River oddities

    Actually, it is quite odd to find a mountain river moving faster than a valley river or plains river.  The research has shown that a river -- at a comparable stage -- has the same velocity from its high mountain avatar to the sea.  If the stream is bankful in the mountains, in the valley, and in the plain, all reaches will have about the same velocity.  This is because its Manning n (a dimensionless number for flow resistance) decreases with increasing cross-sectional area, decreases with increasing depth, decreases with diminishing size of bed material, and decreases with increasing discharge -- all of which occur with distance from the mountain crests.  High gradient in the mountains is not sufficient impetus to overcome small absolute discharge, large rough bed material, ect., so the mountain stream is not flowing faster than the valley stream.  On the other hand, it _looks_ faster because there is a lot of splash and play.  That splash comes at the transitions from sub to critical to supercritical flow (Froude number here), when the water releases its excess energy in hydraulic jumps. Time a stick in a mountain creek, then a piece of driftwood on a big river.  The highest open-channel velocity ever recorded was in the gorge of the Potomac and was, as I recall, 22 fps.

    Mountain rivers are tributary to valley to plains rivers, and carry more water only if

  1. They have responded to a storm cell or snowmelt transient and there has not been time for the lower stream reaches to respond; or
  2. If the lower streams are dewatered for agriculture, typically at the mountain front; or
  3. There is an anomaly at the mountain front, such as a highly porous bajada.
    In their undeveloped state mountain rivers don't have greater depth than valley rivers; this would be very odd indeed unless perhaps in a geologic anomaly, such as karst.  As their gradient diminishes with approach to the sea, given that velocity is a relative constant, the cross-sectional area of rivers must increase to transmit the increasing discharge provided by high-order tributaries.  Their depth steadily increases, and their width increases by about 0.6 with the accretion of each tributary which is roughly equal in discharge to the parent stream.

    Leopold Wolman and Miller _Fluvial Processes in Geomorphology_ is a doubly improbable classic -- a textbook, and one that has actually been reprinted without revision in a quickly obsoleting field.  Leopold has recently published a book on rivers that is eminently readable, contains results of simple genius (the sum of the powers of width, depth, and velocity must equal 1), and is horribly disfigured by terrible copy-editing; a shame to Harvard University Press.

    I'll tell you of what I think is a river oddity.  When on a canyon river you get an encroachment by a side-stream debris-flow or 'dam,' which narrows the main river to <= 0.25w, where w is the width above the constriction, then when the next bankful stage or higher occurs (say, recurrence > 1.5 years on an undeveloped stream) then the river at the constriction forms a normal wave, directly across the current. Velocities increase to the point where bed material greater than 2.5m in diameter can be mobilized.  The normal wave forms without regard to subsurface rocks, apparently solely in response to the constriction.  See Susan Keifer Warner's (Werner?) work in Grand Canyon.

Back to Index Page