Inferred Orientation of Distal Ejecta

Showing 1-59 of 59 messages
Inferred Orientation of Distal Ejecta Cintos 12/3/09 8:32 AM

Our post here is to explain the conceit we leverage in our evaluation of the Carolina bays. This is our first post in our presentation of our proposed Saginaw Impact Manifold.The thread has been expanded as the trial-and-error approach we are taking proceeds. We encourage new readers to participate by reading each post in sequence.

Our working hypothesis holds that the Carolina bays structures are surface features in energetically deposited, highly hydrated, distal ejecta from a cosmic impact. A challenging aspect of the commonly proposed relationship between the bays and a cosmic impact involves the lack of an identifiable impact structure.

Most attempts at following the inferred orientation of the bays back up the trajectories' bearing have failed to produce a focus. We propose this to be caused by at least two variables not considered. First, that the impact may have been an oblique event, which would infer a chaotic focus, and secondly, that the earth rotates during any realistic ejecta loft time, which we attempt to evaluate with the attached kml file. A third variable is the proper accounting for the west-to-east ground-velocity vectors that will be resolved when the ejecta strikes the earth. We will discuss that later; they are measured in m/sec vs km/sec, but were interacting with the relatively slow terminal velocity of the ejecta as it traversed the atmosphere and therefore will prove to be vitally important to the numerical model.

Shoot at where the target will be, not where it is right now

During the time period extending from the moment of the source impact to the eventual deposition of the distal ejecta, we see the de-coupling of the spatial coordinate reference systems in multiple dimensions. The decoupling is driven by the spherical nature of the "playing field" when trajectories cover significant distances. The common term applied to the effect is the Coriolis Force, which is a kinematic force applied to an object to "force" it along a great circle route as a object proceeds along its trajectory. For example, if an object is launched with sufficient velocity on an azimuth of 90 degrees from latitude 45 north (i.e. due East), it will follow a great circle route as it begins to "circle" the earths spherical surface. The cartesian coordinate "bearing" of our example object begins to "turn" south, and eventually the object will cross the equator on an azimuth 45 degrees increased, or 135 degrees. But the above analysis does not account for the fact that the earth is also rotating during any real-world loft trajectory period.

Here we will evaluate the effect on the cartesian cordinate system of azimuths and bearings when the Earth's rotation is considered. The Google Earth facility is employed with the attached set of kml.

During the 12 minute loft time we are modeling in this post (and attached kml), the Earth will rotate three (3) degrees of longitude from the west to the east (regardless of your location on the earth). Therefore, the landing location of the ejecta will actually be three degrees westward of the initial "target".

When the ejecta is deposited at the eventual location, it will still bear the original flight azimuth/bearing. If those geometries are followed back along the trajectory, the focus will be on a location three degrees west of the original launch location. Thus, we feel justified in applying the conceit that, from the perspective of the distal ejecta landing site, the inferred bearing would point back to a suragate impact crater. The suragate would offset on the global map by one degree of longitude westward for every four (4) minutes of loft time.

Please consider, also, that the loft time is a variable affected by both the launch velocity and its loft angle. A trajectory can be generated for a given landing location using shallow lofts (and short transit times) as well as higher loftings which would take longer to get to the same location.

- Michael

Men occasionally stumble over the truth ... but most of them pick themselves up and hurry off as if nothing had happened.
...... Winston Churchill
Re: Inferred Orientation of Distal Ejecta Hill 12/3/09 1:46 PM
Cintos' post refers to the study of the reasons for the origin of the Carolina Bays. There is a thread about the bays and other evidence concerning their origin here.
Click the images to view two important videos about our only home.
A Pale Blue Dot HOME

“Only when the last tree has died, the last river has been poisoned, and the last fish has been caught will we realize we cannot eat money.” Cree Proverb

My avatar Mr. Raven posed for me at the Tower of London.
LiDAR Mapping of Goldsboro Ridge Cintos 12/4/09 8:00 AM

We are utilizing the new Light Detection and Ranging (LIDAR) technology to enhance our view of the enigmatic Carolina bays. The cookie-cutter, repetitive, robust adherence of the bays to a common shape suggest to us a catastrophic origin. The inspiration for our conjecture was an observation in the paper The Goldsboro Ridge, an Enigma, by R. B. Daniels, E. E. Gamble and W.H. Wheeler, 1970.

The Goldsboro ridge is a unique feature on the Sunderland surface and requires special explanation whatever its origin. It must be either an erosional remnant of a once more extensive sediment or a depositional feature. ...The Goldsboro sand overlies the Sunderland Formation conformably. The contact is always abrupt but there is no evidence of deep channeling, basal coarse material, and evidence of weathering at the contact. Even the Carolina Bays do not disturb the underlying Sunderland materials.... The sand in the bay rim is not different from the Goldsboro sand. Therefore, these Carolina Bays are merely surface features associated with the formation of the ridge.

Hence, we speculate that the bays were formed as surface features during an energetic deposition of sand and water (as a super heated slurry) from a remote cosmic impact.

Note that the adjacent bays were created at several elevations. Our conjecture that the bays are surface features in a thin 1- 10 meter thick blanket overlaying the original terrain is supported by this evidence. The elevation profile graphic below was created in Global Mapper.

The attached KMZ file includes details of the Goldsboro Ridge, NC field of Carolina bays. It includes outline overlays of multiple Carolina bays and DEM color-hinted elevation mappings in overlay form.
Visualization of the Proposed Crater Cintos 12/7/09 3:01 PM

This initial post has been edited to reflect the current status of the hypothesis.

For many decades, maps have been drawn which attempt to identify the source of the Carolina bays by using the inferred long-axis orientation of the various bays as an arrival bearing. Drawn of "flat" maps, those attempts failed to account for the fact that the flight path of any meteor, shock wave or ejecta product would not show up on a projected map as a straight line, but rather be a curve following a great circle route. Hill previously posted the classic map from the Eyton & Parkhurst paper in the associated thread HERE.

Google Earth allows for the creation of true great circle lines using the "Add>Path", or by programmatic creation of the appropriate kml. A quick test of paths drawn back from numerous Carolina bays confirms the general trend to the north-west. Using the measured alignments of an initial 40 Carolina bay fields, we generated great circle paths for visualization in Google Earth. This yielded a fuzzy triangulation locus centered at 43.5 N, 89.5 W. Our analysis implies that a great circle triangulation would yield an erroneous “surrogate” impact location, offset to the west. A flight-time adjustment of the crater eastward along the 43.5º N Parallel (taking into account the Earth rotating .25 degrees of arc every clock minute) directs us towards the actual impact site (discussed in more detail in HERE ). We heuristically examined various geological depression found along that transit, and selected the Saginaw area of Michigan’s Lower Peninsula (Lower Peninsula) for further analysis.

Why oval? Craters are circular, right??? Elliptical craters with "butterfly" ejecta patterns make up roughly 5% of the total crater population of terresterial planets and moons. They are caused by impactors which hit the surface at oblique, or very shallow angles, such as this one from Mars:

Image Credit: NASA / JPL / ASU / mosaic by Emily Lakdawalla

A vitally important detail of this lateral ejecta pattern is how well it fits into our proposed distribution of ejecta, both eastward and westward. This topic is discussed in more detail on our web page Oblique Impact. Here is how the Mars oblique impact crater looks when superimposed on our suggested impact site:

You can view a Google Earth mashup of this crater and our proposed triangulation from the Carolina bay orientations using the attached KMZ file.

One of our premisses is that the area was covered 40,000 years ago by a 1-2km thick sheet of ice from the last glacial period. Thus, any crater would be cut more from the ice sheet than from the lake basin, as it is less than 500 meters deep - rather shallow as craters go, even oblique ones. Here is a graphic created to explain the strata which the crater was cut into. Vertical scale is greatly exaggerated!

In our next post, we will provide an "INDEX" kml that shows the 30-some Carolina bay sites with links for retrieving more detailed kml for each individual ejecta field, one at a time.

Best wishes,
Carolina bay Field Indexes Cintos 12/7/09 9:30 PM

This research has been proceeding heuristically, and such a trial-and-error hunting expedition is hard to present in a logical sequence. We have shared the crater identification, now we will try to support the selection.

The attached kmz file titled "Distal Ejecta Fields", contains a list of ~250 Carolina bay "fields". Google Earth will open with each of these shown as a placemark at their global position. Clicking on any one of them will bring up a display that discusses the site, present a "portrait" of our imaging overlays, and provide a link to a more detailed set of kml.

When the kml link is selected, you will get a folder that contains several standard items. I am using network links for the LiDAR kml data, so the kml is small, but the network load as you hit these could be high. There will be a placemark for control, a "Bearing Grid" which includes a reticle to help determine the bay alignments, color ramp elevation LiDAR or DEM overlays for bay identification, oval overlays on some individual bays, occasional "photo" links to present elevation profile data, and in two cases, a sample 3-D loft trajectory.

Thanks for participating. We welcome comments, and recognize that the catastrophic nature of our conjectures are way out of the box, so we don't expect much in the way of acceptance at this point.
Carolina bay Alignment - Resolves on Contact Cintos 12/11/09 8:57 AM

In my first post in this thread I mentioned that a 2nd factor was superimposed in the inferred alignment of the Carolina bays, and that there would be "more on that later".

While the Coriolis Force component is systematic by loft time, this additional factor is systematic by latitude. In the first case, the earth rotates at 0.25 degrees of longitude per second, regardless of where on the earth the consideration is applied. In our second case, we rationalize that the ground speed of any particular spot on the earth is a function of the cosine of its latitude. The end cases are the poles - where the ground velocity w>e due to rotation is negligible - and the equator - where the ground speed w>e is ~1,670 km per hour.

In our specific YDB Manifold, a relevant set of w>e velocities would be our Saginaw Bay impact site - rotating at 1,218 km/hr - and a generic ejecta filed such as Bishopville - rotating at 1382 km/hr. At time of contact the the 165 km/hr w>e velocity difference will resolved by a skewing of the "splat", effectively rotating the inferred bearing in the clockwise direction.

By way of explanation, a droplet of ejecta traveling from the north to the south in its great-circle frame of reference would not be affected by that ground speed difference until it lands on the surface of the earth. At that contact, the west-to-east ground velocities will be resolved. Given a point object such as a missile warhead, this resolution need not be considered. When an object as large as our posited ejecta droplets (say, 100m diameter) lands, this resolution will skew the inferred arrival bearing. In our cases the velocity difference will actually subtract from the expected w>e velocity vector. Since the n>s velocity vector is constant, the effective alignment rotates clockwise slightly.

A relevant analogy would be the "Crab Angle" velocity vector required to compensate for cross-runway winds during an aircraft landing approach. Given high enough cross-winds, the visual effect can be stunning. Take a look at these four sequential photos of a 747 landing. Clicking HERE will link to a web page with two additional movies.

Now, in our situation, it is the land and the atmosphere that is moving rapidly west-to-east under the falling ejecta. That effect increases in magnitude as the landing sites move more southerly. If there were ejecta in the Everglades, for instance, the ground speed difference would be 274 km/hr.

How relevant is this? It may be considered to be a minor factor in the more northerly fields, where there is less w>e velocity differential to detract from a relatively high w>e loft vector. In the more southern fields, the adjustment becomes a much higher factor both in magnitude, and in respect to the smaller w>e loft velocity vector. When superimposed on the counter-clockwise effect of the Coriolis (systematic by loft time), the latter may tend to overpower the former. We suggest that when moving north to south through the fields this general shift of the alignment vector clockwise is seen as "systematic by latitude".

And the results of both? Chaos within a bound set of results. We assume that a compute model can be developed that will iteratively solve for a set of ejecta loft parameters from the Saginaw Impact area (ejection site along butterfly, loft azimuth, loft angle, loft velocity) that would satisfactorily correlate the inferred alignment of each of the Carolina bay structures with the empirically measured results.

Thanks for participating,
Visualizing Possible Skew Cintos 12/21/09 1:44 PM

In my last post I discussed the "Systematic by latitude" alignment factor, driven by the difference between the west-to-east velocities of the impact site and the ejecta landing site. If you accepted for a moment that the Carolina bays were distal ejecta, we can presume it was coming from the north and west. We have run some calculations and determined that the area in the photo below was traveling about 40 meter/sec faster west to east than an impact site at 45N latitude, a velocity that must be resolved as the ejecta strikes the earth. We expect that the entire sheet moves in unison as it "resolves on contact", but we see what might be a different signature in some situations.

The photo below shows what we interpret as being a "skew" of a very watery droplet as it hits the ground. As the droplet spreads, it slides to the west, leaving a trail in its wake. We attempt to provide a visual explanation in the tour "Bowman_Skew" in the attached kml.

Alternatively, these may merely be oxbows. If so, someone has to explain the mechanism necessary to close them and get them all aligned. I have been removing some "fields" of ejecta that are likely oxbows, such as the ones I had for Allenhurst. Also, how it is possible that oxbows - the signature of lazy rivers - could be effective across the significant elevation differences seen in the local bays?

Note that the Global Mapper elevation places the floor of the small bay at top left at 52 meters, while the bay in the lower right is below 40 meters. The distance between them is about 20km.

Some of the bays in this LiDAR image display the concept of "overprint", where a bay looks to be created on top of an older bay. If appropriate, in a later post we will explore this aspect of the enigmatic bays, along with examples of where these overprints are similarly aligned and there they present different alignment bearings.
Crater Hunt II - Saginaw Bay Cintos 12/27/09 8:50 PM
Our initial test fit of Lake Michigan as the proposed cosmic impact site was quickly demoted in light of the solid evidence supporting its glacial heritage. Our second try at identifying another candidate is underway.

The evaluation of the various systematic adjustments to the inferred ejecta trajectory has suggested a rather chaotic set of variations can exist, but that the first order effect is the loft time shift. A mean average of all optimum trajectories was used to generate a proposed single point loci for an impact point. The location at 43.6259 North Latitude and 89.7043 West Longitude was computed. We extended the green line seen in the graphic below east along the indicated latitude to identify possible crater sources, using the surrogate crater conceit.

The following graphic Compiled at the University of Michigan describes the bedrock located within the Michigan Basin geological structure. The central zone is composed of younger, more solidified carbonate rocks, whereas the older underlying rocks are softer shales and sandstones built up prior to the origins of life and the calcium deposits derived from shells. The ice age glacial sheets which removed vast quantities of strata from above the basin were unsuccessful in breaching this carbonate layer to any degree with one major exception - Saginaw Bay.

This leads our search to the Saginaw Bay area. The shift eastward represents a slightly longer loft time of 22 minutes, a higher ejecta velocity and a higher loft ejection angle. Now, the Saginaw Lobe of the Wisconsin Ice sheet is highly likely to have been responsible for this slice into the basin. We are in the process of researching a number of anomalies related to the area in and around the bay. Among the items identified to investigate:

✓ Anomalous remnant ground water oxygen isotope markers water in the Saginaw lobe area; injection of water from Younger Dryas period indicated
✓ Anomalous hydraulic pressures in the surrounding strata layers
✓ Anomalous glacial deposits, juxtapositioned with large bolder fields resting on similarly dated, but smaller sized debris
✓ Identification of Precambrian deposits in glacial till, unique to the Saginaw lobe and not seen in any other Wisconsin-era lobe deposits
✓ Buried sub-glacial runoff channels suggestion the deposition of terrestrial debris on top of glacial sheets
✓ Anomalous buried soil layers suggesting moraine deposits on above warmer climate flora.
✓ Anomalous salt-bearing springs surrounding the Saginaw bay; used for commercial salt production in 1800s
✓ Unusually High Helium Atmospheric noble gas signatures in area aquifer fluxes
✓ Existence of structural anomaly beneath Saginaw bay floor suggested by several researchers; considered to be anticline by some
✓ Carbon dating of natural gas from wells across Michigan Basin show activation ~13kya
✓ Research susggests significant basin re-heating event in past; reactivation of Keweenawan Rift implicated by some

We have created an additional set of kml to visualize this possible crater, and to test fit a set of great-circle lines corresponding to ejecta trajectories from a surrogate crater 5.5 degrees of longitude westward from the Saginaw Bay area. Two points along the length of the crater were used as starting points for the line sets to each of the Carolina bay fields being evaluated. We continue to see general agreement between these trajectories and the actual alignment of the bays.
Bays in Strange Places Cintos 1/3/10 1:55 PM

The attached kml has an updated list of proposed distal ejecta sites. A number of the sites "out west" have been removed, as they likely represent ancient oxbows or more common geological depressions. The overriding concept in declaring a "field" is that there be a substantial number of structures that present similar alignment and length-to-width ratios.

We have added more western fields from the state of Nebraska. Charmingly (or coincidentally?) the ovoid bays in each of these Nebraska sites are aligned towards our first-pass averaged focus at 43.63 North, 89.66 West.

The field of landforms centered at Axtell, Nebraska are known as "Nebraska's Rainwater Basins", and have been likened by some researchers to Carolina bays.

Please note the addition of another "outlier" in the Fields list: The enigmatic Midlothian Gravels, just south-west of Richmond, VA. The bays in this area are formed within gravel rather than sand, somewhat negating a wind-blown dune ancestry. UPDATE: A visit to these bays on 26 March, 2011 confirmed that the rims of these bays are pure sand. While these landforms may well REST upon the gravels, these structures are analogous to the Carolina bays in their rim composition.

There is a useful review of these structures By Bruce K. Goodwin , available online. Like the Nebraska fields, above, these "bays" present an inferred alignment to the first-pass average surrogate impact site.

The teaser graphic below is extracted from that document, analogous to the Cross Section of the bays in the Goldsboro Ridge.

The following photograph demonstrates the flat nature of the bay floor, with a view of the slight rim relief showing in the distance. bay 150310_1259 KML. The view is down a section of railroad bed was abandoned across the bay, and wide open (a buried pipeline is present). In the photo, taken SSE, you can see the paved part of Kings Lynn Road drop down into the bay in the distance. I am just inside the rim here. It is ~ 850m horizontally to the foot of the paved road area, and ~1050m to the top of the rim.
Re: Distal Ejecta in Nebraska? Hill 1/4/10 9:19 PM
Your updates have been interesting reading. Thanks for posting them. smile
Stop the world, I want to Loft Cintos 1/8/10 10:19 AM
Our conjecture holds that the Carolina bays structures are surface features within energetically deposited, highly hydrated, distal ejecta from a cosmic impact. A challenging aspect of the commonly proposed relationship between the bays and a cosmic impact involves the lack of an identifiable impact structure.

Most attempts at following the inferred orientation of the bays back up the trajectories' bearing have failed to produce a focus. We propose this to be caused by several variables not considered by others, and which we have discussed in earlier posts to this thread.

Our selection of the Saginaw Bay area of Michigan, along with the basin area westward across Michigan, as a proposed impact location, allows us to revisit the loft time and trajectories of the identified Carolina bay fields. While we have leveraged the surrogate crater conceit to rationalize the inferred ejecta trajectories bearings, when we evaluate the distance covered in during the event it is appropriate to evaluate the geographic distances involved assuming a stationary earth.

Using a proposed centroid for the Saginaw Bay impact at 43.63 N latitude and 83.94 W longitude, we computed the great circle distances for each of the fields. The chart below presents the fields in a clockwise direction. We feel that the highly correlated loft distances are indicative of a common flight time, loft angle and source for the ejecta at each of the location. The shortest flight distance is found at the Midlothian Gravels site, which may be due to the heavier gravels settling out of the ejecta cloud earlier. Note that the western fields correlate highly with the eastern fields. We propose that the "butterfly" provided slightly more energetic ejecta velocities towards the down-range direction, yielding more loft time and distance.

As discussed earlier in the thread, an oblique impact generating a butterfly ejecta distribution pattern is suggested. We have created a Google Earth overlay which, when visualized on the Google Earth facility, successfully encompasses the identified fields within two narrow bands, one to the east and one to the west.

Let us emphasize here that we fully expect the ejecta blanket may present at distances closer and further than this "ring", but may not have been deposited on hospitable terrain (the Appalachian highlands and the Atlantic Ocean in the east and the Wisconsin Ice sheet in the north, for example).

The attached kml file (Distal_Ejecta_Butterfly) contains the overlay, positioned in a best-fit relationship to the fields, along with the proposed impact structure. When applied to the viewer, the following graphic is generated.

Here is a color ramp DEM graphic overlay displaying the "Saginaw Lobe" geography, which we associate with this crater structure.
Correlating Inferred Carolina bay Orientations Cintos 2/5/10 12:49 PM

Our search for geological formations presenting Carolina bay planforms has allowed us to expand our list of examined locations to a total of 38; 31 in the east and 7 in the west. The attached kml file includes these new sites. Of particular interest is the field at Red Oak, NC. This extensive (8 km x 16 km) set of bays is located quite far inland, in terrain a bit more diverse than others seen.

We encourage you to explore those sites and the associated color-ramp DEM overlays, as they allow these bays to jump off the screen by the hundreds. To review, the Distal_Ejecta_Fields kml element contains placemarks for each field. Clicking on any of those will open a dialogue box in Google Earth that discusses the site, provides a "portrait" of the DEM and orientation arrow, and provides a link to download the full kml element set for that particular field.

At the present time, we have made precious little progress in providing physical evidence to support our conjecture regarding the cosmic impact and resulting distal ejecta blanket. Evidence would need to be obtained from the bay's foundation rim sands, and correlated with data from numerous anomalies identified across the proposed impact site in the Saginaw Bay area of Central Michigan. However, we do feel the exercise here in Google Earth has provided supportive circumstantial evidence, if not an outright proof point.

To review and discuss:

The flight lines, distances and bearings of the Carolina bay fields identified have been analyzed for correlation to the proposed surrogate impact crater site. From our initial bearing analysis, we identified an optimum loci as an average of all "first pass" Coriolis bearings. That value was further rounded for simplicity to a value of 43.5 North Latitudes and 89.5 West Longitude. Our current proposed impact site at Saginaw Bay represents approximately a 22 minutes loft time offset (equating to 5.5 degrees longitude w>e rotation), which suggests a working impact centroid at 43.5 North Latitude and 84 West Longitude.

Our first correlation considers the geographical great circle distance from each field back to the proposed impact centered at Saginaw Bay. As shown in the graph here, a very high degree of correlation is seen.

This suggests several things. First, an ejecta curtain was lofted with a fairly consistent loft angle and velocity. The distances do increases slightly as the landing fields progress radially along the butterfly pattern, suggesting that down-range ejecta was slightly more energetically lofted. lastly, the fields to the west are highly correlated distance-wise with their counterparts to the east. The shortest distance seen, at the Midlothian site, suggests that gravels may have precipitated early from the curtain. Also relevant is a general trend of increased distances as the sites trend towards the downrange-impact trajectory direction.

For each of the evaluated Carolina bay fields, we measured an inferred arrival bearing for the ejecta. These measurements can be validated by reviewing each field in Google Earth using the provided kml elements. Three sets of great circle bearings were generated from each Carolina bay fields back to 1) the surrogate crater centroid (43.5 N, 89.5 W); 2) a point to the NE representing the north-eastern limit of the crater ejection rampart ( 44.5 N, 88.6 W); and 3) a point to the SW representing the south-western limit of the crater ejection rampart ( 42.5N, -90.5 W). These great circle lines can be imported into Google Earth using the SaginawCoriolisBearings.kmz file. The charts presented below plot the inferred bay alignment at 38 sets of sites (western and eastern, ordered in a clockwise walk around the butterfly) against the great circle line bearings back towards the three surrogate crater control points noted above.

Eastern Fields Inferred Bearings vs. Surrogate Crater Control Points

Western Fields Inferred Bearings vs. Surrogate Crater Control Points

Without exception, the inferred bay alignment falls within the bounds of the surrogate crater control points. An obvious trend seen here (analogous to the function in the distance chart above), is that the radial location of the ejecta emplacement site is a function of the the location of ejection along the edge of the crater rampart. Inherent in the graphs is a non-intuitive suggestion that ejecta from the uprange side (NE) of the ellipsoid crater rampart traveled slightly further and on a more down-range bearing than later ejecta from the downrange (SW) rim. (bearings are Cartesian, not relative to the NE>SW centerline of the crater)

The correlation above suggest that the bays are geographic distributed along a narrow and highly symmetrical pair of "butterfly" arcs centered on the triangulated Saginaw impact location. such a distrubution is suggested in much of the current research on oblique impacts. Additionally, the identified down-range "no fly" zone is apparent in the distribution. The following graphic displays the arcs in Google Earth, and also demosnstrates the symmetrical nature of their locations around the implied impactor's azimuth.

We maintain that the high degree of correlation seen in these 38 disparate locations across the continental U.S. support our hypothesis of an ejecta blanket event. Or, perhaps, it is simply an amazing coincidence. If a field of Carolina bay aligned planforms could be identified which does NOT fit into this model it would be definitive negation point. Score 38 to 0.
Kankakee Torrent: Effects of Local Ejecta Cintos 2/12/10 7:01 AM
15,500 years ago found glaciers to the north and east of the Kankakee area melting _ melting very rapidly! Lobes of ice were in areas now occupied by Lake Michigan, the State of Michigan, and eastern Indiana. Moraines to the south pooled Meltwaters, and huge lakes formed. But not for long.“The moraines were breached, and the result was among the greatest floods of the Pleistocene Epoch. This flood had impact not only in the Kankakee area, but in areas far away as well, southern Illinois in fact! The flood is known as the Kankakee Torrent. There were subsequent episodes of flooding, but none so great at this. Guest Column by GeoT, January 5, 2000

When considering the excavation done by these torrents off the Saginaw lobe, it becomes obvious that the volume of water was far higher than that seen from across the moraines of the other Great Lakes during the deglaciation of the Wisconsin Ice Sheet. The common understanding is that a body of water formed against the retreating Saginaw lobe, and at some point it breached the terminal moraines entrapping it. A constraint on the volume of water ponded is the relative height of the Saginaw end moraines, as well as the fact that these moraines are at the highest elevation across the central lower Michigan peninsula understood to have hosted the Saginaw lobe.The lobe's retreat is likely to have occurred in tandem with the Huron/Erie and Michigan lobe retreats, yet those two ponds did not generate an outwash as massive as the Saginaw Torrents. That the Torrent was driven to access the Illinois River Valley by way of the Kankakee suggests that either the Michigan lobe was still in position, or the moraines dividing the two lobes were too high to be broached at that time. The currently available research of the torrent presents a wide range of dates, from 17.k5 kya to 12.9 kya. We are proposing ~15 kya, at the Bölling-Allerød warming period snowfall spike.

The web-based presentation Of Time and the River contains a graphic describing the early-pullback of the Saginaw lobes:

Here is a Google Earth mash up of several elements that allow us to consider an alternative to the moraine-breach concept. The attached kml file contains these elements for your review within Google Earth.

Saginaw Lobe Outwash plain and oversized valleys

We propose that a massive elliptical crater, excavated primarily from the ice sheet, would have quickly become a significant lake in its own right, as it would be surrounded by ~1km thick ice sheet. We suggest that over a short period of time the water level would have risen enough to cause the ice sheet to hydrostatically lift from the terrain along the peripheral edges to the west and south west. Once the undermining got underway, the extensive field of tunnel channels known to exist would have been quickly created. The subsequent catastrophic outflows from the crater basin would easily create both the Kankakee Torrent as well as carving the CKRV. This process may have repeated a few times over the course of the following centuries as the drained pond allowed the sheet to reattached, allowing the crater basin to re-fill again. The channels created are typically 50 meters (150 ft) deep, and often several km wide. They are now occupied by under-fit rivers.

A corollary to the hypothesis would suggest that terrestrial ejecta lofted from the crater floor would have landed on the ice sheet across Michigan, generating vast fields of buried ice. A significant volume of research exists which has attempted to provide a reasonable solution to the extensive anomalies seen in glacial deposits and outburst flood channels in this area.

Several sub topics in our Saginaw Bay discussion are relevant here: Unique Till, Sheetfloods, Buried Soils, Early Deglaciation and Unique Glaciation.
Better KML through Trigonometry Cintos 2/12/10 8:35 PM

With the identification of Michigan's Saginaw area as a probable impact location, our research efforts have been reinvigorated. We can now offer additional approaches to visualizing the dynamics of the impact, and to enhancing the correlations between our conjecture and the landscape. As we discussed earlier, the inferred bearing is a function of the original loft azimuth, which would have deposited the fields of ejecta out in the Atlantic Ocean (for the eastern components) if the earth was stationary. An alternative to the Surrogate crater conceit is to offset the field due east by the Coriolis offset (currently 5.5 degrees), and then use trigonometric formulas to generate a great circle path back up the inferred arrival bearing.

The following graphic displays the Wagram Field , with the programmatically-generated Google Earth paths identified. Many of these elements have previously been discussed and presented in earlier kml files. Our kml generator can now create sets of great circle paths emanating from a point and traveling a suggested distance along a bearing angle. For instance, the yellow path in the graphic starts at the Wagram Carolina bay field and is projected northwest back along the empirically derived inferred arrival bearing ( here it is ~135º). The green path at the bottom right runs to a point 5.5 degrees longitude due east of the Wagram site. The far end of the path defines a "target" location that this ejecta, if lofted from the Saginaw area, would have landed if the earth had not been rotating at 0.25º longitude per minute during a 22 minute loft period. Shoot at where the target will be, not where it is right now.

The next graphic displays the "Wagram.tgt" as a placemark, and another green path segment is generated back towards the proposed crater site, also using the field's inferred arrival bearing to drive the formula. These elements are available for all fields in the KML folder Bay_Triangulation_Computed , and is also in the attached kml file.

In the above graphic, the distal ejecta is hypothetically lofted from near the south western rim of the Saginaw crater, with an initial azimuth towards the "Wagram Tgt" location. However, as the earth rotates west-to-east during the 22 minute loft period, the eventual landing site in North Carolina "rolls" under the falling ejecta. The formula used to generate the great circle lines back from each the fields (both as-landed and the original Target site) is:

lat2 = asin(sin(lat1)*cos(d/R) + cos(lat1)*sin(d/R)*cos(θ))

lon2 = lon1 + atan2(sin(θ)*sin(d/R)*cos(lat1), cos(d/R)−sin(lat1)*sin(lat2))

d/R is the angular distance (in radians), where d is the distance travelled and R is the earth’s radius. θ is the inferred ejecta bearing, as empirically measured at the field.

Where lat 2, lon2 are used to create a point back along the inferred alignment bearing angle. The path is then constructed with the distal field as lat1, lon1. The length of the path (used in the formula above as d) was empirically adjusted to truncate just beyond the identified crater sites.

With this model enabled in our KML generator spreadsheet, we can very easily generate a variety of "what if" scenarios. For instance, it is very easy to regenerate the entire class of great circle paths for our ~40 ejecta fields, using a different Coriolis offset (say 28 minutes). Of course, with the empirical data available, the triangulation (especially with the inclusion of the Nebraska fields) brings forward a fairly tightly constrained loci point for the crater.

Given our new tools, it is even easier to visualize the ejecta trajectories and to reassess and reconfirm our triangulation approach. The final mash-up suggest that the Saginaw Bay area is a probable impact site for distal ejecta lofted to the Carolina bay fields.

Confirmation of this area as an impact will likely take generations of research. Using visualization tools, however, we suggest there is a high correlation from a geometry perspective. Using kml elements in the attached file, the graphic here presents the area in Google Earth using two overlays: a color-ramp digital elevation mapping layer, and an appropriately sized and rotated oval shape file overlay.

We note the strong trend correlation across the northern and souther boundaries, as well as the overlap into the Kankakee Torrent flood plain in the southwestern end.

Experimental and planetary imaging has identified that oblique impact craters generate the deepest excavation at the uprange (arrival) end of the crater. Here, that point for our crater satisfactorily sits in one of the deepest parts of the Huron basin: the Bay City Basin (see the Huron_Bathy overlay in the kml). From that point, the land rises gently across the peninsula, reaching its greatest elevation at the southwest end of the oval overlay. Another common attribute of an oblique impact is the existence of a slight ridge - likely rebounded strata - tracing a line down the center of the structure. In our situation, the Charity Islands rise up from the bay directly along the oval's centerline.

Please note that we fully expect that the Huron lobe of the Wisconsinian Ice sheet would have eventually advanced into the excavated crater from the east, bulldozing the collapsed 1km high ice crater ramparts and leaving the existing set of terminal moraines behind as it deglaciated at the onset of the Holocene.

Thanks for sharing in my adventure...
De Skewing the Inferred Alignment Cintos 3/2/10 12:29 PM
Greetings fellow earth browsers:

Correlation of the Carolina bay "Inferred Alignments" continues to suggest that, should they prove to be distal ejecta, they might be from an impact in the Lower Peninsula of Michigan. Perhaps the most problematic aspect of our analysis thus far has been the empirically-derived 22 minute required loft time. That value is not supportable by common trajectory models, which for a 22 minute transit mandate a steep loft to a very high apogee The resulting near-vertical reentry would not generate the "spread" upon landing required to generate an oval bay shape. At a more realistic 30 - 45 º ejection, the loft time from two ballistic programs used are in the range of 400 to 600 seconds. We are now in a position to resolve that challenging aspect of the solution.

The "Systematic by Loft" model has been numerically extended by applying the "Systematic by Latitude" conjecture. While the Coriolis Force component is systematic by loft time, there is another factor superimposed on alignment that is a function of latitude. In the first case, the earth rotates at 0.25 degrees of longitude per second, regardless of where on the earth the consideration is applied. In our second case, we rationalize that the ground speed of any particular spot on the earth is a function of the cosine of its latitude. The end cases are the poles - where the ground velocity w>e due to rotation is negligible - and the equator - where the ground speed w>e is ~1,670 km per hour.

In our specific Saginaw Manifold example, a relevant set of w>e velocities would be our Sagianw Centroid point (43.58N, 83.94W) - rotating at 1,270 km/hr - and a generic ejecta field such as Bishopville - rotating at 1382 km/hr. At the time of contact, the 165 km/hr w>e velocity difference will resolved by a skewing of the "splat", effectively rotating the bearing in the counter-clockwise direction. To identify the true arrival bearing (which points back to the impact site), a de-skew model must be applied.

By way of explanation, a droplet of ejecta traveling from the north to the south in its great-circle frame of reference would not be affected by that ground speed difference until it approaches the surface of the earth, where the atmospheric breaking effect on terminal velocity would be applied, effectively allowing for a higher W>E landing velocity than supported by the loft geometry. That effect increases in magnitude as the landing sites move more southerly. At our most southerly field, Warner Robins, GA, the ground speed difference is 197 km/hr. Amplifying the latitude effect, the 197 km/hr delta is applied to a W>E loft velocity vector that is smaller relative the those of the more northerly fields, resulting in a higher percentage adjustment. The following graphic attempts to exhibit the numerical adjustments we make to the inferred alignment to arrive at a de-skewed bearing for each bay. The graphic uses the metrics of Wagram, NC's field, and the numerical model generates a new bearing of 143.9º compared to the measured field bearing of 135º. This is the bearing we then apply at the easterly offset field location suggested using "Systematic by Loft Time".

As the heuristic solution continued, we have assumed that a compute model could be developed that would iteratively resolve for a set of ejecta loft parameters from the Saginaw Impact area (ejection site along butterfly, loft azimuth, loft angle, loft velocity) that would satisfactorily correlate each of the Carolina bay structures with the empirically measured results. To further that goal, we have created a numerical model to generate the great-circle lines necessary to plot the west-to east offset caused by the Coriolis force (systematic by loft time) and the earth-surface rotational speed differences (systematic by latitude). By using the model, we solved for a Saginaw area loci and generated the following visualization in Google Earth.

The elements generated include a short green line representing the measured inferred bearing at each field, a pink line to the east representing the loft time offset, a surrogate field placemark (tgt) representing the "stationary earth" trajectory target of the ejecta loft, and a yellow line representing a great circle line back from that target along the de-skewed arrival bearing. KML elements for these visuals are available in the "Saginaw_Rotate_deSkewed_8_min" folder in the attached file.

Here is a graphic which attempts to represent the relationship between the atmospheric drag on the ejecta droplets (Cd) and the resulting terminal velocity and average trajectory ground-velocity for a series of Saginaw loft time solutions. This chart was generated by varying the proposed loft time from 4 to 24 minutes, while varying the droplet Cd to reach a solution for all bays such that their bearing paths triangulate within the Saginaw crater ramparts.

The chart suggests a realistic range of probable velocities and resulting loft time solutions. As the average velocity exceeded ~2 km.sec, the terminal velocities required for a Saginaw loci solution approached a 200 m/sec asymptote. Empirically, this would suggest that the ground vector component of the terminal velocity of the ejecta was ~200 meters/sec. We believe that the 200-300 meters/sec range for terminal velocities that were identified are highly correlated to the ~360 meter/second value seen in a separate terminal-velocity calculation for a 100-meter diameter droplet of ejecta. Note that only the ground vector of the velocity is being considered in the de-skewing calculations, which would be lower than the total 3-D terminal velocity, The value of ground component would depend trigonometrically on the vertical angle of incidence that the ejecta approached at; these calculations assume a 45º angle.

Google Earth path sets for each of the plotted points in the chart above were generated to visualize the loci focus. The charts below are considered likely 'average" solutions to the set and use a loft time of 6 and 8 minutes (360 & 480 seconds). During the actual proposed impact, ejecta density and average velocities would likely vary by field, and thus vary the de-skew factor.

The entire range of correlation charts for the solution sets are displayed on a page at our web site ( Solution Charts LINK ). The attached KML file can be used to visualize the solution we have discussed here.

Our next effort is a web-based java script so that allows for others to input a bay lat/lon value and generate kml to "predict" the bay's inferred alignment for comparison to it's actual geometry on the ground. I'll be refining that over the coming days. What would really be neat is to add a new "Add" item in Google Earth so it could be generated within GE.

We maintain that the high degree of correlation demonstrated in our analysis and charts across these 42 disparate locations across the continental U.S. provide strong support our hypothesis of an ejecta blanket event.
Predicting the Inferred Orientation of a bay Cintos 3/15/10 2:44 PM

The great physicist John von Neumann once disposed of a mathematical argument by saying, "With four adjustable parameters, I can fit an elephant; with five I can wiggle his trunk."

The mathematical model developed here is considered by us to be very simple and elegant. The only variables being perturbed are the average velocity of the ejecta during its flight (with loft time as the tuning proxy) and the terminal velocity of the falling droplet of ejecta ( with the pellet's coefficient of drag Cd as the tuning proxy). At each field of bays, we extract the latitude and longitude as test-case input constants. The model was heuristically focused on the Saginaw Bay proposed structure and it's three control points (NE, Centroid and SW); those latitude and longitude values are similarly applied as constants across all calculations. The 42 bay location "fields" represent many thousands of individual bays, and the solution sets are resolved against all fields simultaneously, with all predicted bearings falling within the control points. While it would seem plausible that the ejecta at any given location may actually have a density or velocity different from other locations, the model did not have to leverage such fine tuning to arrive at simultaneous solutions for all 42 fields.

The model in use can generate correlations directly against an empirically measured bearing which is then de-skewed back to the proposed Saginaw crater location (as in the last post), or by starting with a single point (lat, lon) and generating predicted bearing sets which we compare visually against a bay at that location (which is discussed here) . The series of tests using both methods also generated sensitivity tests for the solutions.

Here are two plots, at 8 minutes and 10 minutes loft time, of the model's predictions of bearings compared to the actually measured bearing.

The only variables are the loft time and the terminal velocity ( as a function of density and Cd). As shown in the figure below, the most acceptable solutions are found with a terminal velocity (in the ground plane) of 200 - 300 m/sec. We find this correlation to be highly supportive of the model's validity. Here is a sensitivity graph showing how the variables relate for the 42-field solution sets.

In the last post, a beta version of a "de-skew" calculator was offered, which only addresses the loft time. A version of that is still being developed that would allow a user to create a two-point path in Google Earth as they perceive the orientation, and then de-skew that input to create a path to the proposed crater.

Here, I'd like to offer another Java web-based calculator which attempts to generate inferred bearings at a user-chosen location. These "predicted" orientations can then be compared by the user to the actual visual orientations. The predictions are based on the systematic-by-loft and systematic-by-latitude computations we have discussed previously. The code is a bit easier than the de-skewing goal, as I only need to parse out a single point rather than a two-point path segment. I believe the forum visitors would benefit from having both capabilities.

The process of using the calculator < Linked HERE > is straight forward for the Google Earth user. A more detailed help page < Linked HERE > is available on line, but a few shots are shown here.

Create a placemark { Add Placemark } within a bay or field of bays. Add a useful name to the placemark. Use the copy function to place the placemark's kml on the clipboard.

Next, access the calculator and paste the copied kml data into the input window (be sure to clear out any old entries there). Click the "generate" button. Click into the result window at the bottom, and { select all } & { copy } to place the generated kml on the clipboard.

Finally, go back to Google Earth and paste in the kml.

The defaults for a solution set are pre-set in the calculator window; they should generate a set of bearings that bracket the site's inferred orientation. It would be neat if I could figure out how to integrate the generator directly into GE, but for now, copy/paste will need to be leveraged.

wave UPDATE!! The calculator has been updated significantly, and will be the subject of a new post soon. IT should be usable now in Windows environments. In the meantime, have a look at the Prediction Calculator .

ocool ocool
Carolina bay Bearing Calculator Cintos 3/19/10 8:39 PM

In out last post we introduced an early version of a java web-based calculator to generate Google Earth kml elements for visualization of our numerical model. The code has improved enough to re-introduce it here, along with a discussion of its new features. The earlier version had me challenges running on many windows installations; we hope this one functions better.

To allow the user to interact with the model, a java web-based application was written. The "Bearing Calculator" can be driven with two different sets of data, resulting in two different types of Google Earth visualizations.
By creating a simple placemark in Google Earth, the calculator can generate predicted bearing sets which you can compare visually against a bay at that location. The calculator can generate correlations directly against a user's (that's you) measured bearing (by starting with a "Bearing Arrow (overlay) you create in the placemark case), which is then programmatically de-skewed back to the triangulation points. The series of tests using both methods also generated sensitivity tests for the solutions.

In the first case ( POINT ) the user creates a single placemark in Google Earth near a bay. That placemark is copied into the calculator and a set of elements is generated for use back inside of Google Earth. The elements identify a "Predicted" set of orientations for the site: one path shows the model's prediction for a path to the oval Saginaw craters' centroid, and two other paths predict the alignment if the ejecta had been lofted from the crater at the northerly and southerly ramparts. Positive correlations is found when the bay's alignment falls within the predicted set. (Link to the Point HELP page)

In the second case ( Arrow ) processes the "Bearing Arrow" (generated above) in Google Earth, after it has been aligned by the user along their best-guess inferred alignment. That overlay kml is copied into the calculator and a set of elements is generated for use back inside of Google Earth. The elements identify a set of orientations for the site: one path shows the user's input path, another path visualizes the loft time offset due east as a function of loft time, and the final path flows back along the model's de-skewed bearing. Positive correlation comes from a series of triangulation's between eastern and western fields which fall within the suggested oval Saginaw crater. (Link to the Arrow HELP page)

The calculator defaults to an 3 km/sec average velocity and a Cd of 0.3. These values can be changed, along with the density of the ejecta and the angle of incidence of the fall. The calculator determines which case is being requested by reviewing the user's input kml.

Please reference the two help pages and the calculator ( Linked HERE ), located at the web site.
Re: Inferred Orientation of Distal Ejecta Cintos 3/22/10 11:19 AM

The calculator at Version 2.2 Version 3.0 has been enhanced with a variation of the trigonometric approach to de-Skewing or re-Skewing based on the terminal velocity. I encourage the reader to try your hand at comparing the calculator's predicted bay alignments against what you perceive to be a correct orientation (create a placemark in the bay and copy/paste that placemark into the calculator), or for you to create a path which is aligned with the orientation you personally perceive, and copy/paste that into the calculator to see where it predicts the bay "came from". (Link to the HELP pages )

The earlier code used arcTangent to deduce the adjusted bearing, using the terminal velocity's n>s component along with the new w>e component (after adding or subtracting the w>e lattitude-driven differential ground velocity vector). That process did not yield reciprocal results for the two cases.

What I overlooked was that, since I am lengthening or shortening the w>e component, and the terminal velocity is a constant for the case, that the n>s component must change for the triangle to remain right. So it is inappropriate to rely on the n>s component in the reconstruction unless that change is accounted for. It is far easier to just simply use the the terminal velocity (hypotenuse) along with the w>e component (adjacent) and use arcCosine to recover the angle.

The numerical model now generates predicted bay bearings and de-skews empirically measured bearings such that they now have an even higher correlation to the proposed Saginaw crater site. Please note that the solved solution sets' Cd (and the calculator's defaults) have are now changed slightly (.3 vs .53 @ 8 min).

The graphs for the two cases, simultaneously-solved across the current 43 sites, are shown in the graphics below. In addition, the kml results for each case is included in the attached keyhole file, for both an 8 minute loft time and a 10 minute loft time.
More Carolina bay "Fields" Documented Cintos 3/30/10 4:15 PM

Much time has been spent scouring the continent for new fields of Carolina bays. The process is quite rewarding on the Eastern Seaboard, and the Ejecta Field count is up to 55, 48 of them in the east. These likely represent 10,000 individual bays. The attached file contains the index placemarks to the 17 latest Eastern fields. As before, the index placemark needs to be selected in Google Earth. In this example, we chose the Larson Mills, NC field:

Clicking on the "KMZ" icon will retrieve a set of kml elements that include a color-ramp DEM image overlay. These really bring out the bay details!

The triangulation to Saginaw is dependent on the Western fields, but at the present we have only documented the Nebraska locations (initially identified by George Howard) which exist across a relatively small area of the proposed ejecta arc. The location is actually quite appropriate, as it is one of the few locations in the west that offer a relatively flat hosting terrain. Complicating the search is the understanding that much of the proposed arc was actually covered by glacial ice at the time, and while those areas across North Dakota, South Dakota, and Minnesota contain thousands of lakes, they are very random in shape and offer no orientation beyond that driven by river channels.

Here is a low resolution elevation map of the western areas. The current fields are located within the circle in the middle, and the relative flatness of the area is evident. The other two circles identify areas in Kansas we are continuing to explore for aligned lakes. The smooth flow-like areas to the north describe the glacial lobe paths.

I find it quite stunning that these states, considered to be "flat and boring" have a considerable amount of rugged terrain, which shows up in the higher resolution NED data I have been using in the search (courtesy of the USGS Seamless Server). An occasional oval feature will show up in Google Earth imagery, but the NED data will then show it to be at the head of a ravine, or perhaps an oxbow on a river. Identification of a true "field" of bays is required, and when they are there, it is quite obvious. Areas such as Heron Lake, Storm Lake and Worthington with a couple of oval lakes are dismissed as unsatisfactory.

Just for a taste, here are three new field portraits:

Cliffdale, NC

St James, NC

and Bonnetsville, NC
New Bearing Calculator Feature Cintos 4/2/10 9:36 PM

A new feature has been added to the Bearing Calculator. In addition to a single placemark (to generate predicted arrival bearings), the calculator can now ingest a "Bearing Arrow" overlay. The Google Earth overlay kml element includes a "rotation" metric, such that when the user positions the bearing arrow overlay so that it aligns well with the local Carolina bay, that information can be passed to the calculator. The overlay also contains a large area of thin graticule lines to verify alignment with numerous bays at once.

This is what the bearing arrow looks like:

The url for the overlay is

A help page has been posted on the P:Z web site to discuss the use in more detail: Link To Bearing Arrow Help .

As a quick start, the graphic below shows a new overlay created at the Wagram bay field kmz .

By selecting the overlay, then doing a "get info", the arrow can be stretched and rotated to suit the user's view of the bay field's orientation.

The overlay element is then copied from Google Earth and pasted into the Bearing Calculator's input window.

When the calculator's output is pasted back into Google Earth, a set of elements are created which "de-Skewes" the user's perceived bearing and traces back to the Saginaw area. The graphic here shows the full working set of about 65 Carolina bay fields as found in the attached file.

Thanks for ploughing through all this! There may not be a crater in Saginaw, but Google Earth is a whole lot of fun to explore in!!! Next up is a discussion of the Nebraska Sand Fields, and all the oriented bays across the area. Guess where their inferred orientation points?

UPDATE! The Bearing Calculator was enhanced beginning with V 2.4 to automatically include a Bearing Arrow as an element when the POINT case is used. Create a new placemark at the center of a bay, paste that into the calculator, and the Bearing Arrow will be provided oriented according to the predicted arrival bearing for that geographic location. You will still need to edit the arrow, as above, for fine tuning to the local bay orientation, then copy/paste the arrow back into the calculator to provide the de-skewed trajectory paths.
More Nebraska bays Cintos 4/12/10 10:29 AM

Our count of "fields" of Carolina bay type landforms in Nebraska has increased to 12. While there is a great deal of research covering the "Carolina bays" on the eastern seaboard of the US, little attention has been paid to the significant quantity of oval-shaped landforms in the eastern areas of Nebraska. These, too, are aligned with each other. Significantly, the inferred alignment is considerably different than that of the orientation of the extensive sand dunes in the area.

While not nearly as extensive as those in the east, the identification of these bays to be critical to the impact site triangulation process, as contrasted with their eastern-US brethren.

The graphic here shows the extent of identified Nebraska bays, and how the Nebraskan bays we have identified lie along a ring around the Saginaw impact site. They are correlated to a high degree using the Bearing Calculator tool. The attached kmz file contains placemarks and kml elements to reproduce this graphic in Google Earth.

The use of USGS-provided digital elevation maps (DEM) of the area has allowed for this identification, as the characteristic shape and orientations are rarely seen or identified on the ground or in satellite imagery through Google Earth. Each field placemark in the attached kmz file includes links to DEM images as Google Earth overlays, which you are encouraged to load and view.

For quick comparison, here are close-ups of two individual bays, showing the normal Goggle Earth Imagery, and then the DEM overlay. A web page is available which shows the same pairings for all 12 areas:The Nebraska bays

The bays are not really very deep, of course. The color ramp DEM images are run with the elevation exaggerated. The last graphic here shows an elevation profile across one of the Garfield, NE bays.

We interpret the placement of the bays as being indicative of them being "pedestals" landforms, which owe their existence to the bowl-shaped interior's capability of retaining moisture. Over the millennia since their emplacement, the majority of the ejecta blanket has been subjected to wind and water erosion, whereas these have been stabilized. The analogy is that of tire tracks in snow, outlasting the surrounding snow, rising above the road surface. A similar process is implicated in the pedestal craters on Mars. Numerous landforms in nearby areas suggest they were at one time 'bays", but have been compromised by encroaching erosion. Such a fate is likely in the future for the Garfield Township bay shown above, which is beginning to be invaded by a stream.

Best wishes,
Exceptional Maryland bays Cintos 4/20/10 12:30 PM

The Inferred Alignment Prediction calculator has been updated to V 2.7, which fixes a small trigonometric error encountered when handling western fields. The correlation of predicted bearings vs the inferred orientations of all identified bays continues to be close to 1.

There is a significant exception to our success: bays in the northwestern corner of the DelMarva Peninsula. There, just east of Washington, DC, lies an extensive field of Carolina bays - probably numbering in the tens of thousands. These bays are almost invisible in visual imager (like Google Earth), but are simply stunning when visualized in high-resolution LiDAR color-ramp imagery.

Bays exist in staggering numbers all across the Delmarva Peninsula, and the Distal Ejecta Fields kml file now includes placemarks and kml support for 15 Fields there. As the fields are traversed south to north, the planform of the bays becomes more and more rounded, yet they can be correlated well with the calculator's numerically predicted arrival bearings. Needless to say, a "round" bay can not suggest an inferred arrival orientation, and thus two fields at the very top of the peninsula suggest an orientation that is off from our predictions by about 15º in the clockwise direction. Here are three graphics, created from 1/9 arc second USGS NED data using Global Mapper. The last of theses shows the elevation profile across a 4 mile path through these bays. Given the small vertical relief, it is no wonder they are virtually invisible to the eye.

If I ever get "into the field" for some ground work, the road cuts in this area (near Price, MD) would be a great place to start! I am placing a network linked overlay file for this last graphic as an attachment to this post.
Cardinal Points for New Jersey Cintos 4/23/10 2:07 PM

If you recall, the numerical model embedded in the Inferred Alignment Calculator uses trigonometry to de-construct and then re-construct the trajectories based on the loft transit time of the ejecta, and its terminal velocity as it re-enters the lower atmosphere. These trig formulas are actually different in each of the four quadrants of the compass cardinal points. To make matters worse, the trig functions are not well behaved when crossing the 0º, 90º, 180º and 270º. Up to now the relationship of the bays and the Saginaw site are not close to any of these, but the New Jersey sites are getting close to 90º east of the crater.

Our earlier solution sets, using bays in North Carolina and southward, were all approximately the same radial distance from the Saginaw focus. That allowed us to use a "loft time" parameter in the calculations. The Maryland, Delaware and New Jersey fields are breaking that model, as they are increasingly closer to the proposed impact site. Here is the graph of distances, walking clockwise around the Saginaw focus.

To address this, the calculator ( Version 2.9 ) now uses a user-adjustable parameter "average velocity" for the transit from crater to the eventual ejecta depositions site. From that value, we are now calculating the loft time for each location. Thus the loft time used for NJ ( 850 km) will now be less than for those in Nebraska (1200 km), which seems to be a reasonable model enhancement. The result is an even higher degree of correlation between all the bay's inferred alignment (as empirically measured) and the calculator's predictions.

As mentioned in the prior post, many of the bays of Maryland, and now for New Jersey, have lost the elongation which aided in deducing an inferred orientation. In fact, it has long been reported by others that bays in the far south and those in the north trend towards a round shape. What is left, significantly, is a predisposition for having a segment of the enclosing ring be fatter & higher than the opposing side. If we continue to deduce the alignment to be from the shallow side to the fat "lip", then we may be able to continue with the process.

Here are three examples of what look to be fields of "Squashed" bays, where the momentum during emplacement was more downward than laterally (higher loft angle?).

The count of bay fields is now at 100. Here is the mashup from our database-generated de-skewed alignments & predicted orientations. The associated kml elements are in the attached file.
Upland Bay Search Cintos 4/30/10 11:02 AM
Finding Carolina bays in the Carolinas is easy, given their plentiful quantities and solid identification, but as the ejecta butterfly arc is walked northerly and southerly, the search becomes more challenging. Some of this is due to the more rounded shapes seen above Maryland and down in Georgia. An additional challenge is the increasingly rough terrain seen when moving inland from the costal plains. Our ejecta blanket sheet model suggests that the bays are present as defects (popped bubbles) in a thin layer of sandy ejecta. These can persist over the thousands of years of erosion only under special circumstances. If the landing area is relatively flat and moist, they will be easily stabilized as bays. If the area is level but very dry, the blanket will be reworked by the wind into a generic dune field, obliterating any bay formation.

When the landing field is in rough terrain, we propose it is quickly sloughed off in erosion. Any bay formations that are draped over elevations that exceed the bay's rim hight can not hold moisture for stabilization. Thus in hilly areas further inland on the east coast, we have pursued the search for bays by identifying areas which exhibit level terrain. These can usually seen as plateaus in the digital elevation maps (DEMs) we are using. The plateaus could be indications of a plateau extent present when the ejecta landed, or be merely the surviving remnants of a larger plateau that has been invaded by erosional valleys. In the latter case, we see that bays which once were stabilized were later compromised by encroaching erosional valleys.

Here is an example of the search methodology, as applied to interior Georgia. The USGS geographic information VIEWER facility is first used to retrieve elevation data for an area.The retrieved data files are loaded into the Global Mapper GIS tool for display in a Color-Ramp image. The full "Thompson" 100K block, with three circles highlighting areas that look to be flat enough to suport bay formation and survival is shown below.

Zooming in to higher resolution view reveals the presence of oval landform shapes in the smooth surface areas. Another level of zoom, below, brings out more of these details. The Global Mapper elevation profiling tool is used to identify the terrain and validate the bowl-shaped nature of a Carolina bay possibility.

Global Mapper can export the color-ramp DEM image as a set of coordinate indexed Google Earth kmz layers. These are imported into Google Earth, where they are used to enhance the normal visual imagery. Once in place, each depression can be evaluated against the visual imagery of the location, to verify if a bay planform is present. If a correlation is found, the site can be tested with the Inferred Orientation Calculator, which will create a set of reference kml for Google Earth display and further correlation. You may note from the image above that several full-rimed ovals are present, and all exhibit roughly the same NNW to SSE orientation. Also apparent are indications of man-made drainage ditches cut into the center of several to drain them. Using Goolge Eart's imagery, and in this case, historic black and white imagery (1999), the array of aligned bay landforms is readily apparent.

The area is identified in Google Earth as being near the town of Wrens, GA. A folder of kml data is assembled for distribution as the "Wrens_GA" field. Similarly, bay suspects were located in each of the other "circled" areas in the Thompson 100K block, and a set of kml developed for each of the other two sites as well: Harts_GA and Hephzibah_GA. These three folders of kml are included in the attached kml file covering 30 fields of bays currently in our Distal Ejecta Fields index kmz file for Georgia. A similar process is used in identification of the state of Nebraska

This post is a subset of the presentation made on our web site on the Searching For Bays page.
Java Code Snippets Cintos 5/28/10 7:32 PM
For starters, here is the latest Saginaw Manifold "de-skewed bearing" Portrait. The file now lists over 140 fields of bays. The attached kml file will recreate this mash-up in your own Goggle Earth instance, including our model's bearing predictions at each bay field and the "walk back" corresponding to the bay's measured inferred alignment.

Hopefully some readers have used the Bearing Calculator with some measure of success. Our motivation for providing the calculator as a web-based tool is to encourage testing of the relevance of the model by those interested in our conjecture. In addition to any Carolina bay location information you may already have access to, we refer the reader to two sources:

* Our extensive catalogue of Carolian Bay "Fields" - Distal Ejecta Fields
* An independent list of individual Carolina bays - Comprehensive SouthEastern US Crater catalog by Thomas Flores

The calculator's numerical model was truly arrived at heuristically during the production of this thread, driven by a forensic analysis of the evidence (a given bay's momentum-generated orientation) and a sense of the potential dynamic and geophysical forces at work. The variables were adjusted until all the evaluated bays' measured alignments were within the calculator's prediction values. The goal was to craft an algorithm that would predict a bay's momentum-driven alignment at any given point along the ejecta ring, which is a significantly different than addressing the simple ballistic trajectory.

I intend on documenting the java code in my model to help explain the algorithm. Allow me to use this opportunity to try some presentation methodologies for the code as it relates to the kml production. Perhaps it can be of use to others in their own kml generation programs.

This is the subroutine to calculate the great circle distance between lat1, lon1 and lat2, lon2: (convdr and convrd convert between radians and degrees)

 private double GreatCircleDistance(double lat1, double lon1
     , double lat2, double lon2) {

  double dLat = (lat2 - lat1);
  double dLon = (lon2 - lon1);
  double a = Math.sin(dLat / 2) * Math.sin(dLat / 2) + Math.cos(lat1)
     * Math.cos(lat2) * Math.sin(dLon / 2) * Math.sin(dLon / 2);
  return (earthRadius * 2 * Math.atan2(Math.sqrt(a), Math.sqrt(1 - a)));  }

Here is the code to calculate the initial bearing from lat1, lon1 towards a point lat2, lon2:

private double GreatCircleBearing(double lat1, double lon1
     , double lat2, double lon2) {
  double dLon = (lon2 - lon1);
  double y = Math.sin(dLon) * Math.cos(lat2);
  double x = Math.cos(lat1) * Math.sin(lat2) - Math.sin(lat1) 
     * Math.cos(lat2) * Math.cos(dLon);
  double Bearing = 180 + (Math.atan2(y, x) * convrd);
  return (Bearing * convdr);

Here is a set of code to create a Google Earth path "linestring" kml element from the forepoint out a distance km, following the initial bearing degrees.

private String GEpathFromBearing(double foreLat, double foreLon,
   double bearing, double distance) {
  final String startKML = "<LineString> <tessellate>1</tessellate> <coordinates>";
  final String endKML = "</coordinates> </LineString>";

  double farLat = Math.asin(Math.sin(foreLat) * Math.cos(distance 
     / earthRadius) + Math.cos(foreLat), Math.sin(distance 
     / earthRadius) * Math.cos(bearing));
  double farLon = foreLon + Math.atan2(Math.sin(bearing)
      * Math.sin(distance / earthRadius) * Math.cos(foreLat)
     , Math.cos(distance / earthRadius) - Math.sin(foreLat) 
     * Math.sin(farLat));

  String foreCoord = foreLon*convrd +  "," + foreLat*convrd +  ",0 ";
  String farCoord = farLon*convrd + "," + farLat*convrd + ",0 ";

  return (startKML + foreCoord + farCoord + endKML);

Rather pedestrian stuff. The next one is more Google Earth kml "exotic" . Returns a sting of kml to place a copy of our Bearing Arrow overlay at a bay site given lat, lon and rotation, with inputs in radians. In the caclulator, this places the bearing arrow properly rotated to reflect the predicted arrival bearing at the bay location. The math identifies the the required points for the overlay location as a 4 kilometer diagonal square using the provided placemark as the center point, by going 2 km out 45º (NE) and 225º (SW).

public String bearingArrowKML(double lat , Double lon , Double rotationValue ){
  double distance = 2 ; /* km for building latlon box */
  double NELat = Math.asin(Math.sin(lat)
     * Math.cos(distance / earthRadius) + Math.cos(lat)
     * Math.sin(distance / earthRadius) * Math.cos(45 * convdr));
  double NELon = lon + Math.atan2(Math.sin(45 * convdr)
     * Math.sin(distance / earthRadius) * Math.cos(lat),
       Math.cos(distance / earthRadius) - Math.sin(lat)
     * Math.sin(NELat));
  double SWLat = Math.asin(Math.sin(lat)
     * Math.cos(distance / earthRadius) + Math.cos(lat)
     * Math.sin(distance / earthRadius) * Math.cos(225 * convdr));
  double SWLon = lon + Math.atan2(Math.sin(225 * convdr)
     * Math.sin(distance / earthRadius) * Math.cos(lat),
      Math.cos(distance / earthRadius) - Math.sin(lat)
     * Math.sin(SWLat));
  Double rotationValueBox = (180 -rotationValue*convrd) % 360;
  final String kmlA = "<GroundOverlay><name>";
     /* elementName */
  final String kmlB = "</name><description><![CDATA[Bearing Arrow Overlay <br>" +
     "<a href= \"\"> " +
     "Bearing Calculator V 2.9 </a> <br> © Cintos 2010 ]]></description>" +
     "<drawOrder>6</drawOrder><Icon>" +
     "<href></href>" +
     " <viewBoundScale>0.75</viewBoundScale></Icon><LatLonBox><north>";
  /* north lat */
  final String kmlC = "</north><south>";
  /* south lat */
  final String kmlD = "</south><east>";
  /* east lon */
  final String kmlE = "</east><west>";
  /* west lon */
  final String kmlF = "</west><rotation>";
  /* rotation value */
  final String kmlG = "</rotation></LatLonBox></GroundOverlay>";

  return (kmlA + elementName +  kmlB + NELat*convrd + kmlC 
     + SWLat * convrd +  kmlD + NELon * convrd + kmlE + SWLon 
     * convrd + kmlF + rotationValueBox + kmlG);
Re: Java Code Snippets JavaGAR 6/1/10 6:34 PM

Thanks for the Java code. It looks really useful, not only for your project, but for other projects involving geographic calculations as well.

A few questions on how to use the code follow:

Should the parameters be passed to the GreatCircleDistance method in radians? That seems to be the case because Java's Math methods are written to operate directly on angles given in radians.

It appears that convrd, convdr, and earthRadius are variables that are declared and assigned their values outside the methods that use them. Is the following what needs to be included in code in order to give them their necessary values?
double convrd = 180.0 / Math.PI;
double convdr = Math.PI / 180.0;
double earthRadius = 6371.0; // in kilometers, or 3959.0 miles

It appears that those variables could be declared, either as constants inside the methods that use them in order to make those methods self-contained - or alternatively the methods can all be placed together in a class, along with those variables, in which case the methods could each access the variables globally.

It also appears that the parameters to the GEpathFromBearing and bearingArrowKML methods should be passed in radians. Is this correct?

This is a really interesting project, and it is great that you are sharing some of the useful code with us.
Best Regards,

Re: Java Code Snippets Cintos 6/2/10 3:52 PM
Greetings JavaGAR:

Thanks for the kind words. As that was my first foray into presenting code on the BBS, I was motivated to try some presentation approaches. The only thing I could find that seemed to work was the "CODE" UBBCode tag. Any comments or suggestions you may have there would be appreciated. Also ! full disclosure ! this was my first trip into Java, being an old FORTRAN guy from the 60's. Take my approach with a grain of salt.

As you surmise, the MATH variables require radians as inputs. Your code for the conversions and earth radius are correct. The subroutines are instantiated within my main, and those are in scope throughout.

I'd like to take the opportunity to add a few more snippets. The Bearing Calculator can ingest two elements, either a Placemark (point), or the arrow overlay. While there is java code out there that handles all manner of tags, I kept it simple and just coded what I needed.

A separate "latLon" class was created to carry the metadata for a location, and did set some default values. I have only created a few children, but eventually I may expand the code to create an array of them as the program is used. When the values Lat & Lon are set using degrees , the routine creates the radian versions lat & lon. (note my hinting: upper case = degrees, lower = radians)

 * Identifies a Google Earth Placemark object on the earth's 
 * surface at the supplied latitude / longitude; altitude= 0
public class latLon {

 static final double PI = 3.141592653589793;
 static final double convdr = PI / 180., convrd = 180 / PI;

 double Lat = 45.0;
 double Lon = -80.0;
 Double lat = Lat * convdr;
 Double lon = Lon * convdr;
 String LonLatGE = Lon + ", " + Lat + ", 0 ";

 void latLonSet (Double latNew, Double lonNew){
      Lat = latNew;
      Lon = lonNew;
      lat = Lat * convdr;
      lon = Lon * convdr;
      LonLatGE = Lon + ", " + Lat + ", 0 ";

A "bay" location is instantiated, along with the Coriolis target location and a pair used in the arrow overlay box kml in the previous post.

latLon bayLoc = new latLon ();
 latLon tgtLoc = new latLon ();
 latLon arrowBoxNE = new latLon ();
 latLon arrowBoxSW = new latLon ();

To parse the placemark or an arrow overlay, I first walk through the input kml text for the for the existence of a either point or arrow,

static  String point = "<Point>";
static  String lineSting = "<LineString>";
static  String comaDelim = ",";
static  String coOrdinates = "<coordinates>";
static  String placeName = "<Placemark>";
static  String latLonBox = "<LatLonBox>";
static  String nameDelim = "</";
static  String nameFlag = "<name>";
static  String northFlag = "<north>";
static  String southFlag = "<south>";
static  String eastFlag = "<east>";
static  String westFlag = "<west>";
static  String rotationFlag = "<rotation>";

userPoint = false;
userArrow = false;
linePos = inKML.indexOf(point)  ;
if (linePos != -1) {   /// do user point option
      userPoint = true;
      int placemarkPt = inKML.indexOf(placeName );
      int namePtr = inKML.indexOf(nameFlag, placemarkPt );
      namePtr = namePtr + 6;
      int nameEnd = inKML.indexOf(nameDelim, namePtr);
      elementName = inKML.substring(namePtr, nameEnd);
      pointCaseCoordinates (linePos);

}  else {
linePos = inKML.indexOf(latLonBox)  ;
if (linePos != -1) {   /// do user point option
      userArrow = true;
      int namePtr = inKML.indexOf(nameFlag );
      namePtr = namePtr + 6;
      int nameEnd = inKML.indexOf(nameDelim, namePtr);
      elementName = inKML.substring(namePtr, nameEnd);
      arrowCaseCoordinates (linePos);
} else  {
            elementName = " No Name";

I check for which type was pasted in by the user (point or arrow). In the point case I then pass the start position to a routine to pull the lat & lon out, placing those values into the bayLoc latLon container using latLonSet.

 private void pointCaseCoordinates (int linePos) {
 Double lat1, lon1;
 int cordPosition = inKML.indexOf( coOrdinates , linePos)  ;

 int lon1Start = cordPosition + 13;
 int comaPosition = inKML.indexOf(comaDelim, lon1Start) ;
 lon1 = Double.valueOf(inKML.substring(lon1Start, comaPosition) );

 int lat1Start = comaPosition + 1;
 comaPosition = inKML.indexOf(comaDelim, lat1Start) ;
 lat1 = Double.valueOf(inKML.substring(lat1Start, comaPosition) );

 bayLoc.latLonSet(lat1.doubleValue(), lon1.doubleValue() );

After setting, I can simply read values as radians or degrees as fits the need.

Here is the code to extract meta data from the arrow overlay kml in the arrow case. Most important is the "rotation" data Google carries along for the overlay. I created the png file for the overlay with the arrow pointing straight up, and Google Earth tracks the angle as the overlay is rotated. Neat... :

private void arrowCaseCoordinates (int linePos) {
 Double LatNE, LonNE, LatSW, LonSW, rotation;
 int coordPosition, coordEnd;

 coordPosition = inKML.indexOf( northFlag , linePos)  ;
 coordEnd = inKML.indexOf(nameDelim, coordPosition);
 LatNE = Double.valueOf(inKML.substring(coordPosition + 7, coordEnd) );

 coordPosition = inKML.indexOf( southFlag , coordEnd)  ;
 coordEnd = inKML.indexOf(nameDelim, coordPosition);
 LatSW = Double.valueOf(inKML.substring(coordPosition + 7, coordEnd ) );

 coordPosition = inKML.indexOf( eastFlag , linePos)  ;
 coordEnd = inKML.indexOf(nameDelim, coordPosition);
 LonNE = Double.valueOf(inKML.substring(coordPosition + 6, coordEnd ) );

 coordPosition = inKML.indexOf( westFlag , coordEnd)  ;
 coordEnd = inKML.indexOf(nameDelim, coordPosition);
 LonSW = Double.valueOf(inKML.substring(coordPosition + 6, coordEnd ) );

 arrowBoxNE.latLonSet( LatNE.doubleValue(), LonNE.doubleValue()  );
 arrowBoxSW.latLonSet( LatSW.doubleValue(), LonSW.doubleValue()  );

 bayLoc.latLonSet((arrowBoxNE.Lat + arrowBoxSW.Lat)/2, (arrowBoxNE.Lon + arrowBoxSW.Lon)/2);
 //rotation : note that provided bearing is counterclockwise ... 
 coordPosition = inKML.indexOf( rotationFlag , coordEnd)  ;
 coordEnd = inKML.indexOf(nameDelim, coordPosition);

 rotation = Double.valueOf(inKML.substring(coordPosition + 10, coordEnd ) );
 BayBearingFromArrow = (360.0 - rotation.doubleValue()) % 360 ;

Looking at my full code base, you'd probably note that a lot of time is spent going from degrees to radians and back, as required for the kml exchange back and forth with Google Earth. You would also note that trig functions expect radial values from -180 to +180, while bearings are done from 0 to 360. ugh.

Another great hardship is that while I can decompose a bearing angle of 135º easily enough with sin and cos, if I try to recompose with same results using acos, you won't get 135.... When I read a bearing, I immediately divide by 90 to yield a "quadrant" (0,1,2,3) to be used to recompose a bearing after adjusting it.

- Michael
Men occasionally stumble over the truth ... but most of them pick themselves up and hurry off as if nothing had happened.
...... Winston Churchill
Comparing Mars Crater to Saginaw Cintos 6/6/10 10:20 AM

OK, I don't know the full scientific relevance of this discussion, all I know is that the comparisons here are kinda spooky. Early in my oblique impact research, I came across an example crater on Mars which presented the oval shape and butterfly distribution of local ejecta. The overlay has been available in our published kml for some time, shown in the Research_Overlays.kmz file, for example.

This Martian crater departs from a pure oval shape along some of its rim, and perhaps by pure coincidence, so does the topography of the Saginaw Bay the same areas! Here are 6 images showing the Google Earth view of Saginaw, with an overlay showing a color ramp elevation DEM, along with the Mars crater overlay (adjusted for orientation and size to match), while changing the transparency of the latter.

Along with the correlation along the rim, the general land mass of northern lower peninsula seems to bear the traces of the ejecta spray. Lake Huron, of course is excised by the glacial flow known to have passed thought that area. Note how the Mars flow stops "at" the northern Lake Erie shoreline...
Certainly, just a coincidence.

Another comparison can be seen in the Lake Huron bathymetry overlay (courtesy NOAA), when compared with the Mars overlay. One characteristic of shallow, oblique impact craters is that the deepest excavation is right at the uprange opening of the crater. Here, that aligns with the Bay City Basin.

Once excavated into a 2km-thick sheet of ice, we expect the "crater" to have filled with water and eventually drained south-west catastrophically as the Kankakee Torrent and CKRV episodes. Glacial incursions across the crater floor from the Huron Lobe would explain the current glacial moraines along the Bay shoreline seen today.

The attached kmz file includes the three overlays and the azimuth line, so the viewer can open these in Google Earth, and using the Get_Info/Edit > Transparency slider, adjust the overlays to see the correlations shown above.
New High Resolution LiDAR from Nebraska Cintos 6/15/10 11:29 AM

The fine folks at the Nebraska Department of Natural Resources (DNR) have just released a treasure trove of high resolution LiDAR data grids for public consumption . The coverage area is south of the North Platte, and covers about 50% of the bays we have examined using USGS 1/3 arc second National Elevation Database files.

The volume of data is enormous, given its 2x2 meter grid with elevation resolution of tens of cm. This gives us to opportunity to identify smaller basins and to see the larger bays in finer detail. I'd like to share a few early "discoveries".

Sand dunes and ridges are commonly found across the Nebraska landscape, and we see these dunes and ridges as over-printing the structural basins. We speculate that the bay depressions were created upon ejecta deposition sometime prior to 25kya. Since that time a significant blanket of late Wisconsin glacial loess have been deposited, rounding off the sharp edges of the bay rims. In some ares local dunes have breached the landforms, but the underling structure continues to show through.

An area near Edgar, NE, shows the dune activity. The bays are oriented at right angles to the prevailing winds, and the dunes and ridges are slowly but surely migrating into the image from the upper left (north is @ top)

We speculate that much of the sand in the Nebraska Sand Hills was originally deposited as distal ejecta during the Saginaw Manifold, but has been compromised by this activity over the past 25 thousand years.

We view this next local with great interest. The Nebraskan bays have been overlain by many meters of late Wisconsin loess, rounding off the edge of their rims. In the Campbell area, there are two bays that have been eroded at one end by a stream, which removed the loess and exposes some of the original rim. This site would be an excellent candidate for additional ground research.

DIrectly south is another, and in both cases the lower left end of the underlying basin structure is visible.

KML to visualize these two areas from within Google Earth are in the attacked kmz file.

More on the Nebraska bays can be seen on our web site.
Re: Inferred Orientation of Distal Ejecta Techlady 7/8/10 7:59 AM
When was it is established that Saginaw Bay is an impact site? You don't cite any sources for that.
An important test for an impact structure is the presence of microscopic shocked crystals. The presence of these can be confirmed only with field work. No amount of staring at Google Earth can confirm an impact structure.
It is my understanding that Michigan is largely a structural basin, a rough circle that sags in the center. Saginaw Bay is a fold created by the sagging. (I recall this from geology courses, so, if it's not current theory, someone correct me.)
At any rate, please let us know how you support the idea that Saginaw Bay is an impact structure.
Saginaw Hypothesis Cintos 7/9/10 9:25 AM
Greetings TechLady:

You are quite correct in your observation that much work needs to be done before any hypothesis offered by me could be considered proven. But I must suggest that the premise of this thread was not to attempt proof of the entire Saginaw Impact Manifold , but rather to share with the GE community the LiDAR images of Carolina bays and the development of the Bearing Calculator that so handily seems to predict their alignments for visualization in GE.

But, since you raise an appropriate objection, I'd be happy to provide some rationale here. From the outset (i.e., at the top of the thread) we had not yet triangulated to the Saginaw Area. At the AGU Fall Meeting, our poster tentatively considered the Lake Michigan area as a source for the 100,000 cubic kilometers of ejecta (much of it pulverized sandstone) that would be spread as a blanket across the US.

As the thread progresses, a diligent reader would note that we demoted that lake and implicated the Saginaw area, based on our triangulation results.

As you correctly note, the “Michigan Basin” is hundreds of millions of years old, and certainly not considered by us as an impact crater. It covers an area >10 times that of our crater’s extent. While the USGS does not have a firm explanation of the genesis of the subsidence, thousands of well-logging strips have mapped the continual subsidence of the basin accompanied by persistent in filling by sedimentary deposits. This results in a layer cake of history. The on-going subsidence at the center of the basin has also resulted in a sort-of stack of bowls (one bowl per stratigraphic layer), and the protruding edge of those bowls has generated a set of rings around the basin, the best known of which is the Niagara Escarpment.

The existence of Lake Michigan's, Lake Huron’s and Lake Erie’s "wrap” around the basin is considered the result of glacial exhumation of older, softer strata on the periphery of the basin, while also following along individual bowl-rim edges (known as cuestas), which present more resistant strata in the sequence. Classical geology has no firm solution for why the "Saginaw Lobe" of the Wisconsinian ice sheet violated those cuestas and penetrated into the central basin, and then continued to excavate uphill along a NE-SE tract. The overall elevation profile also agrees with our proposed 221º arrival vector. I am aware of only one scholarly attempt at solving the riddle of "How Michigan Got Its Thumb". Like the Carolina bays, most scientists avoid such enigmas.

The Saginaw Section of our website highlights numerous anomalies seen across our proposed "crater", and several maps, as overlays, are available in our kml file Research Overlays, which includes the graphic shown above.

So that is how we got to Saginaw. As for the "missing" impact signatures, you are again correct; none have been found. None have been searched for, either.... Our hypothetical impact is not a classic one. 5% of all impacts seen on the planets and moons are "oblique", arriving at near-tangential trajectories. These present entirely different characteristics than those that puncture the surface and come suddenly to a grinding and explosive halt. In the case of tangential "nicks", some of the impactor is expected to continue on its merry way, perhaps maintaining escape velocity.

Recent work by Schultz and Stickle (Lost Impacts) attempts to characterize the effects of highly oblique impacts into surfaces protected by a low-impedance layer, such as a 1km ice sheet. Their result shows minimal deformation to the underlying surfaces. The ice sheet existing over the Saginaw region (riding over it, not through it) at the time of our proposed impact not only protected, but also provided the enormous volume of water required to create our proposed hydrated slurry ejecta. In addition, it provided a vehicle to rework and cart away all the local ejecta deposited on top of the sheet.

The planform of ejecta from these events has been shown to be a "butterfly" shape, with ejecta thrown out laterally as the impactor ploughs through the earth's skin. We propose the Carolina bays to be geometrically wrapped around our crater to the East and the West, as shown in our thread posts here.

As you so correctly point out, using a Google Earth “staring” approach will not prove the hypothesis. Much "ground" work needs to be carried out. However, it is my goal here to leverage Google Earth to validate it as a viable GIS tool in mainstream science. By integrating the LiDAR images, and modeling the trajectories using Google Earth and our publicly-accessible Java calculator (hey, try it), I am comfortable that I have done more than stare.

In closing, I will reiterate my goal of using this thread to promote Google Earth as a viable exploration and visualization tool. In a recent post in the GEC Education/Tools Forum ( Embedded Google Earth Viewer Widgets ), I showed how a user's kml data can be presented within a web page, such as the one embedded in our site's Carolina bay discussion page.

Best wishes,
Carolina Bays Overprinting Ancestral Landforms Cintos 7/23/10 8:58 AM
Douglas Johnson, in his 1942 book "The Origins Of The Carolina Bays", made numerous observations about common bay planforms. Almost exclusively, Carolina bay formations "rest" within an anomalous sand stratum.  It is neither stratified nor laminated, but rather shows a hummocky, turbulated appearance. While there is a "rim", it stands only a very short distance above the surrounding pediment, and that pediment consists of the exact same sand. It is known that the basins do not distort either the surrounding sand nor the underling strata, but rather simply exist within that layer of course sand. Johnson made numerous observations which we feel supports our ejecta layer hypotheses. I have taken the liberty of including some of his text on our web site HERE.

With regard to Dr. Johnson's observations about "Outlet Channels Frequently Traverse Rim Barriers", we have identified a few locales in which we see the bay formations as clearly overlying ancestral drainage channels, and those channels mask up through the bay stratum.

The KMZ file attached to this post contains placemark and overlays to recreate these examples in Google Earth.

Here is one older drainage channel that continues to exist, even after the layer of ejecta blanketed it (in the Edinburgh, NC area):

Here is another similar example in the New Zion, SC area:

Another interesting example is also seen in the Edinburgh LiDAR. I this case, the bay rim actually "blocks" the underlying stream bed. We would expect that in these situations a "spring" would appear at the head of the channel.

The Edinburg LiDAR revels the overprinting of one layer of bays by a second, presumably arriving moments after the first. The first set of basy are seen as "fuzzy" from the ejecta mantle, while the later set of bays have crisp rims. Also shown is the creation of newer bays wholly within the earlier ones.

In the final image, a elevation profile is presented, showing that the bays created on the nearby "valley" floor are 4 meters below the elevation of the nearby bays. The bay orientations and sizes are nearly identical, in spite of the elevation differences, again suggesting the ejecta blanket was simply draped over the original landscape.

Please reference the attached kmz file for a look at these features from within Google Earth.

Allow me to also note that a web page is available that lists all the "fields" of Carolina bays we have been discussing, and each location has a link for the relevant kmz file and one for a jpg of the general LiDAR image: Location of Evaluated Carolina Bay Fields
Colorado's "Carolina bays" Cintos 8/1/10 5:43 PM
The high plains of North-Eastern Colorado is peppered with small aligned basins, similar in presentation to those of Nebraska. We again assume the original basins have been overlain by a thick blanket of late Wisconsinian loess. The orientations correlate well with the Saginaw crater. The USGS only provides 1/3 arc-second DEM data, so the level of detail is low. In spite of this, the basins are readily apparent, if a bit poorly defined. The Google Earth imagery is quite helpful here, as it led to their discovery, and as can be see below, do help define the orientation and confirm the "Carolina bay" planform.

The attached kmz file contains placemarks defining these three current "fields" in Colorado. I encourage the reader to download the kml and make their own assesment of our interpretations.

The overlay yellow lines in the graphics of 11 bays from the Anton, CO region, below, represent the Bearing Calculator's prediction of bay orientation in this area. Both the native Google Earth imagery and a color-ramp DEM image is shown for each location.
Bean Dips vs. Carolina Bays Cintos 8/20/10 10:06 AM

Our discussion here is focused on the area inland of the coastal plain of North and South Carolina, were the bays are rare. We propose that this is due to the higher relief of the landscape, which had caused most of the deposited ejecta to drape over hills and wash into the valleys, prohibiting the formation of the burst-bubble landform. That is, excepting any area of level terrain. Using LiDAR imagery within Google Earth's virtual globe, we have extended the range of Carolina bays back into the Carolina "Sandhills", above the fall line, by locating those level areas. When located, these basins typically conform the the predictions made by our Bearing Calculator.

The term “Sand Hills” has been applied to the area just above the Orangeburg Scarp, where the Middle Coastal Plain meets the Upper Coastal Plain. The bays are not numerous, but at least one researcher has noted the similarity of what is locally referred to as “Bean Dips” to the proper Carolina bay. And, again the sand is enigmatic: No fossils!

“The geologic history of the Carolina Sandhills is regarded as one of the most complex in the United States. Despite the dedication of many geologists, key questions regarding the origins and development of the Upper Coastal Plain remain unanswered. Much of the mystery stems from the great antiquity of this landscape, which is considered to be among the oldest exposed surfaces in the United States. To a large degree, the characteristic sands that mantle the region are responsible for much of the uncertainty that surrounds the origins and development of the Upper Coastal Plain. Due to the extremely porous nature of these sands, the interstream divides in the Sandhills have remained relatively stable and erosion-resistant, despite dramatic climatic fluctuations during the recent geological past. In addition, these sands are not conducive to fossil preservation, forcing geologists to rely, instead, on more imprecise dating methods based on relative stratigraphy.” by Mary McRae James Stevenson*

The following graphic depicts some of the bay fields identified along the NC-SC State line. Mark Twain once said “History Does Not Repeat Itself, But It Sure Does Rhyme”. Much the same could be said about these oval basins.

Here are sample bays, both in standard Goggle Earth Imagery, and with the LiDAR DEM color-ramp image superimposed on the virtual globe. They are displayed in order of increasing altitude, eventually reaching 139 meters (460 feet) above sea level. Given the very shallow relief of these landforms, little of their structure can be seen in the visual imagery, although the LiDAR images the oval planforms, leaving little doubt about their relationship to the bays further towards the coastline. The attached kml file contains links to the LiDAR imagery for these areas.

45 Meters Elevation Cheraw, SC

60 Meters Elevation Cheraw, SC

68 Meters Elevation Bennettsville, SC

78 Meters Elevation Gibson, NC

93 Meters Elevation Wallace, SC

100 Meters Elevation Diggs, NC

112 Meters Elevation McBee, SC

139 Meters Elevation Jefferson, SC

* Legumes In Loamy Soil Communities Of The Carolina Sandhills Their Natural Distributions And Performance Of Seeds And Seedlings Along Complex Ecological Gradients, Master’s Thesis,UNC Chapel Hill, 2000

The full Carolina bay kml: LiDAR INDEX
Oval vs ellipse vs ovoid Cintos 9/19/10 1:13 PM

Attempts to measure the inferred alignment of the Carolina bays requires quite a lot of "interpretation", as it were. It has been recognized that the bays are not truly ovals, but seen as ovoids or ellipses. (See Discussion on the Bay Planform page of our website). Certainly the crisp shapes of the Carolinas are standardized enough to build a shape that represents them. An overlay in this "prototype" shape would make it a bit easier to grab an orientation reference. In a process similar to the "Bearing Arrow", which is used to find the relative orientation of a field of bays, I now utilize a "bay Prototype" overlay to capture individual bay planform metrics.

This shape was generated using a graphic program, and tracing from numerous bays to get a good rendition of the typical North and South Carolina bay planform. It is shown here rotated clockwise 90º for space consideration on the page here. A Java program is being developed to process these overlays. The goal will be to capture the ranges of sizes and orientation of bays within a given field. A beta version is available on the Bay Planform Survey Tool page. The calculator's output is in tab-delimited fields, for import into a spreadsheet for additional manipulation. On goal is to build a histogram of bay sizes, for comparison to generic bubble-field distributions done in other realms.

The actual image is a transparent .PNG file. Here is an example of the overlay when used in Google Earth.

As you likely know, the imaging of bays has been enhanced by using LiDAR imagery. I have discovered a neat trick: if I create a set of image KMZ using Global Mapper, and then manually remove the "Level 5" images, I accomplish two things. First the file size is reduced substantially, allowing for quicker loading. (Some resolution is certainly lost, but for my purposes, that finer level 5 imagery was not all that useful). The second thing the removal does is create a situation where zooming in closer to the earth will eventually drop the LiDAR imagery out, allowing the underlying Google Earth standard imagery to be revealed. The result is the ability to simply rock back and forth between the LiDAR version of the landscape, and the visual version. For example, here is the area show above, backed out a bit, and bringing the LiDAR image into view:

I must say that this "shape" really does capture the general bay planform to a high degree of fidelity. Feel free to open the attached KMZ file and use Google Earth to create additional copies of the overlay. Click on one in the DOM and "copy/Paste" for a new instance, Then use the edit functions (or "get info") to move it over a different bay. After using the handles to rotate and resize, you will likely find it is a neat fit. Using this process, I hope to capture every bay across several "fields".

My ejecta proposal would suggest that the bay's distorted oval is a momentum artifact due to the Earth's rotational speed when the ejecta strikes the land. It remains to be determined just what the actual arrival bearing would lay, given the shape. My estimate - the yellow line - could be off by many degrees.

Adding to the post here: I have completed the first Bay Survey, identifying and capturing ~350 bays across a field of 270 square km in central SC. Two graphs are included here. The first one plots the size distribution of bays as a histogram, with one Hectare per bin. The second plots the length-width ratio of the ovoid bay planform against the area of the bay.

The KMZ file containing the LiDAR link and the individual bay overlays for "SC_Survey_B" is Available HERE.

These metrics need to be generated for (potentially) all 250 "fields" of bays in the field catalogue. Should anyone be in a position to lend a hand, I'd love to discuss a collaboration.

- Michael

PS: bragging rights here.... a high resolution image of the area represented in the KMZ attached to this post has been just awarded first place in the the GSA Annual Meeting 2010 Photo Contest & Exhibition. The entry is in the Abstract Images category (Depict patterns or form by way of photo-micrographs, satellite images, maps, or landscapes that capture a dynamic process or simply show the aesthetic patterns of geology at any scale.) A very high resolution jpg is available HERE.
HSV-Shaded DEM Overlays: USGS Index 100K Cintos 10/19/10 7:58 AM

Our survey of Carolina bay landforms has been carried out thus far using a "field-based" approach, with the count of fields approaching 250. These can be viewed using the Distal_Ejecta_Fields.kml file. These fields are randomly-sized, done in an attempt to keep their sizes down to the absolute minimum, especially when the 1-9 arc-second data is used. But the random sizes are presenting challenges as we attempt to count all the bays...

Most of the larger field DEM files have been recently further reduced by removing the tiles for the highest resolution, as they do not seem to have any additional information. It also allows for the user experience of toggling the DEM image off when zooming in very close to the earth - which then allows for viewing of the Google Earth imagery for comparison purposes.

To support an ongoing effort aimed at cataloguing all the visible Carolina bays, I have generated a set of HSV-shaded DEMs using the USGS's 100k index and 1/3 arc-second data. Each of the resulting tile sets are 1º of longitude in width, and 0.5º of latitude in height. By using a fixed-size regional view, perhaps we can correlate the differences seen in each of these areas. (OK, the areas are actually different areas as we move north to south due to Earth's curvature, but close enough...) A network-linked KML element for this new 100K Quadrant index is attached to the post. In addition, the html-based Google Earth Plug-in LiDAR browser has been updated to include a "100K Quad Index" button.

Here is a graphic showing the extent of our current Eastern Index:

And here is the graphic for the Nebraska area:

In both cases, my index placemarks contain a "KMZ" graphic linked to the actual kml elements to be downloaded and displayed in Google Earth.

The new survey is being attempted with overlays which mimic the shape of the bays. While the bays in any given areas are strikingly similar is shape and length-width ratio, that shape changes from region to region. Here are some of the ones in use currently (the Carolinas, Maryland/NJ, and the Mid-West):

The "squashed" nature of the bay oval planform may be an artifact of the difference between the Earth's rotational speed at the ejection vs the landing site, or it may be due to re-working by overriding dunes and wind erosion, or perhaps both...

- Michael
Give me LiDAR or give me ... B&W Cintos 11/7/10 8:56 PM

I had the pleasure of giving a talk at the 2010 GSA Meeting in Denver this past week. The subject was my use of LiDAR in this research.I demonstrated how the LiDARs were generated and integrated into Google Earth, and also showed how helpful the LiDAR is in visualizing bay planforms & measuring the inferred orientation of the bays.

But what to do when there is no LiDAR elevation data available? The historical Google Earth imagery of the east coast is quite helpful - especially the black and white from the 1990's. The area discussed here is the Sylvania 100K Quad, straddling the South Carolina - Georgia boarder. The KMZ file atached to this post has about 130 overlays using the "bay_south_prototype.png" version shown here, shown pointed due north (orientation 0º):

The DEM overlay in the KMZ is 1/3 arc second, and will drop out as you approach the earth. Less than 10% of the bays are obvious in the DEM alone.

My confidence in the overall hypothesis wains and wanes, but it is always a kick to drop this overlay over a smudge on the landscape and have it fit perfectly. (of course with a rotation and a scale to size...) Time after Time after Time after Time...

Here is a video of 20 of the bays in B&W, both without and with the overlay. Stills of these 20, along with the video, are available on the web site HERE.

The average orientation seen across the quad is 156º; rms 3.36º. Here is the Major/Minor axis ratio plot (slightly less oval than the Carolina archetype):

The neighboring Barnwell 100K Quad Get KML HERE has been annotated with 145 bays, using the same overlay. The average orientation is 156º.

The Laurinburg Quad is ~ 50% done. You can retrieve the KML HERE.

Should anyone be interested in assisting, I have another 60 of these 100K Quads to do. One possible collaboration tool is the Google Fusion Tables. I have been creating and managing the placemarks for each overlay using that facility, using a Java web-application available Here. You can see one of the sets using This LINK.

A new tool is available to generate bay overlays from a folder of placemarks using This LINK.

Best wishes,
Re: Inferred Orientation of Distal Ejecta popeyesmotto 12/3/10 9:19 AM

The past three weeks have brought me no end of wonder at the idea of a changed planetary history of the recent geologic past. The idea is very interesting and would remain a curiosity, much like Atlantis, except the facts of landform images newly revealed through google maps and similar geographic/geologic imaging tools. I enjoy objective facts and the possibility that the subjective reality we have previously believed may be changed by the introduction of new facts. Especially facts that re-illuminate the subjective theories that have been put forth as explanation.
Ah the Carolina Bays, so beautiful and so NUMEROUS when seen with a lidar filter layer. The western continental landforms in the sands of the midwest and their mirrored geometry to the Bays is amazing. To explore a new theory with a new set of tools has me excited.
I have been drawn into this exploration through exposure to Cable TV while traveling, thankfully I do not have to suffer with it here at home, and through a program concerning of all things Atlantis, I was reminded of the differential between current sea level and ice age sea level. A difference of nearly 400 feet. Explorations on the net brought me to the Younger Dryas Impact theory and eventually to the existence of the Carolina Bay structures covering the entire coastal plain of the eastern seaboard. Google Maps in satellite mode and Google Earth have brought hours of entertainment. Posts on various blogs and articles on sites spurred me on to look for other land forms that could be identified as having been effected by impacts or ejecta from a theorized impact or impacts in late pre-history.
I am attracted to Texas and it's environs since I have a fair acquaintance with the place and regularly pass to and through the state. Are there Bay shaped land forms there I wondered? Are there widespread evidence of impacts of recent vintage? I have been drawn to a web site which claims the western end of the North American continent and especially Central Mexico got whacked. Well maybe. Mostly I see volcanic features, heavily eroded. Uplift features heavily eroded. Perhaps a few bay structures in the coastal plain. I may go take a closer look at a few of them next month to see if the discontinuities and shapes can be seen close-up.
But eventually my eye was drawn to some enormous splash shaped features in West Texas and into the Panhandle. I have heard of the Monahans Dune Field and have friends who go there and ride down the dunes. I have never stopped in. I had no idea they were so darn big. And looking at them they appear to have just been laid upon the land, un-attached and un-related to any other feature nearby. The dune fields are tens of miles across and hundreds long. And they have an appearance of splatter. Big splatter. Like thrown out of a crater splatter.
I have investigated a bit on their geology and found a nice study of recent vintage which tries to claim them as having been derived from a formation that exists not too far away but in the wrong direction windwise. There are maps included showing the sand beds and sheet sands that are all associated with these holocene sand fields and darned if the maps don't show the Nebraska sites which have Bay structures to be a similar sand sheet and dune field. Gosh , the maps show all sorts of these sand sheets that are not easily explained. All of them are described as Holocene creations and have no easily correlated sources, but do share very similar descriptions of the sand grains and chemistry. The one map shows all sorts of these fields in the midwest.
Links to Article

Quick Read

It took me a while to realize that the theory being put forth was that of an ejecta blanked coming up out of the impact, and that the ejecta due to its likely high moisture content was coming down as a deep wet blanket of material with large voids of water or dense vapor distributed in some parts of the ejecta's volume resulting in the formation of the Bay shapes. Maybe like very wet sand.
It could be theorized that if the western wing of the ejecta blanked was not entrained with water to the same extent as the eastern blanket, or that the dynamics prohibited the voids to form in the wet sand in the same fashion as occurred in the eastern blanket, the Bay shapes might not form. Or perhaps the sand dried out in the arid climate and then the wind reformed the sand thus obliterating the Bay shapes. I was fascinated to learn that the deep sands on the eastern seaboard do not contain any fossils and the sources for the sand deposits are unknown.

There was also a link to the Monahans Dunes reporting that a skeleton of a paleo-indian female had been found and excavated in 1953 from a site in the dunes and dating had possibly been as far back as 11-12 kya. The remains are not on display and are instead held at a museum at SMU in Dallas. I would just hate to think that this woman was out as the impact to the north hurled this big splat of wet sand at her which instantly killed and buried her. Just if the theory were to have some basis in reality.


I have been exploring more in Google Earth and believe that your scope is too small. I will soon post google earth kmz links to a whole bunch of suspicious shapes which fall into an expanded arc that links the Bays in the East to the sand sheets in the west. I will attempt to remain conservative in my choices as there is a remarkable amount of human worked land throughout the US.
Re: Inferred Orientation of Distal Ejecta Cintos 12/4/10 2:21 PM
Greetings I yam what I yam! :

Thank you for your expressive observations. You have successfully decrypted our message: Its all about the SAND.

The hypothesis suggesting a distal ejecta blanket of pulverized rock ( i.e., sand) demands a whole new frame of reference for the cosmic impact concept, one which the overwhelming majority of the academics and scientists I have communicated with can NOT seem to grasp. They are all hung up on something big coming in and blasting out the bay. Then, appropriately, they point out that the bays are not impact craters... duh! Right, they are not.

Furthermore, you got the concept of a fluidized deposition, one in which the bays are merely voids ( I call them popped bubbles, see image below). You astutely recognize that other ejecta may not have had as much fluid in them - what would that look like? Perhaps sand dunes? And again, perhaps there were bays at the start but the high desert locations did not offer the required water table to keep them stabilized as bays, and they simply blew away. (one other mechanism further north is that the sand fell on top of the glacial shield; there we propose ice-walled-lake-plains as the result).

So, you are likely correct that the extent of the ejecta distribution extends well beyond the sub arcs of the annulus surrounding the proposed Saginaw impact site. However... it remains important to us that we can correlate any landforms documented by identifying the inferred alignment obvious in the Carolina bay depressions. Sand dune depositions will obviously be oriented according to the prevailing winds. ( wind-driven dunes are visible in all the LiDAR, and those dunes are NOT aligned with the bays by any stretch of the imagination).

Allow me to use this post to provide a LINK to the presentation I gave at the 1010 GSA Meeting in Denver. This version has my lecture text in each slide - the original was delivered with bigger graphics and much less text.

The attached file contains the index to our collection of Carolina bay "fields". Each placemark contains a popup with a link to the LiDAR.
Pee Dee River: Dammed by a Bay? Cintos 12/29/10 7:54 AM

Our Survey of all bays is moving along, albeit a bit slowly as there are simply so many of them. 1000+ bays have been mapped in the Elizabethtown 100K Quad , and over 2,000 in the Laurinburg 100K Quad .

While working on the Florence 100K Quad area, I was drawn to the state of the Pee Dee River Valley in the area east of Quinby, SC. At some time in the past, the river created a tortured series of meanders, oxbows and relic channels as it struggled across a narrowing of the channel at 34.3, -79.6.

Coincidentally (?) this location marks the NW end of a large "Carolina bay" planform basin, aligned similarly to all the other bays in the area.

Could the "arrival" of my proposed ejecta blanket of sand created a bay, whose NW rim dammed the Pee Dee at that point? If so, perhaps the Pee Dee backed up, forming a flood plain, which quicly filled in with sediment as the river meandered about. As the river finaly crested the bay rim "dam", the meandering stopped as the river cut through the recently-deposited sediments upstream along an increasingly more direct path.

If there were any truth to this hypothesis, the upriver sedimentary record might well provide a hint for the timing of the event. Then again, this whole scenario may be driven by the lack of LiDAR-quality NED data in the county of Florence, which is in the lower left of the image, and includes the "bay" shaped feature.

The attached kmz file contains elements to display the proposal in Google Earth.

- Michael
5,500 Tagged Carolina Bays, and counting.... Cintos 2/1/11 9:11 PM

The effort to identify and measure the population of Carolina bays continues. At present, the Fusion Table database contains over 5,500 bays. Each one has been measured for length, breath, and orientation. A Overlay KML element exists for each one, and is linked for download through the visualization placemark on the Virtual Globe. I project reaching over 50,000 bays in the future.

The graph here displays a histogram of bays in 1 hectare buckets. The rendering of a perfect power spectrum is suggestive that a unique process is at work in their formation.

The process is time consuming, even with a number of specialized programs at hand to process the data. Should you have any interest in participating, there is a set of web pages being assembled to provide guidance to volunteers, HERE.

Allow me to share an unusual bay, sitting up on a knoll as if it were a man-made reservoir. How would this hilltop support a water table so high above the surrounding drainage? The wave-shapped oval theories all require open water fetch to generate the required wave action.

The attached kmz file contains the working set for the "139316" Octant, in the USGS Laurinburg 100K Quadrant. This 240 square mile area has 540 bays listed. For more information as to how this is structured, please reference This Page .

Best wishes,
10,000 Bays Surveyed! Cintos 2/18/11 11:40 AM

We are celebrating a milestone today: our survey database in the Google Fusion Table facility has surpassed the 10,000 mark.

UPDATE: As of 4/4/2011, the survey has reached the 20,000 bay mark.

Each bay is represented by a discrete overlay, which is annotated with metrics such as the inferred orientation, eccentricity and the surface area of the bay.

With this post, we have also finalized the survey LiDAR and bays for the State of Nebraska. The survey has most of the south eastern section of the state viewable in 1/9 arc second LiDAR, using data from the NEbraska DNR. The attached kmz file offers links to all this, and more, as there are six USGS 100K quads with LiDAR, and another six with the generally available 1/3 arc second DEM overlays. The file is set to only open one Quadrant at a time, so your machine will not be overwhelmed.

The attached kmz opens to the following image:

Note that you must select the placemark balloon to see the pop-up and download the survey overlay. The corresponding Satellite image is shown below:

Please note that the LiDAR resolution is super-high resolution, and the GlobalMapper-generated DEM is using 10x elevation exaggeration. These features, although stunning in the LiDAR, are often imperceivable to a viewer on the surface of the earth. As proof, note that the circular irrigation gantry system has no problem traversing these rims, as shown in the satellite view.

We sense that only the largest of the bays survived the intervening 40,000 years of wind erosion and in-filling by wind-blown sand and dust. Those that do still exists (we identify ~500 in the survey) have characteristics very similar to the Eastern bays, including the overlapping seen in the following slides. What is most exciting to me is the expression of one remnant rim in the floor of an adjacent bay. We interpret the more southerly bay was created moments before the one to its north, as the frothy rain of sand coated the existing terrain and numerous super-heated steam inclusions in the ejecta were "popped".

The 10,00 bays likely represents 10% of the final tally. We could use more help in accomplishing the survey, and to run QA checks on our existing data. I have put a few MOVIES up on the web site that show the process in action.
The Bay-Bells of Maryland, Delaware and New Jersey Cintos 4/9/11 8:25 PM

NOTE: As of Thursday, June 23, the Google Fusion Facility being used to deliver the bay placemarks is again providing placemarks. The survey has reached ~25,000 bays

Do the Carolina bays extend across the North Carolina border into Virginia, Maryland, Delaware and New Jersey? We believe so, and base our interpretation on LiDAR imagery of various basins across the coastal plains of those states. We have no LiDAR for this area NOTE: we just received LiDAR for Virginia's Eastern Shore, which demonstrates, along with the imagery presented in Google Earth, a continuation of the basins across the NC border into VA. We now see the Eastern Shore of Virginia dotted with over 700 symmetrical oval shaped bays, vs the slightly skewed (s/w side flattened) in North Carolina.

Once into Maryland (moving northward along the coast) we see a new planform slowly evolve. We have nicknamed this the “bellBay_Prototype”, and it continues to persist northward, across New Jersey, and right up to the Late Wisconsin glacial moraine and outwash tills. A representation of the overlay used in our Survey is shown here rotated to a generic orientation:

Until the advent of LiDAR, the odd shape these planforms present in aerial imagery did little to help identify their lineage. Indeed, when these were identified as Carolina bays by others, they have intrepeted the orientation as due north. This is especially true as the bays become squatter and more oval. In New Jersey, these landforms are considered to be thermokarst features formed during thaw of permafrost after the last glacial maximum 20,000 years ago. Our working hypothesis suggests that the bays - as “burst bubbles” within a superheated sandy ejecta blanket - will generate planforms based on the skewing generated by the ejecta’s velocity vector and that of the earth at the landing site. This changes based as a function of the cosine of the latitudes of impact site and the ejecta's eventual emplacement location, as discussed in an earlier post in this thread .

The area we consider here is often covered with extensive sheets of sand dunes, likely created during dry periods over the millennia. We see in some areas where these dunes and our bay-Bells interact, but as in the Carolinas, the two planforms are quite distinct and very easy to discriminate between. See my Dunes & Bays Post.

The LiDAR images below progress northward from southern Maryland, and we see the bays become less oval and more pointed on the NW end, even squatter, and - most importantly - flatter on the southeastern side. The attached KMZ file contains placemarks for about 1500 bays contained in our survey

Here is an area in central Maryland presented as an Elevation map:

The examples continue into the state of New Jersey, the first here is immediately east of the Delaware Memorial Bridge...

We reach up into the scholarly Princeton, New Jersey area:

Our most Northerly bay complex is in Mammoth County, just north of the US Navy’s Weapons Station Earl.


The area covered by the above imagery extends across multiple 100K Quadrants. The KMZ files for these are linked to their USGS names, below, listed from south to North:




Wilmington, DE



An alternative to the direct KMZ links above and attached to the thread, you might have a look at the map generated by a Google Fusion Table HERE, to open a new browser window and load directly from Google’s site. Please note that our survey is far from complete in this region, due to a lack of LiDAR coverage and time resources.
SW Alabama - Carolina Bays or Karsts? Cintos 7/17/11 8:15 PM

Our search for bays in additional locations continues, even as we measure more bays in areas already identified. Our Survey total is now over 25,000 bays. Each of these has been documented with numerous metrics and can be reviewed through this Google Earth Fusion Table visualization.

Our distal ejecta sand sheet hypothesis suggest that bays will only be created on level terrain. Our first look in new areas is always to find the level spots on the digital elevation maps. In southwestern Alabama, that led us to the smooth areas seen in the Atmore 100K quad, shown in the image below.

Now, these landforms do not present the typical robust planform and alignment we usually document. Alabama has no LiDAR, however, so all we see is the water/wet areas, not the actual rims. It could be argued that these are simply Karst landforms ( formed by the dissolution of a layer of soluble bedrock,typically carbonate). Only ground truthing will tell. Hints of alignment are seen. Since this area is directly south of the Saginaw region, we would expect the bays to be orientated from the north - actually, more from the NNE due to Coriolis steering. Here is a graphic showing the 1997 black and white Google Earth historical imagery of some of these.

We have only attempted to document a few of these in the 124350 and 124349 octants. These are included in the attached KMZ file. Each placemark can be selected to show a popup balloon. There, you will see the metrics in our table, and a link to retrieve our suggested overlay outline.

- Michael
Pocosin: swamp-on-hill Cintos 8/16/11 8:58 PM

Native Americans referred to a Carolina bay as a "Pocosin". Interesting word. According to Wikipedia®
The word pocosin comes from an Eastern Algonquian word meaning "swamp-on-a-hill."

Now, I may be a bit batty, but I can't for the life of me understand why that appellation has not raised the collective eyebrows of geologists. Swamps should exist in lowlands, not up on hills. Am I right?

The ongoing survey of Carolina bays has reinforced the concept. In many cases, the LiDAR clearly shows that the bays are expressed as depressions in the surrounding landscape, with no true raised rim. Yes, rims do exist in thousands of bays, but could they simply be wind-driven artifacts, created after the depression was generated?

The area around Minturn, SC, displays numerous "swamp-on-a-hill" examples. The attached KMZ has the elevation map overlay generated from LiDAR data, along with a network link to the survey results over the map extent (initially toggled off).

Note that the shape of the bay has no relationship to it's elevation. They are seen from 35m to 60 m, over a relatively small 14km x 18km region. In many cases, the bay floors are elevated 10 meters above the surrounding drainage channels. They have orientations within the range of 142º to 146º.
Re: The Bay-Bells of Maryland, Delaware and New Jersey Cintos 10/14/11 12:44 PM

I had the pleasure of presenting a Poster (get it HERE ) at the 2011 GSA Meeting this week. The topic was specifically these more northerly bays. According to the general casual understanding - as expressed at Wikipedia and sources - the bays lose their robust parallel alignments towards the north-central US, and rotate to a more northerly indicated source. Now that we have the LiDAR, an alternate interpretation can be made.

Our survey of Carolina bays has reached 27,000. When computed as distal ejecta traveling on ballistic trajectories which are steered by the Coriolis forces, they continue to support a triangulation network focused on Saginaw bay.

You can get the latest Google Fusion Table map visualization HERE. The placemarks will link to individual bay overlays for Google Earth viewing.

The attached KMZ file has placemarks which link to survey data in the scope of a 100K USGS Quadrille.

- michael
Re: The Bay-Bells of Maryland, Delaware and New Jersey Hill 10/15/11 9:52 AM
Cintos, I continue to read your investigations with great interest. I never could have imagined when I began my first post about the Carolina Bays in 2005 that Google Earth would ever lead to such important research about them. I'm always waiting for the next update.
Click the images to view two important videos about our only home.
A Pale Blue Dot HOME

“Only when the last tree has died, the last river has been poisoned, and the last fish has been caught will we realize we cannot eat money.” Cree Proverb

My avatar Mr. Raven posed for me at the Tower of London.
Re: Inferred Orientation of Distal Ejecta Peachlaser 1/24/12 9:21 AM
Really enjoying reading your posts.

Yesterday, I was driving through North and South Carolina. One of my missions to was to go to Little Pee Dee State Park in SC while doing some genealogical research. And, I hoped to spot a Carolina Bay or two along the way. Much to my surprise, the park is basically one or more bays in a fairly well preserved state.

A question, is the sand deposited part of the substrate of Michigan or could the comet have also contained some of this sand?
Re: The Bay-Bells of Maryland, Delaware and New Jersey Markopolo 1/24/12 10:16 AM
This thread, with it's potential implications for a new understanding of the geology of North America, is among the best in all of the GEC, I believe. The level of detail, the amount and quality of the data, the illustrations and visual representations of the thesis using the GE browser, all of these are simply captivating, and an excellent use of Google Earth. I have to say that this is Exhibit "A" in how to use Google Earth to make a point. I'm not geologist enough to know whether or not the point is proved to, or accepted by, the community of geological professionals, but I'd think the author has certainly done his part to build the case.

Congratulations, Cintos! Like Hill, I look forward to your updates.
Wherever you go, there you are.

Google Sky Blank Spot Explained
Re: Inferred Orientation of Distal Ejecta Cintos 1/25/12 3:52 PM
Greetings Peachlaser:

Nice to hear of your interest in the beautiful Carolina Bays. All this talk about the geology of the bays should not overshadow the biological diversity seen in them.

The question of the provenance of the sand is a difficult one for me at this point in my project, and your specific question is directly to the point! If there was excavation in Michigan, the sand would be composed of pulverized sandstone from the top layers of the Michigan subsidence basin. The Saginaw Aquifer (Pennsylvanian) is the most likely candidate.

As for the impactor itself, there just might some of that scattered around, too. I have tried to float a few trial balloons about just how a "hydrated silicate" cosmic body would be rendered during such a collision. There obviously would be a lot of vaporized cometary material. Could it be distributed as high-purtity quartz sand? There is no scientific support for the distribution of sand as a by-product of impacts, although I continue to find examples of thick quartz sand (now sandstone) deposits associated with several known impact structures. These seem to be monolithic, "hummocky" deposits of coarse sand of similar grain sizes, and are considered by scientists to be the result of tsunamis or submarine landslides.

The hypothesis is pushing the envelope way beyond the current knowledge base of impact physics here. Sort of like Wegener proposing that the continents went crashing through the oceans, before the physics of mid-oceanic seafloor spreading were discovered. Prior to that discovery, all geologist considered the earth to be shrinking and shriveling up!

Best wishes,
Re: Inferred Orientation of Distal Ejecta Peachlaser 1/26/12 12:45 PM
Have you viewed the area around Little Pee Dee State Park. It is south of Dillon in the far eastern part of South Carolina and almost on the NC border near Lumberton. The white sand really stands out in the Google Maps on my iPad. Most of the sand is on the road out of the park.

I have also heard that the sands of Michigan on the northern coast of Lake Michigan is a special type of sand that was coveted by the foundries before the dunes were protected. The crystals are supposedly rounded so that they do not stick and are valued for this feature when castings were being made. I should of looked closer at the LPD sand. It is definitely very white. Almost looks like snow.
Re: Pocosin: swamp-on-hill Peachlaser 1/26/12 12:52 PM
On Tuesday, I visited the tribal headquarters of the Lumbee Indians. While on a tour of the building, I was shown several paintings and drawings that represented the history of the tribe. On each of these, I noticed that a prominent element was an elongated circle.

I emailed my host and asked if the Lumbee recognize the shape of the Carolina Bay in their traditions? His reply, "Yes, we believe in the circle of life."
Re: Pocosin: swamp-on-hill Cintos 3/14/12 8:14 PM
Greetings Peachlaser:

Extensive archeological research has been done along the rims of Carolina bays in South Carolina. There are extensive findings of evidence for human habitation along the rims over the past 10,000 years. It is good to see that the descendants of those paleo indians carry forward such verbal and graphic histories.

Here is a youTube video about the Savannah River Archaeological Research Program:
Best wishes,
Re: Inferred Orientation of Distal Ejecta Cintos 3/15/12 5:37 AM
Greetings Peachlaser:

The Little Pee Dee River area is in Dillon County, which is covered in available LiDAR elevation data, flown in 2008. (None of the LiDAR flown in 2009 has been published yet).  So,yes, the survey includes the area around the park. The attached KMZ file will load the survey results for Octant #137317.

 Carolina bays are well represented. There are obvious fields of dunes formed on the east bank of the river. In one case the dunes infiltrate a bay, but for the rest, the bays look to be cut into the dunes. I interpret that bays holding water are not over ridden, but instead adsorb and distribute the influx in sediments.  If the bay is drained, such as the large one here, the dunes can grow into the basin. Bays with antecedent drainage channels broaching them were noted by Douglas Johnson as disproving primary impact morphology. The drainage seen into the large bay is certainly over fit for it's duty, suggesting there was once a larger branch stream running north from the Little Pee Dee in that location.

Re: Inferred Orientation of Distal Ejecta wraps 9/9/12 10:52 AM
this is very interesting, because Hill pointed me here after I made a post correcting  a pond my grandfather enlarged in northwest florida, possible carolina bay type .I live about 25 miles to the northwest of it now, but I have always been finding small magnetic blobs of metal on a hill where my mom and dad live.
City of Aiken, SC is Riddled by bays Cintos 10/15/12 7:08 AM

Recent work to render the old Aiken County LiDAR elevation data set has resulted in a usable tiled set of imagery overlays for Google Earth. In addition to the county LiDAR, I also updated the Aiken Quad KMZ set for the survey. 

The data on the seamless servers is 1'x1' gridded, which makes for an enormous file over the extent of a 100K quad, so that is reduced to 10'x10' grid. I did generate a smaller area which encompasses the city of Aiken in the 1'x1' grid, which can be visualize using the attached AikenCityHighRez.kmz , and while the actual imagery is huge, the networked link will bring it in as a paged tiled set. 

The city has a few large Carolina bay depressions. One has been preserved as a park and nature area, focusing on the bay and its habitat. Actually, the water surface only covers a small area at the center of the depression, but most of the bay extend is preserved.  I discussed this bay in a recent Carolina Bay of the Day post on Google+. The largest bay has been drained and is in use as a ballpark and recreational facility, as seen in the images attached here.

The Carolina bays in this region are "Bay South" archetype, and display alignments which correlate precisely to the Saginaw Bay triangulation calculator's predictions.

Re: Inferred Orientation of Distal Ejecta David R. Kimbel 11/8/13 9:03 AM


I was invited to join Younger Dryas impact research community ten years ago.  One of my objectives has been to help correct the explanation for the Carolina Bays promoted by the NC Parks Department at Jones’ Lake and Lake Waccamaw.

Over the past ten years, I have been generously trained, supplied and given co-authorship status of YDB papers for my work, as well as personally benefiting by staying active and interested.

Recently, I’ve worked out a scenario which seems to fit all the Bay forms I digitized in North Carolina using NC DOT LIDAR, some years ago.

It includes effects on the composition and distribution of the “black mat” which may be of interest to the YDB research community.

As a very rough draft of a presentation to the Parks Department, I offer:


What about the Carolina Bays?

Rare ice fall




Carolina Bays (poccosins) are the only clearly visible, direct evidence of high-velocity impacts of extraterrestrial objects through the Laurentide ice sheet + 13,000 years ago starting the Younger Dryas “mini” ice age.

The basic shape of a Carolina Bay is circular, however, a large number are elongated (miss-identified as elliptical).  


“Downburst damage can be differentiated from that of a tornado because the resulting destruction is circular and radiates away from the center. Tornado damage radiates inward, towards the center of the damage. “


Circular downbursts in, and moving with, the outflow from ET impact sites dropped straight toward the center of the earth and excavated the Bay forms.  Straight-line elongation of a Bay is on a bearing, and for a distance related to, the duration of the downburst and the atmosphere’s movement.



GE overlay of 13,218 North Carolina Bays, Cumberland County LIDAR DEM (color)


Visible Bays in North Carolina are (generally) excavated into unconsolidated sediments left behind as the ocean receded to its current beach line.  NC’s Bays represent only a fraction of the still-visible Bay forms (to be) found in coastal plain sediments of the eastern United States.


Published data of findings in multiple locations, often within a fine-sediment enriched, “black mat” layer, has shown widespread dispersal of impact debris and established a Younger Dryas Boundary (YDB) strata.

The "black mat" has not been found inside any Bay sampled, but other YDB evidence has been found along the bottom of Bays’ bowl-shaped excavation (SK Bay, GH Bay) and the base of Bays’ ejecta rims (Blackville Bay, Clay Bay #1).


Clay Bay #1 (Hoke County) sample data and experimental results suggest that sand blasting produced about 50 grams of fine powder (smaller than 53 microns) per kilogram of ejecta during the Bay’s formation.



Preliminary analysis of Clay Bay #1 samples suggests layers of surface sediment

before and after downburst sandblasting


Based on the Clay Bay #1 computation, the 119 Bays in Cumberland County would have ejected an estimated 1,055,130,398 kilograms of fines across their rims into very turbulent, upwelling atmosphere surrounding each downburst.


Taken together, North Carolina’s Bays contributed an estimated 74,757,637,951 kilograms* of shattered, fine sediments to the mix of the “black mat”.


North Carolina’s Bays, however, may be as little as one twentieth of the total continental downburst coverage.  The vast majority of the continent’s downburst ground effects  have been, and are being, destroyed by alluvial, eolian, vegetative, and human activities.


Movement of the downburst that created a particular NC Bay can be seen by connecting the center of the circle blown out on touchdown of the downburst, at the northwest end, to the center of the circle being blown out as the downburst stops at the southeast end (C-to-C).


The reverse bearing of the C-to-C line (southeast to northwest) points roughly toward the Laurentine ice sheet impact sites where thousands of cubic kilometers of ice were either: dissociated to elements in plasma, then burned back to water vapor (ET bolide moving through ice); or, powdered by the horizontal shock wave through the ice; or, boiled to steam inside ET’s earth crater(s); and then, thrown above the atmosphere as molecular ice.


The special properties of water - as ice, (4% greater volume than water); - as liquid, (does not compress); and, -as steam (compressible up to 7300 psi) disperses energy from the ET impact(s) to the trajectories of downburst-forming, molecular ice clouds.

Note that any wet ejecta, coming from the ET’s crater, is “freeze dried” as it leaves the atmosphere, purifying the water as it crystallizes to molecular ice (snow).



Snow vortices form within 3 meters

Re-entering the atmosphere at thousands of kilometers per hour, the twisting, snake-like snow clouds burn into and pull air down. 

Centrifugal force and the Coriolis Effect cause the dropping air/ice mix to spin into vortices that deliver extreme high pressure, high temperature, air/steam mix in tight, circular, sustained downbursts, perpendicular to the ground.


The size and duration of the downburst moving along the C-to-C line of any particular Bay is related to: the overall trajectory and cross section of its snow cloud out of, and back into the atmosphere; the snow cloud’s length along that trajectory; the distance traveled outside the atmosphere; the cloud’s density variations - related to the ice sheet’s crevasses, mountains, etc. at the ET impact site; and, the cloud’s coalescing  and lengthening as the ice particles crash together (negating their closing velocities).


The bearing of the C–to–C line reflects the downburst being carried across the surface by the outflow from ET impact site(s), and the atmospheric turbulence created by proceeding or nearby downbursts.


The three-Bay complex at site Clay Bay # 1


As the downburst moves along this center line, sediment is blasted out in all directions from the center, leaving a circular northwest rim; wider, long-radius curved, side rim sections (ejecta is “fanned out” as the downburst moves by); and, a high, circular southeast rim.


Visible, “whiter-than-white” sand rims are made of repeatedly-blasted sand accumulating as the downburst excavation moves along the C-to-C line, pulverizing and blowing away softer materials, leaving (mostly) clean, shattered-along-crystal-planes quartz.  These smooth surfaces, some with micro-planer fractures, reflect light in the same manner as snow.


A particular downburst’s movement across the ground, C-to-C, toward the southeast is due to: “forward”, out-of-balance loading at the top of the vortex, causing skipping; and, more importantly, the outflow of the atmosphere away from still-steaming ET impact sites.  In rare instances (Brunswick County shown), it appears that outflow from a different ET impact arrived mid-downburst.  


Successive multiple Bays tend to be aligned north (left of C- to – C line) due to downburst rotation.



       Harrison Bay complex Cumberland County, NC                                    Brunswick County, NC


The vast range of sizes of Bays, with virtually the same ground effects, plus the orientation of their C-to-C lines, strongly suggests that ice ejecta from multiple large ET strikes into the Lauretide ice sheet created the Bay forms, and in the process, added to the composition of, and distribution of the “black mat” and other YDB indicators.




David R. Kimbel,

Independent Researcher





*The ejecta calculations use: Clay Bay #1’s loss of 50 grams of fines per kilogram of total ejecta; 1,922 Kg/m3  weight of wet sand; formulae derived from measurements of Cumberland County’s Bays; uniform Bay depth of 15’ from center to 2/3 of the radius, then depth-tapered to perimeter; and, data generated by Global Mapper.


The calculations do not include the still-burning, biosphere debris (left behind by ET’s supersonic heat and pressure waves) that would have been pulverized by the downburst and added to airborne “black mat” materials for some distance around each Bay. 

<span style="font-size:13.5pt;font-famil

Re: Inferred Orientation of Distal Ejecta Fwob 1/15/14 10:40 AM
Where can I find a summarized version of everything here? 
Re: Inferred Orientation of Distal Ejecta Cintos 1/20/14 11:53 AM
Greetings Fwob:

My efforts at researching the Carolina Bays can be reviewed beginning the root of my Cintos web site. A synopsis (although a few years old) of my thoughts here is on a sub page @ Saginaw Manifold Overview.

While "following the data", I have been pushing back the presumed date of the event - currently somewhere between 800 ka and 150 ka.

More topics »