18
between adjacent fluid laminae. Turbulent flow in
open channels, pipes, and in air was investigated,
and the change from laminar to turbulent flow with
increased velocity was documented with the use of
the Reynolds number (Reynold's classic work in this
field dates back to 1883). In 1936, Shields provided
the first satisfactory theory regarding the initiation
of bedload movement on a flat bed, in which the
bottom shear stress or shear velocity was shown to be
the critical parameter. Hjulstrom (1935) used average velocity to predict the critical conditions for the
beginning of sediment movement. This approach
has been used recently in paleohydraulic reconstructions, but is less accurate. Rubey ( 1938) reexamined
Gilbert's (1914) flume data in the light of these theoretical developments, although he d�es nOt seem to
have been aware of Shield's (1936) contributions.
Rubey was able to explain reasonably well the movement of material of medium sand to pebble grade,
but stated that "a satisfactory theory of the force
required to start movement of fine sand and silt
must await additional observations''.
Theoretical developments on the calculation of
total load are incomplete. Rouse (1939), and later
Vanoni (1946), used the concept of momentum diffusion to predict concentrations of suspended sedi�
ment. The earliest successful work on bedload
tranportation was that of Du Boys (1879), who deduced that the rate of movement of bed load is
proportional to the product of the shear stress and
the difference between shear stress and the critical
shear stress for the initiation of movement. This
physical model has been widely used. However, it
fails to take into account bed roughness (grain
roughness and form roughness - that induced by the
presence of bedforms), a problem first tackled by
Einstein (1950).
Developments in fluvial hydraulics up to the mid1950s were summarized by Leliavsky (1955). Geologists made little use of this field 'until Sundberg's
(1956) major paper on the River Klaralven was published. In this publication, Sundberg applied available theory regarding flow dynamics, erosion, and
entrainment of sediment to a study of the bedload
and suspended load of a specific river. The geomorphology of the river, and the texture and structure of
its deposits were also described in great detail. The
paper thus represented a major bridge between the
disciplines of hydraulics, geomorphology, and geology, and it was widely quoted during the 1960s and
1970s.
This brief summary has attempted to cover most
of the more fundamental developments in the theory
Historical Background
of sediment transport. A more complete discussion
is beyond the scope of this chapter, and the reader is
referred to Bush and to Briggs and Middleton (in
Middleton 1965) for further historical details. Recent
reviews of the topic have been given by Church and
Gilbert (1975), Middleton and Southard ( 1977), Shen
(1978), and Fielding (1993b).
Subaerial debris flows contribute much coarse,
poorly sorted sediment to alluvial fan surfaces, as
first documented in detail by Blackwelder (1928). He
referred to these catastrophic events as mudflows,
but data compiled by Bull (1964) and Lustig (1965)
showed that their actual mud content may be less
than 10%. Observations on modern floods by
Chawner (1935) and Sharp and Nobles (1953) demonstrated the transporting power of debris flows
and the conditions under which such events occur,
notably infrequent but torrential rainfall on unvegetated upland regions, where sufficient time has been
allowed for the accumulation of a mass of loose
debris. More recent studies of the mechanics of debris flows have been published by Middleton and
Hampton (1969), Statham (1976), and Rodine and
johnson (1976). Other special processes operating
on alluvial fans, including the tendency of flows to
lose competence by inflltration, leaving «sieve de�
posits", were described by Hooke (1967).
Textural studies, that is, investigations of grainsize parameters (such as mean, sorting, skewness,
and kurtosis) and other grain description·s, includ�
ing roundness, shape, and sphericity, have long fascinated clastic sedimentologists. Since the work of
Udden (1914), the hope has been that simple laboratory measurements on a suite of samples (better still,
a single sample) would reveal their depositional environment, and the literature on grain-size analysis
is now vast. Modern work on the subject began with
the grain-size scale for pebble-. to day-sized material
erected by Wentworth (1922; based on Udden 1914)
and the investigation of the logarithmic phi scale
for grain-size description by Krumbein (1934). Attempts to use sorting parameters as indicators of
mode of deposition began with the work of Inman
(1949, 1952), and Folk and Ward (1957) proposed
methods for describing the sorting of samples of
clastic rocks using percentile values of various grainsize intervals derived from the graphs of the cumula�
tive grain-size distributions. These methods were
reviewed by Folk (1966). Many attempts have been
made to define the sorting characteristics of sand
deposits formed in various environments using
graphic (Passega 1957; Friedman 1961, 1967; Spencer
1963; Visher 1965a, 1969) and statistical (Sabu 1964;
between adjacent fluid laminae. Turbulent flow in
open channels, pipes, and in air was investigated,
and the change from laminar to turbulent flow with
increased velocity was documented with the use of
the Reynolds number (Reynold's classic work in this
field dates back to 1883). In 1936, Shields provided
the first satisfactory theory regarding the initiation
of bedload movement on a flat bed, in which the
bottom shear stress or shear velocity was shown to be
the critical parameter. Hjulstrom (1935) used average velocity to predict the critical conditions for the
beginning of sediment movement. This approach
has been used recently in paleohydraulic reconstructions, but is less accurate. Rubey ( 1938) reexamined
Gilbert's (1914) flume data in the light of these theoretical developments, although he d�es nOt seem to
have been aware of Shield's (1936) contributions.
Rubey was able to explain reasonably well the movement of material of medium sand to pebble grade,
but stated that "a satisfactory theory of the force
required to start movement of fine sand and silt
must await additional observations''.
Theoretical developments on the calculation of
total load are incomplete. Rouse (1939), and later
Vanoni (1946), used the concept of momentum diffusion to predict concentrations of suspended sedi�
ment. The earliest successful work on bedload
tranportation was that of Du Boys (1879), who deduced that the rate of movement of bed load is
proportional to the product of the shear stress and
the difference between shear stress and the critical
shear stress for the initiation of movement. This
physical model has been widely used. However, it
fails to take into account bed roughness (grain
roughness and form roughness - that induced by the
presence of bedforms), a problem first tackled by
Einstein (1950).
Developments in fluvial hydraulics up to the mid1950s were summarized by Leliavsky (1955). Geologists made little use of this field 'until Sundberg's
(1956) major paper on the River Klaralven was published. In this publication, Sundberg applied available theory regarding flow dynamics, erosion, and
entrainment of sediment to a study of the bedload
and suspended load of a specific river. The geomorphology of the river, and the texture and structure of
its deposits were also described in great detail. The
paper thus represented a major bridge between the
disciplines of hydraulics, geomorphology, and geology, and it was widely quoted during the 1960s and
1970s.
This brief summary has attempted to cover most
of the more fundamental developments in the theory
Historical Background
of sediment transport. A more complete discussion
is beyond the scope of this chapter, and the reader is
referred to Bush and to Briggs and Middleton (in
Middleton 1965) for further historical details. Recent
reviews of the topic have been given by Church and
Gilbert (1975), Middleton and Southard ( 1977), Shen
(1978), and Fielding (1993b).
Subaerial debris flows contribute much coarse,
poorly sorted sediment to alluvial fan surfaces, as
first documented in detail by Blackwelder (1928). He
referred to these catastrophic events as mudflows,
but data compiled by Bull (1964) and Lustig (1965)
showed that their actual mud content may be less
than 10%. Observations on modern floods by
Chawner (1935) and Sharp and Nobles (1953) demonstrated the transporting power of debris flows
and the conditions under which such events occur,
notably infrequent but torrential rainfall on unvegetated upland regions, where sufficient time has been
allowed for the accumulation of a mass of loose
debris. More recent studies of the mechanics of debris flows have been published by Middleton and
Hampton (1969), Statham (1976), and Rodine and
johnson (1976). Other special processes operating
on alluvial fans, including the tendency of flows to
lose competence by inflltration, leaving «sieve de�
posits", were described by Hooke (1967).
Textural studies, that is, investigations of grainsize parameters (such as mean, sorting, skewness,
and kurtosis) and other grain description·s, includ�
ing roundness, shape, and sphericity, have long fascinated clastic sedimentologists. Since the work of
Udden (1914), the hope has been that simple laboratory measurements on a suite of samples (better still,
a single sample) would reveal their depositional environment, and the literature on grain-size analysis
is now vast. Modern work on the subject began with
the grain-size scale for pebble-. to day-sized material
erected by Wentworth (1922; based on Udden 1914)
and the investigation of the logarithmic phi scale
for grain-size description by Krumbein (1934). Attempts to use sorting parameters as indicators of
mode of deposition began with the work of Inman
(1949, 1952), and Folk and Ward (1957) proposed
methods for describing the sorting of samples of
clastic rocks using percentile values of various grainsize intervals derived from the graphs of the cumula�
tive grain-size distributions. These methods were
reviewed by Folk (1966). Many attempts have been
made to define the sorting characteristics of sand
deposits formed in various environments using
graphic (Passega 1957; Friedman 1961, 1967; Spencer
1963; Visher 1965a, 1969) and statistical (Sabu 1964;