountains of Titan observed by Cassini RADAR.). This gives an alluvial wedge height estimate 700 m (or half the original rim height). The volume of the entire fan structure is thus 560 km3 or less (a wedge shape was used to make the estimation easier, in reality it would rest against the original rim angle of repose and thus not be a right triangle).

Click to view attachmentClick to view attachment

So an original 40 km rim section of 832 km3 volume is estimated to have eroded down to a thinner rim section of 350 km3 and an alluvial structure of 560 km3. The erodible surface of the original rim section was 40 km x 16 km or 640 km2.

Since the projected age of a crater just ready to transition to a breached crater is 2 Gyr, this gives an erosion rate of 560 km3 eroded from a 640 km2 surface over 2 Gyr.

So we estimate an erosion rate of 4.4 E-10 km3 * km-2 * yr -1

(= 4.4 E-10 cubic kilometers per square kilometer per year.)

-Mike

[EDIT 2/12/2008: The T23 crater estimated age should be closer to 2.8 Gyr. This would give a corrected estimated erosion rate of 560 km3/(640km2*2G.8Gyr)= 3E-10 km yr-1 = 0.3 m Myr-1]

Good job on your study, Mike.

It actually matters not if some features are not impact. Even if they are tectonic in nature, they represent a topographic feature that is undergoing erosive processes. I would suspect that these processes are more akin to those on Venus than anywhere else.

--Bill

Yeah, I'm cool like that!

After thinking about it some more, though, there still might be something there. Consider the possibility that erosion on Titan (at least for large/midsized landforms) might be slower overall then on the icy moons: the latter are exposed to micrometeorites, greater temperature extremes, etc. We don't seem to know much about the physical properties of Titan's crust, much less its composition. The additional assumption is that the rings formed catastrophically & spread debris of a continuum of sizes throughout the Saturn system, focused eventually on the equatorial plane.

I don't think this is likely at all, frankly (although I do think that erosion on Titan moves much more rapidly in the polar regions where the subsurface methanofiers seem to be), but thought I'd throw it out there for input. Titan's so different that not only do you have to think out of the box, you have to have the box shipped to Eris & have a memory wipe so that you do not think of the box ever again...

How often does it rain on Titan in the Equatorial zone?

With an erosion rate estimate above and the streamflow estimates during flood rain events we should be able to estimate rainfall rates in the Equatorial Zone.

(Lotsa math and units conversions, feel free to skip to the end)

By looking at channel sizes observed on Titan, the streamflow during rain events was estimated (ref: Jaumann et al. LPSC 38 (2007) Abstract 2100 "Surface Erosion on Titan") at both Bohai Sinus on Quivira and the Huygens Landing site. Both locations are located in the Equatorial Zone.

From Post 223, Titan's Lakes Revealed thread (link here), a rough estimate of average sediment load for rivers on Titan based on Earth rivers was obtained (0.2 kg/m3 flow). The density of ice sediment is assumed to be 0.8 kg/1000 cm.

To calculate sediment load during rain events in Bohai Sinus (units converted where necessary):

Averaged Stream discharge for Bohai Sinus: 8.6E3 m3/s
Drainage (collection) Area (km2) of Bohai Sinus: 7.9E9 m2 (= 7.9E3 km2 in reference)
Solvent exposure = 1E-6 m3/s of flow (solvent flush) per m2 of surface during basin saturating rain event
Assume 0.2 kg sediment load/m3 flow (= 2.5E-4 m3 ice sediment moved/m3 flow),
we calculate 1E-6 m3/s flow per m2 * 2.5E-4 m3 ice sediment moved/m3 flow

= 2.5 E-10 m3 ice sediment moved/s per m2 surface area during rain events at Bohai Sinus

For sediment load calculations at Huygens Island Stream:
Averaged Stream discharge for Huygens Island stream : 9E2 m3/s
Drainage Area (km2) for Huygens Island stream: 1.5E8 m2 (= 154 km2 in reference)
Solvent exposure = 6E-6 m3/s of flow (solvent flush) per m2 of surface during basin saturating rain event
Again, assume 0.2 kg sediment load/m3 flow:
6E-6 m3/s flow flow per m2 * 2.5E-4 m3 ice sediment moved/m3 flow

= 1.5E-9 m3 ice sediment moved/s per m2 surface area during flooding rain events at Huygens Island Stram.

An average estimate of both values is 9E-10 m3 ice sediment moved/s per m2 surface area during flooding rain events

Converting the value from post above into m3 * m-2* s-1 units we get:
4.4E-7 m3 sediment moved *m-2 surface area * year-1
Dividing by (9E-10 m3*m-2*s-1 rain event) we get

= 488 s flooding rain event/year on average in the Equatorial Zone in Titan.

Or 1 day (24 hours) of flooding rain on Titan every 177 years.
Or a 2 week-long deluge (1 Titan orbit) every 2400 years.
Or a big month-long deluge every 5000 years.
On average, it's pretty dry in the Equatorial Zone.

-Mike

Checklist of items for future Titanauts.
.......
umbrella
.......


What about the rain coming from the ground ?
Will there be an effect from the methane possibly escaping from underground ice clusters ?

Is there another crater inside Menrva?

Staring at the erosion on the rim of Menrva for the umpteenth time, and it a small fresher crater near the SW wall of Menrva jumped out at me. Context image of T3 section (PIA03555 does a good job of showing this view):
Click to view attachment
I think this impact might have been responsible for the breach and drainage of the outer area to the SW into the Menrva crater basin. (The RADAR rim can be traced by the foldover pattern. It is deformed outwards and around the crater) The purple box shows the breach area, right next to and alongside the pedestal of the putative crater inside Menrva. Blue lines show drainage pattern, red and green putative crater rim and debris field. Annotated and "naked" pictures from the T3 RADAR Swath:
Click to view attachment[attachment=12459:Menrva_3__naked_
jpg
(Anyone else noticed this before?)

-Mike

Here's an Earth analog of a crater in the Multiple-Breached erosion state: The 15 Mton nuclear bomb on Bikini atoll left a ca. 1.5 km crater in the atoll. Over the past 50 years, this has been eroded by wave action, overwash from storms, and a few close nuclear bomb tests as well (they reused bomb test sites).

Click to view attachmentClick to view attachment

What I found interesting is the pattern of rim/apron breaching at the SE quadrant (indicated in second image). Most of the material in the SE was emplaced at the original (and accidently too large) Castle Bravo test. Looking very closely at the Google Earth image at [11deg41'51.70"N, 165deg16'29.8"E] you can see the pattern of the undersea banks which show the channels in the breached apron.

An aerial view of the crater and atoll showing overwash can be seen at this site: http://www.radiochemistry.org/history/nuke...rater636c10.jpg

Here is a tracing of the Castle Bravo crater showing the emplaced and breached banks in the SW quadrant. (The original atoll has been grayed out):

Click to view attachment

And here are images of two craters on Titan which display similar patterns in the eroded rim/mantle/apron structure (both much larger than the atoll crater):

Click to view attachmentClick to view attachment

-Mike

So what are you saying here Mike? Comparison with terrestrial analogs suggests Titanians
wiped themselves out in nuclear conflagration ?

That wasn't it, Ralph; everyone knows that Soylent Blue was made out of Titanians (and was best served chilled)...

Very interesting comparison, Mike. These eroded craterforms in low latitudes make a compelling argument for fluvial erosion, all right, but from what source? The processes we see in the equatorial regions right now from Cassini seem to be almost exclusively aeolian (dunes & more dunes), whereas your Bikini analog is fluvial.

What's making me bump my head here is that heavy monsoons could certainly provide an adequate mechanism to erode these craters as you've described, but why the hell does the Equatorial Sand Sea even exist if these monsoons do occur? Can massive dune structures as observed reconstitute themselves in less than 15 Earth years or so after presumably extensive flooding?

Obviously it seems that Titanians have better weapons than we do. We must surrender immediately.



I would state that "it is interesting that a crater formed by a nuclear explosion and subjected to fluvial and tidal erosion on Earth appears similar to craters on Titan formed by impact and exposed to conditions in the sand sea basins".

I would speculate that the impact craters shown on Titan may have been subjected to intensive fluvial and even tidal erosion from a past hydrocarbon ocean long ago. Then as the climate/atmosphere changed, it went to an extremely arid climate with occasional monsoons that we observe in the present epoch. The occasional (seasonal? maybe....) monsoon phase is enough to help locally erode rims, fill channels, emplace alluvials, but not big enough to destroy the dunes or make permanent oceans that would be a sediment trap.

Something really tremendous made the Huygens Channel, but it wasn't a big monsoon. It had to be much, much larger.

The now-dried terrain of Lake Bonneville [E California, all of Nevada, W Utah) might be the best Earth analog of the present Equatorial Sand Seas. Just much, much drier, with rarer but still impressive gully washers, and ancient ocean sediments underneath fresh alluvial mantles.

-Mike

Recent article in space.com suggesting a megaflood is now thought to have shaped a Box Canyon in Idaho. Previously, it was thought that slow, steady erosion carved out the canyon.

Link to article here: http://www.space.com/scienceastronomy/0805...nyon-flood.html

(Replace the word "Titan" for "Mars" and the word "methane or nitrogen" for "water" and "ice" for "rock or basalt" in the space.com article).

The Science article is not yet available.

-Mike

I can't help but wonder about the contribution of subsurface erosional agents such as methane in the breakdown of crater walls, rock-ice islands and the older continents like Xanadu. Perhaps just calculating the rainfall over a region doesn't give a complete picture of the fluvial erosion at work on a region. For example the Huygens landing site looks pretty unimpressive from Cassini heights above the ground, the striking branching rivers (1) and what appears to be methane springs (2) don't show up, yet DISR images from Huygens (providing "ground truth") shows prime examples of sapping in the arid equatorial zone landing site.
Likewise on many of the focally eroded craters it appears dark erosional channels are arising from high up on the wall of the crater (3) rather than from surface run-off implying an erosional agent flows underground emerging as springs. The increased "whiff" of methane upon landing of Huygens, the abundance of CH4 on the spectral analysis of the air at ground level, the haze down to ground level seem to speak of methane just under the surface.
1) Click to view attachment 2) Click to view attachment 3) Click to view attachment

What we don't know is how porous Titan's surface material is. Or how much this varies over the surface.

Really hard impermeable stuff might cause a large amount of runoff during a torrential downpour. (Like the dendritic pattern in Figure 1 above). Lots of erosion and carving right from where the raindrops hit.

Really soft porous stuff might allow methane rains to floof right into the surface and percolate underground before emerging as sapping springs (like the stubby pattern in Figure 2). This would have limited erosion at the surface until right where the liquid collected and sapped out into a stream.

I think Huygens Island is evidence of both types of terrain. To the NW, the dendritic pattern shows a relatively impermeable terrain, while to the SW, the stubby pattern shows a very permeable terrain.

Without having any absolute topography data as yet (T41 + T8???), (save for the DEM's from the limited Huygen's view), I'd still be pretty confident that the Huygens Island is a local topographic rise. I'd be harder pressed to come up with an explanation for how the local methanofer is pumping out liquid on a topographic high. The most simple explanation for Huygens Island is that the origin for the fluid originating on the was from rainfall coming from above.

Below is an image that has been bugging me for quite a while. Why does Ksa have a channel skirting the base of the debris apron on the downwind side? Could the impact basin have compacted the sediments and caused the base of Ksa to be cut off from underground communication? Could this then cause it to fill like a cup during "fluid flow" events (above ground or below) and then slowly percolate out the base of the debris apron. Or does the debris apron causes a topographic block and the eroded ice sands are less prone to percolation and you get a short surface flow over this section. (Note how the dark channel goes into the sand sea area and then quickly dissipates away. Methinks the sand sea basins are very very porous)

Click to view attachment

Is the related to why the anabranched streams of Menrva originate well away from the crater rim? Is there percolation as well through the debris apron of Menrva? Did Menrva fill up like a crater lake during monsoon events (big delta deposits in there as well as the canyon breach on the SW side) and then slowly percolate out at the base of the debris apron?

Could the permeable vs. non-permeable terrains types explain the lake morphology in the N?

Could it also explain the canyon-land terrain (soft and percolable terrain) in the T39 RADAR swath compared to the big delta deposit in T7 (hard and impermeable)?

-Mike

Very interesting analysis. I note that the assumed impact craters often appear very dark in radar images. And the impact crater you show here is the most intriguing to me. Its radar appearance seems quite uniform and almost as dark as the arctic lakes. And the encompassed branching extension is a sign of ice fractures or flows.
Do you think it's possible to envisage that a blanket of water ice is floating on a liquid hydrocarbon layer in the area of the presumed impact crater ? Is it physically possible?

No. Water ice is denser than liquid hydrocarbons

Largely true, but see Alex Hayes paper just published in GRL on this question. He touches on it this in
his DPS abstract from last year...

Title:
Titan's North Polar Region: Lake Distribution, Statistics, and Implied Methane Hydrology from Cassini SAR
Authors:
Hayes, Alexander; Aharonson, O.; Lewis, K.; Mitchell, K.; Lunine, J.; Lorenz, R. D.; Wall, S.; Mitri, G.; Elachi, C.; Cassini RADAR Team


Recent observations of Titan's Surface from Cassini's Synthetic Aperture Radar (SAR) have revealed quasi-circular to complex features which are interpreted as liquid hydrocarbon lakes (Stofan et al., 2007). We use the global distribution of lake features to investigate methane transport in Titan's hydrologic cycle, which includes atmospheric, surface, and sub-surface interaction. Specifically, the latitudinal and longitudinal division of hydrocarbon lakes combined with derived topographic information is used to model subsurface transport and place limitations on the properties of an isotropic porous regolith. Our analysis of the dataset, which covers 22% of the surface, has led to the identification of multiple lake morphologies which are correlated across the polar region. Radar dark lake features are limited to latitudes above 65°N and vary in size from the limits of observation (a few km2) to more than 100,000 km2. Granular and sub-granular lake features, which are distinguished by increased radar backscatter relative to their surroundings as compared to dark lakes, can be found as low as 55°N. Sub-granular lake features are inferred to be empty basins while granular lake features are interpreted as transitional between dark and sub-granular. The orientation, size, and statistical correlations between dark, granular, and sub-granular lake features provide constraints on precipitation conditions and the importance of subsurface transport. Using preliminary porous media properties inferred from Huygens probe results at 10°S, timescales for flow between observed dark and empty lakes are calculated by solving the groundwater flow equation. Derived lake equilibration timescales are compared to the time between collocated SAR observations in order to place limitations on the permeability of an isotropic porous regolith. For permeabilities of 10-5cm2, equilibrium timescales are found to be in the 10's of years and are similar to Titan's seasonal cycles and lake evaporation estimates (Mitri et al., 2007).

Maybe not liquid hydrocarbons....water ice has a density of about 0.9 g cm-3 (= 0.9 g mL-1). Off the top of my head, most normal hydrocarbon stuff has a density in the range 0.3-0.8 g cm-3. [The exception is with halogenated solvents which are usually much denser than 1 g cm-3 - there are probably not a lot of halogenated solvents on Titan.] Methane is the least dense of the hydrocarbons - makes sense, in it's molar volume you've got four atoms, three of which are hydrogen. So in principle the water ice chunks should sink to the bottom of a hydrocarbon layer.

But if the ice layer is particularly porous, it's density might get lower, but still might not be enough to float on top of the methane/ethane phase compared to foamy stuff on Earth. (Presumably the macropores will be filled with N2 or methane gas).


However, an impact in the Sand Sea basin will throw up ice crust debris on top of the solid organic dunes sands. So this will make an metastable system where the heavier solid is lying on top of the lower density dunes sands. [I'm using dielectric constant and spectral characteristics to guess at relative densities. - Based on VIMS and dielectic constant the crust should be more water ice rich than the dune sands. More ice rich stuff should be denser and closer to 0.9 g cm-3 compared to organic stuff.]

If the dune sands are kinda mobile, they might act like a deformable solid mass and slowly work their way up through cracks in the icy debris. This could be similar to a salt dome or maybe more like a sand pipe.

[I really like thinking of the dune sands as like little plastic beads. Put a heavy object in a box full of plastic beads, give a good shake and the heavy object should sink down.]

I strongly suspect that organic material has infilled many of the craters. This might not just be dune sands. This material could have come from below, carried in by subsurface methanofers. This might be the same material and process that filled in the Shiwanna Virgae graben in N Tseghi. This would explain the VIMS deep black unit seen in the landlocked Shiwanna Virgae graben.

One current theory is that the dune sands formed by a sintering process as they drifted down to the surface. My speculation is that the dune sands formed as dark blue ice sand material rolled around in the Deep Black organic material during ocean events and gave the ice sands a coating of "more polar hydrocarbon gunk" (soaplike stuff - with the polar parts of the gunk molecule coating the water ice and the nonpolar part sticking out into the hydrocarbon liquid.) Sort of like making breaded chicken nuggets. As the oceans dried, the non-polar sands got mobile. (Tar-coated water ice). This hypothesis explains why the dunes sands appear limited to Sand sea basins, sintering should occur globally.

The Deep Black material is hydrocarbon-dissolved gunk material moving around in the subsurface and oozing out in the deep basins, deep graben, deep craters, or being coughed out of some features like Omacatl Macula.

(According to this hypothesis, there should be a dune sand layer trapped at the bottoms of the polar lakes, but buried under a second sediment layer of airfalled organic material.)

-Mike

If I understand the abstract correctly, they were concerned with the SAR RADAR response level of the lake material: the dark lakes are up north, and the grey "damp" lakes are a little lower, and the dried crusty remnants area lower. IIRC from the Titan flagship proposal, topographically the north poles dips down into a very broad depression. So it might be a combination of latitude and base altitude that determines the RADAR return (translating to liquid, damp, or dried materials).

Another analysis is discussed here, here they discuss shoreline morphologies:
Mitchell et al., Ices, Oceans, and Fire: Satellites of the Outer Solar System (2007) Abstract 6042. "Titan's North Polar Lakes as Observed by Cassini RADAR: An update." (freely available here)

From the abstract:
"On-going mapping efforts have revealed that lakes of a given morphological class often appear in clusters, with steep-sided small lakes more often appearing at
the lower latitudes. Many lakes are seen to be fed by channels, some short and stubby, indicating intersection with subsurface liquid methane reservoirs (equivalent to aquifer or water table), others long and sinuous, probably indicating that they are fed pluvially or via artesian springs. This plethora of landforms suggests a dynamic system of liquid hydrocarbons, equivalent to terrestrial hydrologic system (informally we use the expression “methanologic”), underlain by variable surface materials [1,6,7]."

Another description is found in the Mitchell Titan lakes CHARM presentation (20071127) freely available here (WARNING: 6.3 Mb).

In slide 2, he has classified the lakes into:
"Shoreline classes:
• Steep margins: seepage or methanifer
• Diffuse scalloped margins,decrease backscatter towards lake centre: drainage basins"

and in slide 7:
"T19 revealed one more morphology: larger lakes, rugged coastlines, like Lake Powell or Scandinavian fjords"

Slide 9 in the presentation shows a nice graphic with the classification of the lakes into four different regimes:
blue = larger lakes, rugged coastlines
green = shallow sided
purple = steep sided
yellow = dry

There appears to be a regional control to the shoreline style. [The observation that all three dark lake styles are observed between 75-80 N indicates that latitude is not the dominant control]. Is it topographical control or is it permeability control?

Looking at PIA10353, these two rough shoreline lakes appear to be at different levels. The RADAR blacker and apparently deeper-looking lake is primarily colored blue in the chart (only 900 m below 0 level). The RADAR shallower-looking lake to the left is purple, this corresponds to 1200 m below 0 level. If there were perfect communication between the two lakes, the lake on the left should be dark and deep, while the lake on the right should be shallow. Assuming that the difference between the two topographic shadings is significant (and it might not be) it appears that these two rough-sided lakes are not in equilibrium. They have permeability issues.

-Mike

Mike, I have long suspected that permeable vs. impermeable must be be a fundamental parameter in characterising the different surface types on Titan, including different types of channels and different types of lakes. Your detailed case by case consideration of a number of locations with this in mind - much more specific detail than I have at my command - is therefore of great interest to me (even more so than usual. ).

On the question of ice floating on liquid methane I think this is possible if the ice is in the form of nitrogenated ice-pumice, in other words flash-frozen soap suds.

I have mentioned before the possibility that many of Titan's surface materials, including dunes, buoyant pebbles and some at least of the highland areas may turn out to have surprisingly low densities. This has already been observed for parts of Xanadu.

The Huygen's landing site has sapping (S), tectonic (T) and dendritic (D) patterns in the bright terrain near the lower, darker landing site terrain. Actually one could argue the dendritic channels have rather short stubby tributaries that are of the sapping type as well. I may have missed an earlier post but what of the bright channel in the left lower corner of the image? Has there been any further discussion on what this channel is? (early speculation suggested it was subsurface water ice extruded to the surface.)
Click to view attachment

Its been a bit quiet lately on this Titan blog so I’ve been looking at images of other worlds lately! I was struck by several earthly photos from TIME's Nature's Wonders,( K. Knauer ed., 2008) a beautiful new special publication about Earth's natural processes. Couldn't help but do a little Earth-Titan look alike comparison of two of the photos from this great glossy; on the left is a channel in a Greenland ice block and its look alike in "Huygen Island" on Titan and on the right are two famous faults, the San Andreas, California Earth and one from Xanadu, Titan.

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

(images reduced and may be subject to copyright)

While the appearances are similar the state of the elements and direction of the forces responsible differ. Greenlands ice block is melting and water is liquid in the channel. Titan's channel may be tectonic and the 'water' in it extruded hard-rock water-ammonia ice mix. The San-Andreas is a slip-strike fault, Xanadu's may be created by compressional or extensional forces.

From which SAR swath did you pull that Xanadu fault? Doesn't seem to be on the T12 west-east pass...

It's from here:
http://saturn.jpl.nasa.gov/multimedia/imag...fm?imageID=3082

For those of you with access, this is now in press at Icarus, which seems pretty relevant to the discussion here.

"Fluvial Erosion and Post-erosional Processes on Titan

Ralf Jaumann et al."

There's not yet an abstract for public viewing, but there will be soon I suppose. It's rather long, so I won't quote it fully, but here are the main results:

" Thus the observed surface erosion fits with the methane convective storm models as well as with the rates needed to transport sediment."

and
"Only frequent storms with heavy rainfall or cryovolcanic induced melting can explain these [massive eroded area seen by VIMS at 7S, 30W, with changing spectral characteristics] erosional features."

Another recently released paper:

Lorenz et al. Planetary and Space Science 56 (2008) 1132-1144. "Fluvial Channels on Titan: Initial Cassini RADAR observations."
(Pay-for article: link to abstract here.)

Key quotes clipped from the abstract (my juxtapositional bias):

"An overall impression from data so far is of radar-bright but shallow channels at low latitudes where fines have been removed, more deeply incised channels at mid-latitudes, and radar-dark sometimes meandering depositional channels at high llatitudes."

"These observations show that fluvial activity occurs at least occasionally at all latitudes, not only at the Huygens landing site, and can produce channels much larger in scale than those observed here."

"The corresponding global sediment volume inferred [by fluvial channels incised in bright terrain] is not enough to account for the extensive sand seas."

-Mike


[EDIT: fixed link]

Thanks Mike. For some reason I can't get that link to work, but I found this here:
http://adsabs.harvard.edu/abs/2008P&SS...56.1132T

Is that the same abstract?

Interesting that channels preferentially flow eastward. Is this evidence that some surface flows are wind-driven rather than gravity-driven? A sort of fire-hose effect? It can't be downhill all the way going eastward (though some Californians may disagree).

Thanks, Nigel! I corrected the link.

There appears to be a W to E directional gradient as well as a poleward directional gradient in the areas observed.

According to the authors, the W to E gradient was observed in T3 (fluvial) and Ta (cryovolcanic flows). Both of these are in the same rough general area NE of Xanadu, so it could be a broad regional gradient.

From the article:
"Taken together with the southerly flow of the T7 and T13 channels, this can be considered indicative either of central or north Xanadu being topographically high (at least at the time of channel formation), or perhaps of a general equator to pole topographic gradient."

[An equator to pole topographic gradient would be the optimum solution if the crust was freely allowed to shift around due to polar wander. It would place the bulgy stuff (topographic highs) at the equator. An altimetry trace that was in the Titan Flagship mission proposal seemed to indicate a topographic drop to the north polar region. Hmmmm. That's where the lakes are as well. Is this all starting to fit together?...perhaps......]

"Despite the broad indication of slope, the morphology of the channels suggests overall slopes are quite gentle across the T3 region (at least at the time of channel formation) otherwise meandering and braiding or anabranching would not occur"

From the article, it appears that the gradients on Titan are subtle and possibly regional in scope.

(And there are dark "lanes" that I'll attribute to sloughs in the T25/T28 region that appear to have flowed westward. For example, see the identified RADAR darker features in this zoom of the T28 Swath (Mountains of Titan thread, post 6, link here.)

-Mike

I agree it's possible that the eastward flow bias could just be observational selection and that gradients always rule the flows, but I think the observations at both small (Huygens) and large scales do leave room to question that terracentric assumption.
In terms of the forces they can exert the ratio "Wind/Gravity" is so much bigger for Titan.

Regarding the net flow poleward - that's just another thing to add to preferential deposition of organics at high latitudes that is redistributing mass away from the equator. These processes can't go on indefinitely without eventually destabilising the system and causing a 'polar flop'. The big unknown - again - is the relative timescales involved. Note also that an area can be topographically high and still have a mass deficit if the terrain is constructed from low density materials. I'm betting that the northern seas and seabeds are considerably denser than parts of the Xanadu highlands.

Geology and even more so, Titanology are not my specialties, however;

A large portion of the Titanian surface covered with ~ <1 kilometer high dunes, will show a much higher number of kilometers of channels on the gentler leeward slopes as opposed to the steeper windward slopes. And this pattern can persist over the entire surface. Should windage be primarily from W to E, the eastern slopes are a higher % of the total area.

This might factor into the skewed Iowa stats, ~25% SW, ~75% SE

(Windward surface flow from equator to pole will also bias stats N and S of equator.)

I'm not so sure.

The northern seas and seabeds are *probably* made of lower density hydrocarbons and organics.
(Methane density about 0.45 IIRC, other hydrocarbons more like 0.6).

The mountains and crustal stuff is *probably* water ice based with a density of 1 (or 0.9). So the topographically high mountains would also have the most mass. The lakes and organic sludgies in the northern polar lakes would be lower density stuff. Any organic schizzle deposited would also be pretty low density (and if's it fluffy, which I suspect it is, even lower density).

Low density stuff lying low at the north poles, and topographically high (relatively) and massive (relatively) stuff at the Equator = stable configuration for polar wander. Dunno how the South polar terrain fits into the equation. I haven't seen any altimetry traces or much RADAR data released. (There don't seem to be as many lakes, does this indicate that it is higher terrain?)

-Mike