The Meander Mechanism
from Alternative Stream Researchby Lloyd Ramsey, who says:
"Common notions about streams range only from wrong, to backwards."
all photos are COPYRIGHTED, to be viewed only at this website
|
PREFACE The original hope of this study was that from a large amount and range of observations, some general notions of water behavior would emerge. This webpage presents current results as The Meander Mechanism. The theory derives only from direct observations of real stream flow, and is about only real stream flow as observed. No other theory is supported. The entirety of this theory and all parts of this presentation and all modifications are continually protected from all other claim and copy and distribution under Copyright and Intellectual Property laws which you agree to follow by proceeding. |
|
ABSTRACT The general theory is that streams develop channel forms only as temporary states, as water continually responds to 1) gravity, 2) internal cohesion, 3) obstructions not yet moved, and 4) material currently moved. The key finding is that storm flows develop an accumulation of movable materials on the bed forming a central and generally straight ridge. Where the Central Ridge approaches a channel bank, lower flows are impounded so that the ridge is breached by flow across the ridge. The crossing flow erodes a descending Breach Ramp that will increasingly direct general flows so that the stream bed and receiving bank erode to the next stage of forms characterized as Meanders. Where water falls, the Rim Deposit resulting from the Plunge Pit will similarly be breached at an unpredictable point to serve as the ramp for similar redirection of general flow and bank erosion. |
|
EXPLANATION Gravity The force driving water flow is gravity, so that water will move until there is no direction that is down. Flowing over a bed, however, is not the same as falling directly over an edge as at a log or dam. Falling water is continually accelerating, while water flowing on a bed is restrained to a velocity relative to the slope of the bed. Note that a stream is independent of valley slope, since the channel is by definition separated from the valley surface. The direction of falling water provides extreme erosion that has no logical limit, nor any limit suggested by this study. The Plunge Pit that develops varies in form from nearly circular to elliptical, and will excavate under any upstream support. The pit is continually being flushed, with material generally deposited as a raised rim. This Rim Deposit will be breached downstream, commonly to both sides of the channel leaving a tongue-shaped deposit. A dominant flow may develop to one side in the manner of a meander, leaving the rim to develop as a minor bar. Flowing water will then spread the deposit material as wide shallows or Shoals. |
 |
 |
|
Photo at left is looking upstream to where water falls over end of rocks and rubble dumped and washed down from another Plunge Pit upstream. Urban storm flow is continually deepening and enlarging this pit. The Rim Deposit has been breached at view right by flow around a boulder placed as a spur to "deflect water away from the bank" but flow is sent into the bank. A normal breach of the rim can be seen at view bottom center, looking like a tiny "riffle" but more importantly directing flow against the bank. There is already a bar deposit forming on the other side. Photo at right is of a larger urban Plunge Pit caused by a brush and trash jam seen at view top. The silt fill of this area has allowed a very geometric form to the pit and rim deposit. |
 |
 |
|
Photo at left is view upstream where water is eroding a pit around a large natural boulder in view upper left to make a turn. The Rim Deposit across view middle has been breached more strongly on this side, leaving a Tongue Deposit at view middle right. Photo at right is downstream from bridge over the Mussellshell River, Montana. Material flushed from bed at structures has collected as a rough Rim Deposit broken into shifting tongues by various breaches. The main breach flow is on the side to view right, with bank erosion evident. |
 |
 |
|
Photo at left is a small Check Dam built as an arc with single boulders on added rock and gravel bed for a city park in Utah. With no water present, you can clearly see a Plunge Pit eroded beneath the falls and undercutting the boulders. The trapezoidal one is slipping down, and when it goes, the rest will soon follow. That is why Rosgen stacks two layers of boulders for his various "step-pool" structures -- but, why make such things at all? Bed material washed over the dam in storm flows is continually flushed into a mound to view left. Flow would be going around both sides equally, forming the typical Tongue Deposit. Photo at right is looking upstream to a small "gabion" check dam. A Gabion Mattress of rocks in heavy wire mesh was placed below the intended fall to prevent pit erosion. But at the end of the mattress, as at the end of any more resistent material, a Plunge Pit has developed with a typical ellipsoid shape. The end of the mattress is being undercut and is sagging. White caliche shows the Rim Deposit characteristic, with dominant flow tending to view right despite very steep bank. The artificial "riffles" that Rosgen teaches people to install will produce this same End of Surface Effect, with an unplanned Plunge Pit and disruptive deposits as the likely eventual results. |
 |
 |
|
These two photos are both from a fish habitat project in Idaho, typical of older "bioengineering". A check dam using a single log on the left has been undercut completely by enlarging erosion of the Plunge Pit. The photo at right was two stacked logs and shows remains of wire mesh with plastic geofabric that was attached to the upstream side to retain rock bed fill, but is undercut and failing the same way. This same development occurs commonly at sewer pipes or other such items that cross a stream. Any source of falling water will make the same sort of enlarging pit in any movable material. Dave Rosgen tries more recently to use boulders too large, he thinks, to be moved -- but for the boulders to move, all that has to move, is what is under the boulders. The next two photos are from the most elaborate and expensive check dam seen by this study, and was apparently solving the pit problem -- until the end of the added rock bed. This indicates the potentially infinite extent of Plunge Pit erosion: |
 |
 |
|
The photo on the left is a Low Drop Structure built by the Corps of Engineers, near its home base in Mississippi, at a cost of about a half million, using steel sheet-piling with a concrete cap. There is a "baffle" of sheet steel at view left to spread the water, putting the rim deposit onto a patch on the bank. Shortly downstream, rip-rap for the bed and banks stopped, and in 2001 the resulting pit extended for about an acre. The other side is becoming a major sediment deposit growing trees, with this side becoming a massive meander. Rosgen would have you believe that this "riffle-pool sequence" is a great benefit, and he teaches installing such conditions on all projects wherever you might like. Deeper water is pressed down by gravity acting cumulatively on water above, so that the force available for erosion is increased by higher flows and in deeper parts of a bed. This means that deep places will always be flushing material out to remain deep and likely become deeper, and generally adding to the rim deposit and downstream distribution of it, and in that way contributing to meandering of the channel. The strangest gravity effect is that a higher water level in a given channel will therefore go faster, as if falling off itself. Having only other water for a "bed", the tendency is to flow straight, generating the ridge/ramp process described following. It should be remembered that, except for bedrock or other solid conditions, what seems to be the bed of a stream is really a totally saturated accumulation of previously moved material, which becomes effectively fluid itself with stronger and faster flows. Internal cohesion The internal cohesion of water, one of its properties as a fluid, is how distributed precipitation gathers into distinct flows to begin with, and how it then offers buoyancy. After that, cohesion is not easily seen in stream flow, and seems to be contradictory. Within any channel, there will be various distinct flows. Where two streams meet, they will typically remain intact for considerable distance. Deep flows will have layers of behavior, so that bottom flow that falls over a bed feature will then form a different "bed" so that water above does not respond to the actual bed feature. Similarly, wide flows are many adjacent flows responding to different features and influences. |
 |
 |
|
The photo at left looking upstream shows a single flow from two flows that join at view top from the right and the left. The water from the one on our right is a different green color, and you can see the different flows staying distinct even with a wiggly boundary. The photo on the right is looking downstream at the Niobrara River in Nebraska, with surface patterns showing the many different flows across its width. Water fills a discontinuous channel by both diverging from a dominant flow direction, and by carrying more water with it as it seeks a retaining boundary. The role of friction seems to be zero. Obstructions At an Obstruction, water: a) does go around, b) can go under, eroding the bed in both cases, and c) may go over, and then resumes flowing through the available channel. At an instream rock there will commonly be a trench eroded upstream and beside. Sediment settles downstream as flow continues past, unless water falls over and so erodes as a pit in that general case. If water cannot flow around both sides of an obstruction, it will, after a way is eroded; and meanwhile will flow to refill the channel. The behavior at any object, in any location within any flow, will produce the same sort of results, continually. |
 |
 |
|
The photo at left is a view upstream at Big Hole R., MT. Natural boulder is deeply anchored in valley fill, smoothed by gradual abrasion of sand and gravel during higher flows during geologic time. Large bar deposit within very wide dispersed or "braided" channel is kept clear around this isolated obstruction, particularly at view left. It is this sort of natural facts that was uniquely the subject of this study: what is true, that can be known by simply looking. Photo at right is a view downstream on Norton Br., TN, where Robbin Sotir put very large imported boulders along banks throughout this part of project. Water is not "deflected away from bank" as promised, and paid for. Rather, carefully sloped and "vegetated" -- and very expensive -- bank is being rapidly removed by flow continuing around the boulder. See same result downstream at next boulder in view upper right. The scale does not change the basic effects: |
 |
 |
|
The photo at left is at Smith Cr., AR, with flow to upper left. Cobble-size rock on gravel bed shows typical continuous eroded trench around an object. In this case, there are not enough particles small enough relative to flow to form downstream drift deposit. The photo at right is at Long Cr., NC, with again flow to upper left. The larger object near view center is barely the size of a golf ball. All show typical bed erosion: undercutting at upstream face to view right, trench continuing past both sides, with drift deposit extending from downstream side to view left. Mudball to right has drift deposit flush with its flat top; middle mudball had water going over, see two depressions like form of its top; left pebble with curved top has middle height to deposit. Note little chunks gathered in all upstream pits, with finer grains stripped away by the same process. To prevent flow under a rock, Rosgen has moved towards excavating the bed to bury larger rocks as a base or "footer". Besides requiring larger equipment and more disruption in the bed, that also makes deeper trenching around a deeper obstruction, with more bed material moved and deposited somewhere. Sediment As the stuff of the earth for all of time and evolution, sediment cannot reasonably be considered a "pollutant". Calling it that allows endless make-work projects that never have stopped the natural behaviors of streams. Streams do not "transport sediment" like some freight service. Water can neither support particles like a truck, nor push them like a bulldozer. Streams do not exist for human purposes, and demanding that they do is more of the monstrous arrogance by which humans are dismantling the surface of this planet. Water does have mass and does move, and so can exert force; but what is required for water to move a particle is Turbulence, to counteract gravity. Once in motion, sediment behaves similarly to sand dunes and snow drifts. Bed ripples are transverse dune forms, with similar slope and slip face aspects and movement. The longitudinal central ridge deposits of storm flow proposed here, and the deposits downstream of a regular obstruction, are similar to seif or sword dune forms. |
 |
 |
|
Photo at left is at N. Buffalo Cr., NC, downstream view showing exposed results of a recent storm carrying a lot of sand for long enough to form these large-scale Transverse Dune forms. Photo at right is from Fannegusha Cr., MS, view downstream from bridge. The sharp ridge to the left would be a direct model of what are seif dunes in deserts; but it is partly just erosion collapsing the other side. The broader smooth long bar to the right is more typical of a drift deposit behind a bridge pier. But toward the end see small line of a thin sharp top of ridge, formed by equal flow on each side meeting across the top as a seif dune would be formed. This study has found no differences in forms of deposits by particle size or size distribution, and considers all studies of those details to be a waste and diversion of resources. This study has found the association of plants with deposits to be a matter of increasing or reducing erosive turbulence, showing nothing of "trapping" or "slowing the flow" as commonly claimed by others. |
 |
 |
|
The photo at left is a detail of wild plants on a bar in an urban stream, that are kept laid down by storm flow toward view right. Alignment is with the flow, not across flow as various Mats are placed on all "bioengineering" projects. Leaves are generally filling in as they fall, rather than being "trapped" in dense jams; and fresh sand is visible in view lower left. Any such deposits are protected by such smooth cover which prevents turbulence at the ground surface, contradicting all impositions of "roughness" by projects. The photo at right is at formerly Dry Walnut Cr., KS, engineered with flat, low-angle slopes for flood drainage. Four years after first being viewed, there has been almost no change. It is the smoothness of mown grass that is preventing the turbulence that would have started bank erosion. This study has noted Berm Deposits at all longitudinal edges, not just the "riparian levee" and "inner berm" studied by others. This study has seen nothing of a channel "aggrading" as a general or common condition, with bed erosion taken as a given of channel development. The confusion of the term "degrade" as both erosion of the stream bed and a judgement of a stream's condition generally, is deliberate, manipulated to justify endless stream projects. For this study, delta outwash and braiding are special cases of the flushing mentioned above with pit development, and not given special attention. The Meander Mechanism The center of a storm flow has been shown to flow fastest, and it carries the bulk of the sediment during the straight flow. This study finds that sediment is then deposited as a central ridge that is similarly straight. Because the sides of a channel are not going to be so straight, the Central Ridge Deposit will impound water where a bank may be nearest. During a flow higher than the ridge, there will be a breach that will wash out a flat ramp. The Breach Ramp then drains the Impoundment at lower flows. As that flow falls to the other side, it will erode something of a pit that will then develop to a trench along the base or "toe" of the bank, into which that bank may collapse. As gravity deepens the trench and the increasing water flow lengthens the trench, the Trench Flow becomes a more general channel feature. More of higher flows will be affected by this shift, driving further redirection and bank collapse in the manner called Meander. |
 |
 |
| The first photo is of Brushy Fk., NC, July 2000. Former straight canal shows long meander to view left with wild plants on previous bank collapse. Central Ridge Deposit is now a series of low bars running straight from view lower right toward view upper left. Flow crosses that ridge by Breach Ramp off view bottom right. Impoundment is long pool blocked by ridge above view center, see next Breach Ramp to right of view center. Above view center is another meander bank on other side of central ridge deposits, with small Breach Ramp to be shown following. Photo at right is view downstream July 2003, from left "bench" or level area inside channel. Straight central bed ridge shows more plants on more deposits at view center. Note also that breach upstream, in view bottom right, is now a longer Shoals angled to a left bank collapse extending downstream (see dark shadow), a later stage of development of same scenario. Next is same, five months later, from top of bank. |
 |
 |
|
View downstream Dec. 2003 at left is again from left terrace. Central Ridge is now more massive, and right deposit is looking more like a "point bar" as meander swings away. Shine on ramp flow shows longer crossing form too, and note more bank collapse downstream. Yet all three bar deposits are still in same basic line of the Central Ridge (view lower right is only tail part). Photo at right is view downstream April 2005. Ridge deposit to view right is now rounded and sloped as a Point Bar and blending more with upper terrace. Impoundment is more a continuous and curved or "sinuous" channel, note natural slope of high bank at view left. Main storm flow, however, is still straight, shown by larger rocks going over bar to view lower right. The Breach Ramp has acquired the character that has been called "riffle", followed by the Trench Flow which has been complimented as a "pool". Or, is it a "pool" followed by a "riffle"? The next breach developed in a similar way, showing that this is a general mechanism of channel formation: |
 |
 |
| Photo at left is a view downstream at next Breach Ramp July 2000. Where the Central Ridge approaches bank, as seen in view bottom left, flow is an Impoundment, see deeper water along view bottom. Active breach of ridge and the ramp to other side is at view center. Plants at view top are on previous bank collapse material or "slump": note that they extend to waterline. Photo at right at same location Dec. 2002 shows further deposits along persisting central ridge, see humps of sand to both sides of same breach. Slope of ramp is steeper too, with pit typical of such fall now eroding as longer trench along base or "toe". Right bank is collapsing upstream toward this flow, note vertical face in view upper left with no slope or plants at base. |
 |
 |
|
Next view downstream July 2003 shows central ridge has more deposits across previous breach, with another weak breach at view left, and more plants. Ramp flow leads to deeper and longer pit at view center. Right bank in view upper left has collapsed farther back, see typical Slump Slope now at base, as meander follows breach flow. Photo at right is view downstream April 2005, showing more bank collapse upstream at view right. Ramp is more of a shoals followed by the Trench Flow as a smooth curve along continuous bank collapse, with no plants even on slump toe to view upper left. The next two photos show same meander and ramp in May 2010: |
 |
 |
|
The photo at left shows the high collapse bank moved farther out in the middle portion, in line with current ramp and not "migrating downstream" as is taught as typical. There is a pit at the end of the incoming ramp flow in view lower left, while to view right the point bar is higher and growing willows at the downstream end. Photo at right can be compared to photo above it to see that the ramp has developed into a straight riffle section, and the next meander bank has collapsed a lot more, see tree roots draped over the face. The breaching of a ridge is essentially random in development. However, the dominant crossings will typically develop as an oscillation of flow direction, with alternating Point Bars as remnants of the central ridge, continually modified by further deposits and erosion. The general storm channel and previous central ridge deposits are shown by "connecting the dots" of the later alternating point bars: |
 |
 |
| Photo at left is a view downstream April 2005 to the ramp and meander shown in first set above. Fourth photo following will be to this place; fifth photo following will be from this place. Viewpoint is near center of previous canal, with three significant bar deposits in that same central line. To right of view center, see larger gravel and rocks going up the deposit along that same line. In view bottom right is an Impoundment upstream, followed by this Breach Ramp across Central Ridge Deposit, to next Impoundment as seen in first set above. Photo at right is on same date from top of bar seen in view center. Line of same ridge is crossed by next ramp as at second set above. Note hump of continued central deposits below view center, with larger gravel and rocks still in line of previous central flow, and then on next part at view center. Next is continuing downstream: |
 |
 |
|
In photo at left, the end of Central Ridge Deposit is at view center, with previous canal continuing fairly straight downstream. Breach across lower view, making next meander to the left, is allowing another bar at view center to rise with further deposits to blend with high terrace to view right. The tree debris at view right is from continuing major bank collapse at the downstream part of that meader. Photo at right is a view upstream to the viewpoint seen above for this set. The smoothly rounded form here is a typical Berm Deposit on this older bench. Because this meander contains most of the volume of most storm flows, it can now dominate flow tendencies to become independent of the previous general channel. The former central ridge here is becoming a mostly level area or "bench" acting as a minor "flood plain". |
 |
 |
|
The photo at left is from same vantage as first of this set, showing further development in May 2010. Point bar is higher, blending more with original terrace level to view right, and growing more woody plants so that further view is now obscured along with the sense of the former straight canal. But note that the crossing ramp is breaking downward still in same line of the Central Ridge Deposit as the larger rocks rising over the bar, carried by same straight path of storm flows moving those rocks. Photo at right shows detail of the "size distribution" of the bar deposit, a standard of stream assessment for justifying projects, but ignored for what it might have taught about general channel development. The older a given channel, the more erosion and deposit and bank collapse are overlaid, with point bars rising, sloping, and growing plants to blend with the upper flood terrace. It is in this way that the essential process is afterwards hidden, as if "sinuosity" were a given of stream flow itself -- and anything else was an error that could be corrected by building the "correct" channel per Rosgen. But, this process never stops to become "stable" as proposed (but never shown) by Luna Leopold and adopted as the pupose of rebuilding channels by Rosgen and his thousands of disciples. This study has found no evidence of, nor any reason to believe that there would be, any limit to the meander process. Meanders commonly extend at a rate and to a degree regulated only by the frequency and force of flows, and the resistance of materials encountered. |
 |
 |
| Photo at left is on Smith Cr., AR, view downstream, showing typical case of long central bar crossed by flow at Breach Ramp to Trench along collapsing bank. The layers exposed on the bank show previous geologic valley fill of solidly packed rip-rap-like rocks that are offering no deterrent to erosion, while supplying the bed material. Photo at right is on Thomas Fk., ID, view downstream. Flow shine shows direction straight into ridge with a turn to view right. There is a typical pit at the end of the Ramp flow, shown by green deeper water extending as Trench to view right, and the bank collapses into it as storm flows drive the pit deeper and farther. High bench to view right is growing longer and higher, and more resistant with more well-watered plants, as meander continues. |
 |
 |
|
Photo at left is from So. Table Cr., NE, view downstream. Flat bar with plants in view lower left shows long development of nearly 180° reverse loop, with a higher ridge offering no limitation at any time. Material is shale, again holding a nearly vertical face. Photo at right is on Lance Cr., SD, view upstream to a larger "gooseneck" reverse turn. Height of this eroding ridge is about 100' (30 m), with no apparent limit to its continued collapse. Note that water runs clear here and at all such banks, with sediment incorporated into normal aquatic life including appropriate fish. High banks like these are taken as the primary excuse for requesting and approving stream projects, based on a false assumption that there must be "pollution" and "degradation" of the stream from the resulting sediment. There is only the inevitable development of the stream's channel, as has been true for all of living time. When Rosgen and others "improve" a stream by building curves and ramps and pits, they are installing only an un-natural version of the Meander Mechanism that will remove their artificial works. This study has not found any consistency in how the material of a bank falls, finding such a range of "block" and "rotation" and shatter and ravel, as to call all of it Slump. What happens to what may fall is also variable, with a Slump Slope commonly forming at the base and quickly growing wild plants starting with sod from the terrace or previous growth on the slope. |
 |
 |
|
Photo at left is from Panther Cr., IL, view upstream. In wetter climate, more of the bank falls when it falls, and grows more plants faster. Slump Slope toward view left is making a covered buttress, much harder to erode and usually gathering fine sediment as Berm Deposit. The angle of such slopes is remarkably uniform across occurrences, about 1.5 : 1 as common to Civil Engineering, while completely variable per occurrence so that where and how it is measured would predetermine the result, making such measurements meaningless. Photo to right is downstream on same stream, view downsteam, with same sort of slump at view left, and to view right. The much steeper area at the highest part is also growing some plants, despite being the most actively removed part, and was basically the same the following year. Note that the overhanging sod cap at the top is the result of "dry ravel" as would happen at a similar road bank, not related to the stream. Besides being entirely natural and inevitable, such vertical banks are the necessary condition for nesting by Belted Kingfisher, birds native to the entire continental U.S. that feed mainly on small fish. The presence of such banks is so common and dependable that this species has evolved with physical as well as behavioral modifications to allow it to tunnel into such features for nesting: |
 |
 |
| Photo at left is an old Cutbank on Otter Cr., MT. The Slump Slope is long since stabilized with plants. Scarp shows only "rill" erosion, typical of any exposed face. Hole at top above view center is nest hole for Belted Kingfisher. If it is nesting there, the stream must be very healthy to support complete aquatic ecology with abundant and constant fish supply. In photo to right, see young birds near entrance for feeding, and close to fledging. Tunnels are typically two feet or longer, often sloped upward to prevent flooding. Note that Kingfisher hole is slightly widened with characteristic groove to each side, shaped like an eye. |
 |
 |
|
Photo at left is on So. Fk. Mitchell R., NC, Oct. 2000. Kingfisher nest hole is again near top, in shaded area above view center. Typical Slump Slope at bottom is recent, see edge of cleavage to view left and previous slope washed away to view right. Typical of thousands of stream projects every year, this bank was main excuse for a complete Rosgen-style reconstruction, shown in progress in photo at right also in Oct. 2000, leaving no such nesting habitat. The first photo at bae.ncsu.edu to fraudulently promote a different project, is also this bank; the person in orange vest here is approaching to confiscate the camera to prevent any information other than such promotion by NRCS, which has made a major business of such projects and commonly prevents viewing by others. There is another bird and other species that are evolved to use Cutbanks: |
 |
 |
| Photo at left, taken at same site and time as above, shows a smaller hole near the top for a smaller bird: Northern Rough-wing Swallow. It does not feed on fish, nor need to find its necessary nest site at a stream, yet commonly does because such situations are so common in nature. Its hole is round, with a flat ramp worn at bottom; and this nesting site was destroyed by same NRCS project. The photo at right shows at view center a nest entrance for probably a Muskrat, which will dig such tunnels where a water lodge is not appropriate. |
 |
 |
|
Photo at left shows nest hole definitely for River Otter family that were seen at the time, taking advantage of a minor bank collapse with the typical slump slope at base. Photo at right shows a fish nest at a collapsing meander bank on a previously channelized urban stream. A small Red-breast Sunfish has made its nest in the silty bed there, and was seen aggressively patrolling the area. This species is considered to have "intermediate tolerance" of poor water quality, in keeping with the rating for the entire basin. Blaming such cutbanks for the immense amount of silt from the disturbed watershed is another immense lie that has enabled endless stream reconstruction projects, none of which change to any degree the watershed characteristics. The uniform flat banks commonly imposed by stream reconstructions have not been found by this study as a natural result of any stream development in any valley material. While Slump Slopes with plant cover are often as "stable" as stream banks can be, this study has found no convincing evidence that trees or any other roots "hold banks". The real question would be whether a bank will hold the roots. |
 |
 |
|
 |
 |
|
The above four photos are from various places, all showing that bank material will erode away from roots of any size and density. The two leaning trees are on different streams, yet show very similar stages of falling, as both surely will. When they have fallen into the flow, they will form a variety of obstructions and consequent erosion, increasing with collecting other debris into a jam. Most projects will have stripped all trees from banks to excavate uniform slopes which are supposed to be "stabilized" by smaller plants. Here is an attempt by Rosgen to use grass cover on a constructed "bankfull bench": |
 |
 |
|
This is the same project on Lake Fork shown at home webpage, view upstream. As the steep toe at eroding trench flow is washed away, sod drapes over in a deceptively smooth, curved roll as in view upper right, then breaks and becomes the next vertical face. Bare dirt of bank at view left is current collapse, dark shadow below waterline is undercut making the next. Note high dense grasses on built terrace, and transplanted shrubs intended to make a thicket. Photo at right, view downstream, shows detail of typical slump crack at that same bank. This condition, typical of any bank loss, was not prevented by Rosgen's rock Vanes, nor by thick turf on both sides of crack, with dense stout roots visible on dirt face. This study has found almost no occurrence, on streams of the scale directly accessible to this study, of an "oxbow pond" form resulting from excessive meandering. Rosgen introduced that notion as a source for fill for reconstructing channels, and an excuse for doing something else costly and requiring huge machinery. Entrenchment It is tempting to include entrenchment as an important part or even as the major factor in channel development. There does seem to be a rate relationship, so that more entrenchment means more channel erosion. However, that is circular: the more entrenchment, the more water is inside the channel doing what water does. The stream in a canyon in fact shows the same general characteristics as other streams. At present, entrenchment is left as an interesting variance, rather than a variable, in the general process of channel formation. Sinuosity Water flow does not produce curves. The curves we think we see are perceptual simplifications of the sort called "gestalt" in psychology. Any closer examination of any apparently regular form at any starting scale will reveal perturbations that increase with closer examination; yet at each scale, one will continue to simplify the conditions to averaged directions and boundaries to again "see" curves. There is a math for producing such curves from such scattered actual locations, proving only the power of math to distract, and it has been deliberately ignored by this study of real water flow. Streams have no necessary direction other than down, as commonly running straight as curved, and with meanders as often proceeding upstream as downstream. It is only the continuing tendency of storm flow to go straight that drives erosion farther to a given side away from a previous or general direction, and from which it would return by the same tendency, moving always and only however down could be. |
 |
 |
|
Photo at left is a pasture gully near Oakland, IA. This would be called "ephemeral" because it carries only rain etc. runoff. The path would have been the grass swale between the slopes, then a rill or bare track, then the wandering or "meandering" gully that is growing deeper and wider and more kinked. At view center is a loop about 180° with typical steep concave bare sides at outside of turn. Note very narrow remnant of terrace at that switchback. Lower angle of straight slopes below view center is typical of older slump with plants now growing along previous path. Present path is random and is now independent of land forms, being able to erode any mass in its way, and entirely similar to other streams of any larger scale, with nothing anywhere of regular curves or distances between turns as postulated by Leopold and demanded by Rosgen. Photo at right is view upstream to the Salmon R., ID. No particular materials or external forms are determining the path of this naturally straight part of a river. Nobody put it there, and it's been there at least as long as the large trees on the true floodplain. OCCURRENCE The Meander Mechanism process and other aspects of the theory have been found to occur at all scales in all movable materials. No fundamental differences in the nature of stream forms have been found to result from valley or bed slope or proportions or constituent materials. Forms are found to be so similar across such a range of conditions as to demand a theory of general mechanisms, rather than a theory based on "classification". The photos presented currently at this website show both the development of the theory and various aspects of the theory, and together serve as substantiation of the theory by offering real examples. FURTHER DISCUSSION The theory presented here not only does not support other theories; it could be argued to refute the bulk of "fluvial geomorphology" and the studies and practices deriving from it, particularly those of Rosgen and thereby the bulk of stream projects worldwide. However, it is the purpose of this website and this treatise to report, with the hope of teaching at least something of observation skills. Luna Leopold had imagined and propounded a mysterious force arising from the flow of water that produces a mathematically expressible curvature and oscillation of direction that will tend toward a stable channel form. This study has seen that the only direction water will reliably go is down, and will continue in a direction until turned by gravity or an obstruction. Leopold argued from "dimensionless graphs" that are therefore about nothing. What he actually wrote was that he had devised a way to calculate an average sinuosity of "five to seven bankfull widths" from an actual range of two to eleven such ratios. This study has not found, on the ground or on maps and aerial photos, any such sequence of even two consecutive turns, or even such "sine-generated" turns. His romantic vision and mathematicalizations are supplanted here simply by showing the actual, practical processes. Luna himself denied ever intending that his theory be used to control streams, as Rosgen has made millions claiming to do; and Leopold doubted that it would do so. For that last point, he was quite correct. Leopold developed the notion of a "bankfull" condition that was the essential cause and therefore reference for any stream's form. This study has seen nothing of such a consistent aspect of any stream. Any amount of water in any stream-like circumstance is a "channel-forming flow" at that scale; and the channel conditions encountered by any flow will affect that flow and so its effect on the channel. That recursive loop is the source of meanders, meaning instability, not "balance". Whether there is a bank for some flow to exceed, and even if there is a distinct bank and floodplain, is a variable at every part of every stream and for every flow. Many people have been taught to hunt for and find some appearance that they are taught to call "bankfull indicators"; but that relates only to the trainability of those people. David Rosgen's derivative proposal of a calculated and artificially produced "natural stream channel form" that would therefore be "stable" has nowhere been found by this study to be supported by his own and his followers' or anyone's stream projects. The curves and banks produced do not naturally occur; and no such forms as the "structures" are ever produced by streams. Rather, all such stream reconstructions studied exhibit the Meander Mechanism and other aspects of this theory in definite and even rapid developments tending to obliterate those unnatural projects. This study has found no such continual line of flow as the fanciful "thalweg", in nature or in what remains of such designed projects. Please note that the so-called "riffle-pool sequence" is not the same as this theory. Those commonly recognized, but obviously not well observed, forms are taught as objects that can be installed at will and by means of machinery. It is not possible for machines to do, or make, anything natural. A true riffle is a longitudinal bed form that sometimes occurs with higher rock content. They are uncommon, because that is not what water flow commonly does with movable materials. Similarly, a true instream pool is a continually produced pit or trench that is continually producing deposits downstream as shown above. The breach ramps and shoals of this theory occur with widely varying location and character, with associated erosion also widely variable. For the Meander Mechanism theory, "crossing" does not refer to the long straight between turns, which is only a possible development of the channel after a crossing of the central ridge deposit as described here. The Meander Mechanism theory proposes general processes that contradict notions of "stream classification". The extensive and intensive observations of this study would not have suggested and do not support Rosgen's proposed discontinuity of forms at some abstract percent of bed slope. Nothing in the natural setting of streams is either abstract or dependably reproduced. Channel forms such as Rosgen's E have been seen adjacent to A, with a piece of D in an otherwise B, with the general mechanisms summarized here occurring throughout. Easily "communicating" what is false, is not helpful. The insignificance of particle size has been mentioned above. This study does not find "width-depth ratio" to be a real or useful factor. Rather, it contradicts the notion of "U-shaped" or "concave cross section" as being the way bed erosion can occur, with flat beds seen so often and so widely as to be assumed as the normal condition. Neither has any sense to "bankfull" been found in the highly varying channels and flows seen everywhere outside a few meadows. Any number of flow channels can be found nested within a general channel, increasing with entrenchment and also with total width, with each such flow following the general factors of this theory. How some magic could be found in one such channel to "stabilize" the results of all other flows in the general channel, has not been found by this study. The notion of a "reference reach" defined by abstract notions rather than any observations and then to be copied, is as absurd as regularly recurring "rain events". The notion of "roughness", used to justify any sort of imposition on a channel, is basically backwards. Any source of turbulence, as all roughness will make in some flow, produces the erosion that people do not want. Neither has it been seen possible to simply "slow the flow" without therefore impounding that flow and so producing further, and generally undesirable, consequences. There must be some material that can be moved by available flows for there to be erosion; and any material being moved that cannot be moved in another circumstance will then be deposited. But this study has found no connection between these two typical conditions. There is no "balance" necessary within a given channel. The only natural places that have revealed any longevity to this study are bedrock with extensive plants on the sides. If roots did "hold the banks", then there would be only burbling brooks shaded by those plants. This study finds that any upright or resisting element, plant or not, poses an obstruction and such developments as noted above. Where plants offer a smooth surface to storm flow, particularly grass and forbs laid flat and parallel to the flow, the resulting lack of turbulence is all the "protection" that has been observed. Deposit does not rely on this condition; but any that occurs is also so protected, and so can increase, giving the false impression of causing deposits. This study has found no constructed banks to be more stable than rock revetment allowed (which is rare) to silt in and grow plants and so becoming smooth; and plain grassed slopes, as shown above, have been seen as stable as those with concrete blocks in them. Of the various structures placed in beds with claims of "redirecting the flow", none have been seen to convincingly do so as claimed. Water flow does not bounce, and so cannot be "deflected"; and the redirection that does occur, as described above, usually erodes the structure. The only direction water has when falling over anything, is down; and what it does at whatever bottom, is erode that bottom. After that, water can only find a way through the resulting turbulence to again flow along the channel. It is not possible to compress water nor keep it from filling the available channel; so it is not likely that water could be "redirected to the center of the channel" as claimed for various disrupting structures. Any increase in rate of flow by constriction at a structure has been seen to produce greater bed erosion, generating the deposit scenarios as above, as well as wrecking the structure. This study has found no evidence of anything happening in streams as if by the action of scissors, indicating that "shear forces" are sheer fantasy. The practice of "sizing the rock" is based only on repeating that practice, not on any action of water observed by this study. Nor has any sign of a helical flow at turns been found by this study, though upwelling has been seen during storm flows. For this study, the term "water column" would mean something vertical and realistically apply only to recurring local upwelling, usually at a deep bed obstruction; so any use of that term about streams is an alarm for ignorance of reality and for confused thinking. CONCLUSION The Meander Mechanism theory holds that it is the bed, not the banks, that generates stream channels, so that what has been done to banks has been at best irrelevant to any purposes of control, and what is done to beds is only, and increasingly violently per Rosgen, disruptive. The general summary of this study is: Return to TOP of essay. Return to home-page for Alternative Stream Research with illustrated discussion of David Rosgen. Copyrights by and Comments to: Lloyd Ramsey, Alternative Stream Research 4948 Old Baux Mtn. Rd., Winston-Salem NC 27105 |