Channel and island change in the lower Platte River, Eastern Nebraska, USA: 1855­2005
R.M. Joeckel a,
, G.M. Henebry b
a
Conservation and Survey Division, School of Natural Resources and Department of Geosciences, 615 Hardin Hall, University of Nebraska-Lincoln, 68588-0996, USA
b
Geographic Information Science Center of Excellence, Wecota Hall, 506B, Brookings, SD 57007-3510, USA
A B S T R A C TA R T I C L E I N F O
Article history:
Received 6 August 2007
Received in revised form 22 April 2008
Accepted 22 April 2008
Available online 8 May 2008
Keywords:
River planform change
Platte River
Nebraska
Island stabilization
Anabranches
The lower Platte River has undergone considerable change in channel and bar characteristics since the mid-1850s
in four 20­25 km-long study stretches. The same net effect of historical channel shrinkage that was detected
upstream from Grand Island, Nebraska, can also be detected in the lower river but differences in the behaviors of
study stretches upstream and downstream from major tributaries are striking. The least relative decrease
occurred downstream from the Loup River confluence, and the stretch downstream from the Elkhorn River
confluence actually showed an increase in channel area during the 1940s. Bank erosion was also greater
downstream of the tributaries between ca. 1860 and 1938/1941, particularly in stretch RG, which showed more
lateral migration. The cumulative island area and the ratio of island area to channel area relative to the 1938/1941
baseline data showed comparatively great fluctuations in median island size in both downstream stretches. The
erratic behavior of island size distributions over time indicates that large islands were accreted to the banks at
different times, and that some small, newly-stabilized islands were episodically "flushed" out of the system. In the
upstream stretches the stabilization of mobile bars to create new, small islands had a more consistent impact over
time. Channel decrease by the abandonment of large, long-lived anabranches and by the in-place narrowing
resulting from island accretion were more prominent in these upstream stretches. Across all of the study area,
channel area appears to be stabilizing gradually as the rate of decrease lessens. This trend began earliest in stretch
RG inthe late 1950s and was accompanied byshifts inthe size distributions of stabilized islands inthatstretchinto
the 1960s. Elsewhere, even in the easternmost study stretch, stabilizing was occurring by the late 1960s, the same
time frame documented by investigations of the Platte system upstream of the study area. Comprehensive
management plans for the lower Platte River should account, at least in theory, for the observed differences in
stream behavior upstream and downstream of the major eastern tributaries.
 2008 Elsevier B.V. All rights reserved.
1. Introduction
At the beginning of the twenty-first century, land, water, and
wildlife management issues in the Platte River Basin (~220000 km2
)
clearly have major significance in the central United States and in
North America as a whole (National Academy of Sciences, 2004;
Condon, 2005). An understanding of geomorphic change along the
river should be a critical aspect of integrated understanding of the
Platte River ecosystem. Historical geomorphic change is particularly
relevant in the case of the Platte River because of emerging concerns
about threatened and endangered species, whose life histories are
intimately tied to the physical condition of the river's bed, mobile bars,
and banks (National Academy of Sciences, 2004). Long before the
strongest concern about habitat change on the Platte River emerged,
benchmark studies of historical change (Williams, 1978; Eschner,
1983; Eschner et al., 1983; Kircher and Karlinger, 1983) documented
dramatic channel narrowing, the stabilization of mobile bars and
banks by woody vegetation, and extensive island-to-bank accretion
since the 1860s. These studies focused on the river west of Grand
Island, Nebraska, and on the tributaries North and South Platte River
even farther west, relying on serial measurements of channel width at
single stations. More recent studies of the Platte employed improved
techniques (including examinations of change in channel area using
GIS) and greatly enhanced the understanding of riparian dynamics
and recent environmental history (e.g., Johnson, 1994, 1997; Currier
and Davis, 2000; Johnson and Boettcher, 2000), but they still tended to
concentrate on short stretches of the central to western Platte River.
Despite the weight of previous studies, the historical dynamics of the
Platte River remain far from completely characterized. Our study, in
contrast to earlier ones, concentrates on the lower Platte River (Central
City to near Plattsmouth, NE: Fig. 1) and employs geographic information
systems (GIS) technology to generate area measurements of both
channel and stabilized in-channel island area in long (~20­25 km)
stretches of the river (Figs. 1­3). The reasons for concentrating on the
lower Platte River are fourfold. First, the understanding of historical
change in that area is very incomplete. Second, patterns of change can be
contrasted upstream and downstream of two major tributaries, the Loup
and Elkhorn Rivers (Fig.1), which together have discharges equivalent to
Geomorphology 102 (2008) 407­418
 Corresponding author. Tel.: +1 402 472 7520; fax: +1 402 472 4608.
E-mail addresses: rjoeckel3@unl.edu (R.M. Joeckel), Geoffrey.Henebry@sdstate.edu
(G.M. Henebry).
0169-555X/$ ­ see front matter  2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.geomorph.2008.04.016
Contents lists available at ScienceDirect
Geomorphology
journal homepage: www.elsevier.com/locate/geomorph
about 30% of the total discharge of the Platte at its mouth. Third, the
study area is environmentally distinct from central and western
Nebraska. For example, there is maximum difference in mean annual
precipitation of nearly 200% (415 mm at Scottsbluff versus 788 mm at
Plattsmouth). Fourth, the study area overlaps the rapidly growing
800,000-population Omaha metropolitan area, wherein concerns about
human­river interactions and management should only increase in the
future. The goal of this study is to address some fundamental
geomorphic underpinnings of the lower Platte River ecosystem by
seeking a robust assessment of change and by understanding that
change in terms of processes.
2. Regional setting
The lower Platte River flows through multiple physiographic
subregions (Fig. 1). It flows in a broad valley cut through rolling, loessmantled
plains between Grand Island and Fremont, NE, through its Big
Bend or North Bend (Fig. 1). From Ashland, NE, the river turns abruptly
eastward in its lowermost 35 km to flow through a shallowgorge eroded
into Pennsylvanian sedimentary rocks (Lugn, 1935, p. 19), hereafter
referred to as the Eastern Platte River Gorge (Fig. 1: EPRG). The EPRG
includes the much smaller but still prominent South Bend (Fig. 1). The
EPRG is a geologically young feature, dating to the Wisconsinan stage
Fig. 1. Location of study area and study stretches CC, SC, RG, and EP. Also depisted are Forks of the Platte (FP), Big Bend (BB) and South Bend (SB).
Fig. 2. Stretch EP showing time series of the development of stabilized islands (black). Numbers and subordinate letter pairs indicate islands or groups of islands (e.g., 1a) that are
gradually accreted to the banks (e.g., 1b: hatched pattern). Small "functional" anabranches created by the isolation of braid channels after the stabilization of islands (e.g., landward
side of island 1a) are too small to be depicted at scale of diagram.
408 R.M. Joeckel, G.M. Henebry / Geomorphology 102 (2008) 407­418
and records the capture by the Platte of the lowermost Elkhorn River (cf.
Condra, 1903). The Platte River must have captured a major part of the
Loup River asit shifted its course northward after the earliest Pleistocene
(cf. Swinehart et al., 1994).
Study stretch EP lies within the EPRG. Stretches CC, SC, and RG (in
contrast to stretch EP) are characterized by broad valley widths and a
lack of bedrock constrictions.
We employ the standard use of the term "anabranch" to describe
subsidiary channels that divert from the main channel belt and rejoin
downstream. Likewise, we consider "braid channels" to be smaller
channels within the main channel belt, but our general usage of the
term "channel" hereafter refers to the entire belt of braid channels as a
whole. We have observed that changes in island stabilization and bank
accretion eventually convert some braid channels into "functional"
anabranches by isolating them from the main river channel.
From the forks of the Platte (the confluence of the North and South
Platte Rivers; Fig. 1) eastward into the study area, however, much
larger anabranches a few kilometers to a few tens of kilometers in
length have been prominent from at least the time of first
Euramerican contact (early eighteenth century), and presumably
much earlier. Large, long-lived anabranches such as these are
particularly prominent between the present towns of North Platte
and Grand Island, NE. Stabilized tracts of braidplain between
anabranches are considered, for the purposes of discussion but not
statistical analyses, to be large to very large islands even though their
origins are clearly different from those of stabilized in-channel bars.
The most conspicuously large and long-lived anabranches are located
near Grand Island, 32 km upstream from study stretch CC, where
seven major anabranches exist within a 29-km-long stretch. Downstream
from Grand Island, Nebraska ca. 1860, anabranches on the
Platte River become much shorter and much less numerous. In the
twentieth century, however, every study stretch included anabranches
of some kind, whether the short-lived "functional" ones
isolated during the process of island stabilization and accretion or the
Fig. 3. Stretch RG showing time series of the development of stabilized islands (black). Numbers and subordinate letter pairs indicate islands or groups of islands (e.g., 1a) that are
gradually accreted to the banks (e.g., 1b: hatched pattern). Short anabranches (a) are abandoned in the process of island accretion between 1957 and 1993 (2b and 3b).
Fig. 4. Stretch RC showing time series of the development of stabilized islands (black). Numbers and subordinate letter pairs indicate islands or groups of islands (e.g., 1a) that are
gradually accreted to the banks (e.g., 1b: hatched pattern). Arrows indicate positions at which groups of islands have stabilized downstream of open bends.
409R.M. Joeckel, G.M. Henebry / Geomorphology 102 (2008) 407­418
much rarer, larger, long-lived ones that had been present since at least
the mid-nineteenth century, and probably much earlier.
3. Materials and methods
Manually georeferenced sets of historical aerial stereophotographs
and digital orthophoto quadrangles dating from 1938 to 2005 are the
principal data sources for our assessment of historical channel change.
U.S. General Land Office (GLO) surveys (see Cazier, 1977) from the
1850s and 1860s (Nebraska State Surveyor's Office, undated), although
constituting a very different kind of documentation, are nonetheless
the earliest reliable maps of the Platte River and therefore provided
ancillary baseline data. Four stretches of the lower Platte River
(Figs. 1­5) were selected for analysis: (i) stretch EP near Louisville,
selected because of its position downstream from the mouth of the
Elkhorn River (Fig. 2); (ii) stretch RG, immediately downstream from
the Loup-Platte confluence (Fig. 3); (iii) stretch SC, in the vicinity of
Silver Creek, Nebraska, just upstream from the Loup-Platte confluence
(Fig. 4); and (iv) stretch CC, near Central City (Fig. 5), which extends
eastward from the point at which recent analyses of the river by
Johnson (1997) ended.
Scanned and adjusted stereophotograph images were georeferenced
in ArcMapTM. Image data were layered over Landsat TM images
in ArcViewTM GIS. Channels and stabilized islands of the Platte River
were rendered onscreen over these layers at a scale of 1:20 000. A
fundamental distinction was made between what appeared to be
islands stabilized by woody vegetation (dark photograph tones, with
discernible trees in the highest resolution imagery) and temporarily
emergent bars. Only stabilized islands were rendered (Figs. 2­5). After
an initial onscreen rendering at a scale of 1:20000, resultant channel
and stabilized island surface area data were scrutinized for potential
anomalies by the examination of plots and the comparison of statistics
from a given stereophotograph series with those from the next series
in time sequence. Subsequently, each rendering was carefully edited at
1:10000 and was assessed for continuity of selection criteria for
groups of individual stabilized islands and for active channel segments
in successive aerial stereophotograph series (Figs. 2­5). Resultant data
were once again examined, rendered features were re-examined in
time sequence, and a third edit of rendered features was made at a
scale of 1:10 000. Point-to-point measurements of channel width
were also carried out at four or five equally spaced positions along
each of the four stretches in the GIS project at a scale of 1:10000.
Width-measurement points were selected to avoid obvious and
immediate interference by bridges and other manmade structures. A
single operator performed all activities for the purposes of consistency
and overall precision. Unfortunately, serial measurements of stream
depth and surveyed cross sections (the "third dimension" of change)
are unavailable, but earlier studies (e.g., Williams,1978; Eschner,1983;
Eschner et al., 1983; Kircher and Karlinger, 1983) demonstrated that
these constraints do not negate the value of examining river planform
change over decadal timescales. Moreover, no a priori reasons exist to
assume that channel depths have increased dramatically as planform
Fig. 5. Stretch CC showing time series of the development of stabilized islands (black). A long-lived anabranch (a) has gradually dwindled since 1938.
Table 1
Sinuosity of study stretches ca. 1860 and in 2005
Year EP RG SC CC
ca. 1860 1.00 1.04 1.02 1.00
2005 1.01 1.05 1.03 1.01
Table 2
Apparent bank erosion (m2
/km) between GLO survey (ca. 1860) and first aerial
stereophotograph year (1938 or 1941) in four study stretches
Stretch Erosion (m2
/km)
EP 65260
RG 149760
SC 32980
CC 17270
410 R.M. Joeckel, G.M. Henebry / Geomorphology 102 (2008) 407­418
Fig. 6. Stretch CC in 1999 (including smaller anabranch "a"; see Fig. 5) relative to same stretch from early-1860s GLO surveys of different townships (e.g., T. 14 N., R. 5 W.). Named and
numbered (e.g., Prairie Island, "Island No. 2") islands did not exist as islands per se by 1938 because the South Channel anabranch had already dwindled. See Fig. 1 for location.
Fig. 7. Major changes in Platte River planform since GLO mapping of early 1860s along boundary betweenT.15 N., R. 3 W. and T.16 N., R. 3 WE in stretch SC. (aa) Divergence of South Channel
anabranch (SCh), which is entirely separate entity from "South Channel" in Fig. 6, from main channel of Platte River around large island of stabilized braidplain (i) in 1861 GLO survey. (b)
Same area in 1999; abandonment of SCh occurred prior to 1938; "1" represents south bank of SCh as taken from 1861 survey and "2" represents prominent break between smooth terrain
and bar-and-swale topography that may represent an even earlier south bank position of SCh, potentially indicating shrinkage in width and diminution of flow in the anabranch prior to
1861. Post-1938 accretion of stabilized islands has grown across the openingof SCh.(c) and (d) fate of short anabranch around island of stabilized braidplain (i). In both areas, the appearance
of gravel pits (p) which begins prior to 1938, indicates that anabranches had long been abandoned and that accretion lands are already relatively stable. See Fig. 1 for locations.
411R.M. Joeckel, G.M. Henebry / Geomorphology 102 (2008) 407­418
changes have occurred. Sparse anecdotal data exist for some level of
channel downcutting at some sites on the lower river.
4. Channel and island change
4.1. General observations
Channel area in the lower Platte River has decreased since the mid-
1800s, and this decrease can be well documented from the date of the
first aerial stereophotography in the area (1938) onward. The decrease
in channel area resulted from the stabilization of mobile bars by
woody vegetation (i.e., the production of new islands), the accretion of
islands to the river's banks, and the abandonment of braid channels
and anabranches (Figs. 2­5). Upstream from the study area, marked
woodland expansion on the North Platte, South Platte, and upper to
middle Platte River involved the widespread growth of cottonwoods
and willows in formerly active channels and on formerly mobile bars,
mostly between 1900 and the 1930s (Johnson, 1994). Woodland
expansion on the lower Platte is less easily contrasted with preEuramerican
settlement conditions because wooded islands were
present upstream to the mouth of the Loup River long before intense
Euramerican interaction with the river (Secretary of War, 1836;
Auerbach and Alter, 1940; Rollins, 1995). Surveyor's notes from the
GLO surveys of the mid-1850s to early 1860s (Appendix) implied that
riparian and valleyside woodlands had always been better developed
in the EPRG. GLO survey notes from the mid-1850s (Appendix)
described at least nine large wooded islands in the Platte in the EPRG,
similar to those visible in aerial imagery in stretch EP from 1941
onward. Some of these islands were prominent enough to receive
place names, such as "Cedar Island," which are still in use. Even in the
lowermost Platte River, however, woody vegetation has stabilized
additional in-channel bars (creating new islands) and proliferated on
pre-existing islands since at least 1938, and probably decades earlier.
In stretch SC, groups of islands have tended to stabilize, grow, and
begin to accrete to the banks downstream of large, open bends in the
river (Fig. 4); whereas in the other study stretches, such a pattern is
not readily apparent (Figs. 2, 3, 5).
The lower Platte has remained more or less within the braid belt
delineated by late-1850s and early-1860s GLO surveys, and the
sinuosities of the four study stretches have not changed significantly
since ca. 1860 (Table 1), but apparent bank erosion per kilometer
between the date of the GLO survey (ca. 1860) and the first aerial
stereophotograph year in each stretch (1938 or 1941) varies
considerably among the stretches. Stretch RG experienced the most
bank erosion during that early period of historical change, EP the
second most, followed by SC, and finally CC, which experienced very
little bank erosion (Table 2).
The fate of anabranches and large to very large islands (as opposed
to in-channel bars gradually stabilized after 1938 by vegetation or
joined groups of smaller, stabilized bars) is especially noteworthy.
Prairie Island, a 13 km-long and 2.5 km-wide area of stabilized
braidplain between the main (current) channel of the Platte River and
an abandoned anabranch (the South Channel), appears in an 1862 GLO
survey upstream along study stretch CC (Fig. 6). Although the place
name "Prairie Island" persists today, the South Channel in this area
had been abandoned by 1938 and an island per se no longer exists.
Several other islands named in the GLO survey of the area have since
been accreted through the abandonment of the South Channel
anabranch (Fig. 6).
In stretch SC, a much longer (~28 km) anabranch also called the
"South Channel" (but distinct from the anabranch of the same name
that flowed around Prairie Island) also diverged from the main
channel of the Platte in the area; this anabranch has since dwindled
and islands within it have been accreted (Fig. 7a, b; SCh). In the same
Fig. 8. Change since late 1850s around the Loup-Platte confluence near Columbus, Nebraska westward from upstream end of stretch RG (indicated at right). "Grand Island" is term used
in GLO survey of T.16 N, R.1 E. for stabilized tract of braidplain between South Channel anabranch (same anabranch as in Fig. 7a, b) and does not seem to equate to other historical uses
of the place name for intra-anabranch tracts near the present city of Grand Island, farther upstream. Confluence has moved from point ×1 to point ×2 since 1858. A short anabranch
(a) and smaller tract of stabilized braidplain (i) at right are also abandoned after 1858. Note slight lateral migration of Platte b associated with bank erosion. See Fig. 1 for location.
412 R.M. Joeckel, G.M. Henebry / Geomorphology 102 (2008) 407­418
stretch, an unnamed 3.6-km-long and 0.7-km-wide island mapped in
the early 1860s was accreted when the river narrowed by the
abandonment of a short anabranch north of the island sometime
before 1938 (Fig. 7c, d; a). Downstream near Columbus, Nebraska, the
same South Channel (Fig. 8) is uniquely referred to as a "bayou" in the
late 1850s GLO survey, suggesting that flow was already more sluggish
there than in the main channel and that abandonment had already
begun. Furthermore, a short, broad anabranch existed immediately
upstream from stretch RG in the early 1860s (Fig. 8), but it too has
since been abandoned. This anabranch (Fig. 8; a) was separated from
the main channel by a large tract or "island" (Fig. 8; i) of stabilized
braidplain.
4.2. Channel width and area statistics
Change in channel widths is not consistent between all stations in
the four stretches. Width at individual stations has increased very
slightly, decreased slightly, or decreased precipitously at particular
stations since 1938 (Fig. 9a­d). Some individual stations show as much
as a 233% difference between minimum and maximum widths
attained during the period 1938­2005 (Fig. 9b), but the difference
between minimum and maximum channel width attained is b8.0% for
at least one station in all four stretches (Fig. 7). Overall, stretch CC (Fig.
9d) is the least variable stretch in terms of channel width, with the
exception of an abrupt apparent narrowing between 1999 and 2005 (a
period of significant regional drought) that relates to the abandonment
of a long-lived anabranch, which (on the basis of serial aerial
photographic observations) had nonetheless been dwindling for
decades (Fig. 9d, no. 4; aa). Channel widths in stretch RG varied the
most because of the abandonment of two larger, long-lived
anabranches (Fig. 7b, nos. 1 and 2; aa) during the study period. In
the other stretches (EP and SC), small, "functional" anabranches
(produced through the isolation of former braid channels and the
stabilization of large islands after 1938) were subsequently abandoned.
When the GLO surveys are used as ancillary data sources for
the period 1855­1862 and when width is viewed relative to the
baseline data set at 1860, the overall trend is also one of decreasing
channel width (Fig. 10). Through this exercise, the abandonment of
large anabranches is identified as a significant contributor to overall
channel narrowing over time (Fig. 10c, d).
Channel surface area (Fig. 11a, b) decreased consistently during the
period 1938­1999 in stretches CC, SC, and RG, regardless of any other
changes. Channel area decreases in these stretches proceeded through
a succession of three phases: (i) sustained decreases in channel area
between 1938 and the late 1960s; (ii) stabilization of channel area,
commencing first in stretch RG in the late 1950s, and then occurring in
SC and CC by the late 1960s; and (iii) relatively slight change in
channel area after the late 1960s. Stretch EP, downstream from the
Elkhorn River (Fig. 11a, b), shows a slightly different pattern of (i) a
slight increase in channel area during 1941­1949, followed; (ii) a very
Fig. 9. Station-to-station channel widths for stretches EP (a), RG (b), SC (c), and CC (d). Abandonments of large, long-lived anabranches (aa) and small, "functional" anabranches
formed by isolation of channels through island stabilization and growth (aa') are indicated.
413R.M. Joeckel, G.M. Henebry / Geomorphology 102 (2008) 407­418
slight decrease during 1949­1955; (iii) a more pronounced decrease
in area during 1971­1993, followed by (iv) very slight, and perhaps
insignificant, fluctuations in channel area during 1993­2005. The net
trend in stretch EP is still a decrease in channel surface area after 1941,
although the relative decrease is lesser in comparison with other
stretches (Fig. 11b). Overall, stretches CC and SC (the stretches lying
upstream from the two major tributary confluences in the study area)
show the most marked absolute decreases in channel area since 1938
(Fig. 11).
4.3. Island statistics
Statistics were compiled for small to large islands that appeared to
have grown more or less de novo after ca. 1860 in each of the study
stretches, but only aerial imagery (1938­2005) was employed as data
sources. From 1938/1941 to 2005, the number of islands per river
kilometer was low in stretches EP and RG but considerably higher in
stretchesCC and SC upstreamof the major tributaryconfluences(Fig.12a).
The number of islands per kilometer in stretches CC and SC increased
from 1938 to the end of the 1940s, remained high through the mid-1970s,
and then decreased precipitously until the early 1990s (Fig. 12a).
The cumulative area of stabilized islands in each stretch during 1938­
2005 shows different patterns of progression over the studied time
interval compared to channel surface area (Fig.12b, c). Stretches EP and CC
had relatively low cumulative areas of stabilized islands during the study
period, but EP also consistently contained fewer islands overall (Fig. 12a)
and showed an overall decrease in cumulative island area (Fig. 12b).
Stretch CC, in comparison, experienced a slight increase in cumulative
island area until the late 1960s, followed by a long period of comparative
stability (Fig. 12b). Stretch RG showed an increase in cumulative island
area until the end of the 1940s, followed by a precipitous decline until the
early 1990s, and then a state of comparative stability (Fig.12b, c). Relative
change in cumulative island areas since 1938/1941 is greatest in SC and CC
and least in RG and EP. A significant change in conditions with respect to
cumulative island area in the late 1960s is more evident in both SC and CC
than in the downstream study stretches (Fig. 12c).
Fig. 10. Examination of station-to-station channel widths for four study stretches. (a) Stretch EP, where several small "functional" anabranches produced by isolation of channels as
islands stabilized and grew, are abandoned over time (aa') at three stations. (b) Stretch RG, where both small (aa') and large (aa) anabranches are abandoned. (c) Stretch SC, showing
effects of abandonment of large anabranches between 1858 and 1938 at stations 1, 3, 4, and 5: plain numbers and solid lines (e.g. 3) represent width of main channel alone at stations,
whereas numbers with asterisk (e.g., 3) represent widths including large anabranches and including large stabilized islands between anabranches; stations 3, 4, and 5 include Prairie
Island (Fig. 6). (d) Stretch CC, also showing effects of abandonment of large anabranches between prior to 1938 at stations 2, 3, and 4; large, gradually dwindling anabranch mapped in
original GLO survey is abandoned between 1999 and 2005 at station 4, probably because of prolonged drought donditions. 1858­1862 GLO survey baseline widths are all set at 1860
for ease of comparison. Individual stations are identified by same numbers as in Fig. 9.
414 R.M. Joeckel, G.M. Henebry / Geomorphology 102 (2008) 407­418
Furthermore, the median area of stabilized islands during the
period 1938­2005 changed least in stretches CC and SC (Fig. 12d).
Median island area increased in both of these upstream stretches after
the late 1960s. Stretch CC shows a slightly greater increase in median
island size between the late 1960s and the early 1980s than does SC
(Fig. 12d). Median island size changed more, and indeed erratically,
during the period 1938/1941­2005 in stretches EP and RG. The
greatest changes appear as sharp increases in median stabilized island
sizes in EP during the mid-1950s and in RG during the late 1960s
(Fig. 12d). The total size range of stabilized islands observed in all four
study stretches spans five orders of magnitude (Fig. 13). The size
distribution of stabilized islands shifted least through the period
1938­2005 in stretch CC (Fig. 13D). In stretch EP, the size distribution
of islands appears to have shifted erratically over time (Fig. 13A), but
temporal patterns can be surmised from size distributions in the other
stretches (Fig. 13B­D). In RG, a general increase in area occurred in the
part of the island size range below the 40th percentile and above the
85th percentile (Fig.13B). Shifts in the size distribution occurred in the
years 1957 and 1969, the period in which channel surface area was
stabilizing (Fig.13B). In stretches SC and CC the surface areas of islands
above the 80th percentile also increase over time (Fig. 13C,D),
although in stretch CC the surface areas of islands below the 40th
percentile generally decrease over time, indicated by an overall
downward shift in the plotted size distribution below this percentile
between 1938 and 2005 (Fig. 13D).
Plots of the ratio of cumulative stabilized island area to channel area
during 1938­2005 (Fig. 14A) and of the same ratio presented in terms
of change relative to 1938/1941 baseline values are nearly identical to
the corresponding plots of cumulative island area (Fig. 12b,c).
5. Discussion
Early studies, predominantly concentrating on stretches west of
the study area (e.g., Williams,1978; Eschner,1983; Eschner et al.,1983;
Kircher and Karlinger, 1983), demonstrated striking decreases in
channel width since the 1860s. Johnson (1994), in a singularly
important study, examined decreases in channel area and increases
in island stabilization over small stretches of the upper and middle
Platte and related them to the expansion of cottonwood- and willowdominated
woodland, which was regulated by patterns of (i) June
flows, (ii) summer droughts, and (iii) wintertime conditions in the
channel belt. These factors are particularly relevant to the life histories
of cottonwood and willow trees, specifically in their effects on
seedling survivorship (Johnson, 1994). Reductions in June flow related
to upstream impoundments and diversions (mostly upstream of the
present study area) during the twentieth century "exposed much of
the riverbed and elevated recruitment and seedling survivorship"
(Johnson, 1994). Seedling survivorship and flow reduction, however,
have probably had ever more diminished importance eastward on the
Platte River. From the Loup River downstream, yearly rainfall
increases, major tributaries enter the river, and woody vegetation
has, by all accounts, always more abundant.
Our results show a general decrease in channel surface area in the
study stretches over time, similar to that detected in earlier studies
farther upstream in the Platte system. The farthest downstream reach in
our study, however, actually shows a slight increase in channel area in
the 1940s. Width measurements of the four study stretches of the lower
Platte River vary over time between individual stations on the same
stretch, underscoring the robustness of serial channel area measurements.
Nonetheless, the comparison of both width and area measurements
over time remains an important means of characterizing the
nature of change, particularly anabranch abandonment, which may
contribute relatively little to area measurements but greatly to decreases
in width. In addition, the distinction between "functional" and longlived
is prompted by scrutiny of width measurement changes.
Throughout the aerial stereophotograph study period, the number of
stabilized islands per river kilometer in the study stretches was, overall,
greater upstream from the major tributary confluences than it was
downstream of them. Stretches SC and CC, which lie upstream from the
Loup-Platte confluence, both show overall increases in the number of
islands per river kilometer from the mid-1950s to the late 1960s. These
trends strongly separate the upstream stretches from stretches EP and
RG downstream of major tributary confluences (Fig. 12a). The relatively
small median island sizes in stretches CC and SC (Fig. 12d), when
interpreted in conjunctionwith the islands-per-kilometer data (Fig.12a)
and, even more importantly, with qualitative observations from serial
aerial photographs (Figs. 4, 5), reflects the continual stabilization of
many small, in-channel bars even as older, stabilized bars were joined
with each other to form large, vegetated islands (Fig. 13C,D).
In stretch EP (Fig. 2), at the far downstream end of the study area,
qualitative observations indicate a very different history with respect
to islands. Both large and small islands were already present in stretch
EP in 1941, and an examination of the GLO survey for the area indicates
at least one of these large islands was a tract of stabilized floodplain
excised from the bank by the river after 1855. By 1955, however, many
of the smaller islands (individual, stabilized in-channel bars) in stretch
EP had disappeared, while most of the large islands from 1941
generally remained in the same form. New, smaller islands appeared
in subsequent years, but by 1971, several of the large islands present in
earlier aerial photo years had been accreted to the banks by the
abandonment of channels and, presumably, by some amount of sand
deposition in those channels (Fig. 2). Similarly, in stretch RG, large
islands were accreted to the banks; and anabranches were abandoned
Fig.11. (a) Cumulative channel area over time in study stretches. (b) Cumulative channel
area as a percentage of baseline values from first aerial photograph year in each stretch,
either 1938 or 1941.
415R.M. Joeckel, G.M. Henebry / Geomorphology 102 (2008) 407­418
as channel width decreased over much of the reach after the late
1960s (Fig. 3).
The very strong similarities in cumulative stable island area and
the ratio of island area to channel area in all stretches (Figs. 12B, 14Bb)
indicate that, since 1938, the Platte River underwent change
dominantly by the growth and accretion of islands and only to a
much lesser extent by the abandonment of anabranches. Nonetheless,
estimates of pre-1938/1941 bank erosion (Table 2) also demonstrate
that the two downstream stretches, and particularly stretch RG,
underwent more lateral migration than the two upstream stretches.
Conversely, relatively little bank erosion occurred in stretches SC and
CC during the same period (Table 2). In stretches SC and CC station
widths decreased considerably after ca. 1860 (Fig. 10C,D). These
decreases in width certainly corresponded to decreases in channel
surface area as well, but comparing GLO survey data and aerial
stereophotography with respect to exact values of island and channel
surface area would be highly questionable considering the fundamental
differences between the two data sets. In any case, anabranch
abandonment likely played a bigger role in hypothetical decreases in
channel area between ca. 1860 and 1938 because, originally, multiple,
large anabranches existed in stretches SC and CC. Nonetheless, when
the effect of the abandonment of large anabranches is removed from
consideration and all stretches are compared, the relative width of the
"main" channel (i.e., the one that consistently contains flow from the
mid-1850s forward) is actually the most stable in stretch CC, even
though that channel was gradually narrowed over time by island
accretion. Although in all of the study stretches the Platte has largely
stayed within its ca. 1860 channel belt, the combination of observations
outlined above indicate an increasing, upstream tendency of inplace
narrowing of the channel. Downstream of the major tributaries
where discharge was higher and more consistent, though, more lateral
migration by bank erosion took place.
6. Conclusions
Channel area stands out as a very robust and consistent measurement
of historical change on the lower Platte River. Channel shrinkage
is prominent, as has been demonstrated upstream (e.g., Johnson,
1994). Even within the lower Platte River, however, the difference in
behavior between stretches upstream and downstream from major
tributaries (Loup and Elkhorn Rivers) is striking. The least relative
decrease in channel area and the most bank erosion occurred in the
two stretches of the lower Platte River that lie downstream of major
tributary confluences. The farthest downstream stretch (EP) actually
showed an initial increase in channel area during the 1940s. Channel
area has been stabilizing gradually across in the study area since the
1950s.
Measurements of island area greatly expand the assessment of
historical change. The overall decrease in cumulative island area and
the ratio of island area to channel area in stretches EP and RG relative
to 1938/1941 baseline data, the comparatively great fluctuations in
median island size in both stretches, and the erratic behavior of island
Fig.12. (a) Number of islands per kilometer over time in study stretches. (b) Cumulative area of stabilized islands over time in study stretches. (c) Cumulative area of stabilized islands
as a percentage of baseline values from first aerial photograph year in each stretch, either 1938 or 1941. (d) Median island size over time in study stretches.
416 R.M. Joeckel, G.M. Henebry / Geomorphology 102 (2008) 407­418
size distributions over time in stretch EP indicate two processes. First,
large islands were accreted to the banks at different times in the
subsequent 150 years. Second, some small stabilized in-channel bars
were episodically "flushed" out of the system in these stretches,
probably as a part of an overall mechanism of dynamic equilibrium
that maintains a certain range of channel width under conditions of
comparatively higher discharge. The upstream stretches (SC and CC),
in comparison, showed a more consistent impact from the creation of
new small islands by bar stabilization.
We conclude that a comprehensive management plan for the
Platte River downstream from the Elkhorn and Loup Rivers should
differ from the management philosophy proposed by Johnson (1994)
for the river upstream from Grand Island, and also that monitoring
changes in bar elevation, channel grade, and sediment supply would
greatly improve the assessment of future changes by providing the
"third dimension" of change related to human activities. The
observation of apparently dwindling anabranches of the lower Platte
in 1850s GLO survey notes, however, provides a glimpse of longer term
changes essentially unconnected to human activities. Therefore,
knowledge about the river's characteristics in the recent geologic
past is equal in relevance to monitoring programs in the formulation
of management policies. The establishment of river characteristics
relative to major tributary confluences in this study suggests that it
may be possible to interpret the sedimentary record of the ancient
Platte River in the context of the putative captures of major tributaries
(Loup and Elkhorn Rivers). Moreover, precise channel-belt and
ground-penetrating radar surveys, detailed soil and shallow stratigraphic
studies, extensive radiocarbon and luminescence dates also
have the potential to document a record of riparian change during
early historic (~1700 to 1854) and late prehistoric times. Such a record
would provide a legitimate means by which hypotheses about the
"natural" condition of the Platte River could be tested.
Acknowledgement
This study was made possible through the award of NASA/JPL
Contract 1247442 in conjunction with the Association of American
State Geologists. P.R. Hanson, J.A. Mason, C.R. Fielding, and D.B. Loope
provided helpful discussions.
Appendix A
Field notes of 1850s government land survey in Eastern Platte River
Gorge (EPRG) and vicinity, archived as unpublished typescripts at
Fig. 13. Size distributions of stabilized islands in stretches EP (a), RG (b), SC (c), and CC (d).
417R.M. Joeckel, G.M. Henebry / Geomorphology 102 (2008) 407­418
Sarpy County, NE Surveyor's Office (Papillion, USA). Typescripts were
generated from transcripts of original surveyors' notes made at the
U.S. Surveyor General's Office, Plattsmouth, 1882.
"Field Notes of the Subdivision Lines and Meanders of Platte River
in Township 12 North, Range 10 East of the 6th Principal Meridian of
Nebraska by Philander C. Patterson, Deputy Surveyor."
"Field Notes of the Survey and Establishment of the Third Standard
Parallel North of Base Line through Ranges 10, 11, 12, 13, and 14 East of
6th P.M. Executed by Charles A. Manners U.S. Deputy Surveyor"
"Field Notes of the Survey of Exterior Township Lines of Fractional
Townships 13 and 14 North of Base Line of Ranges 10, 11, 12, 13, and 14
East of 6th P.M. and Fractional Township 12 North of Ranges 10 and 11
East of 6th P.M. in Nebraska Executed by W.N. Byers, Joseph H. Wagner,
and Merriwhether Thompson U.S. Deputy Surveyors"
"Field Notes of Survey of 3rd Standard Parallel along North
Boundary Section 6 and North and South Quarter Line and Meanders
of Missouri and Platte Rivers in Sarpy County Nebraska"
"Field Notes of Township 12 North, Range 11 East of Sixth Principal
Meridian in Nebraska As Surveyed by I.H. Wagner Deputy Surveyor."
"Field Notes of the Subdivision and Meander Lines of Township 13
North, range 10 East of the 6th Principal Meridian in Nebraska as
Surveyed by Charles E. Watson Deputy Surveyor"
"Field Notes of the Subdivision Lines of Township 13 North, Range
11 East of 6th Principal Meridian in Nebraska as Surveyed by John I.
Paynter Deputy Surveyor"
"Field Notes of the Subdivision and Meander Lines of Township 13
North, Range 12 East of the 6th Principal Meridian in Nebraska as
Surveyed by John I. Paynter Deputy Surveyor"
"Field Notes of the Subdivision Lines and Meanders of Township 13
North, Range 13 East of the 6th Principal Meridian in Nebraska as
Surveyed by John I. Paynter Deputy Surveyor"
"Field Notes of the Subdivision and Meander Lines of Township 13
North, Range 14 East of the 6th Principal Meridian in Nebraska as
Surveyed by William N. Byers Deputy Surveyor"
"Field Notes of the Subdivision and Meander Lines of Township 14
North, Range 10 East of the 6th Principal Meridian in Nebraska as
Surveyed by Charles E. Watson Deputy Surveyor"
"Field Notes of the Subdivision Lines of Township 14 North, Range
11 East of the 6th Principal Meridian in Nebraska as Surveyed by John I.
Paynter Deputy Surveyor"
"Field Notes of the Subdivision Lines of Township 14 North, Range
12 East of the 6th Principal Meridian in Nebraska as Surveyed by John
I. Paynter Deputy Surveyor"
"Field Notes of the Subdivision and Meander Lines of Township 14
North, Range 13 East of the 6th Principal Meridian in Nebraska
Executed by John I. Paynter Deputy Surveyor"
References
Auerbach, H.S., Alter, J.C., 1940. Life in the Rocky Mountains, 1830­1835, by Warren
Angus Ferris. Rocky Mountain Book Shop, Salt Lake City.
Cazier, L., 1977. Surveys and Surveyors of the Public Domain 1785­1975. Department of
the Interior, Washington, D.C.: United States.
Condon, S.M., 2005. Geologic studies of the Platte River, South-Central Nebraska and
adjacent areas -- Geologic maps, subsurface study, and geologic history. United
States Geological Survey Professional Paper, vol. 1706. CD-ROM.
Condra, G.E., 1903. An old Platte channel. American Geologist 31, 361­369.
Currier, P.J., Davis, C.A., 2000. The Platte as a prairie river: a response to Johnson and
Boettcher. Great Plains Research 10, 69­84.
Eschner, T.R., 1983. Hydraulic geometry of the Platte River near Overton, south-central
Nebraska. United States Geological Survey Professional Paper, vol. 1277-C, pp. 1­32.
Eschner, T.R., Hadley, R.F., Crowley, K.D., 1983. Hydrologic and morphologic changes in
channels of the Platte River basin in Colorado, Wyoming, and Nebraska: an historical
perspective. United States Geological Survey Professional Paper, vol. 1277-A,
pp. 1­39.
Johnson, W.C., 1994. Woodland expansion in the Platte River, Nebraska: patterns and
causes. Ecological Monographs 64, 45­84.
Johnson, W.C., 1997. Equilibrium response of riparian vegetation to flow regulation in
the Platte River. Regulated Rivers: Research and Management 13, 403­415.
Johnson, W.C., Boettcher, S.E., 2000. The presettlement Platte: wooded or prairie river.
Great Plains Research 10, 39­68.
Kircher, J.E., Karlinger, M.R., 1983. Effects of water development on surface-water
hydrology, Platte River basin in Colorado, Wyoming, and Nebraska upstream from
Duncan, Nebraska. United States Geological Survey Professional Paper, vol. 1277-B,
pp. 1­49.
Lugn, A.L., 1935. The Pleistocene Geology of Nebraska. Nebraska Geological Survey
Bulletin, vol. 10, pp. 1­223.
National Academy of Sciences, 2004. Endangered and Threatened Species of the Platte
River. The National Academies Press, Washington, D.C.
Nebraska State Surveyor's Office, undated, Information Online: Original and GLO
Resurvey Plats, (web page), http://www.sso.state.ne.us/.
Rollins, P.A. (Ed.),1995. The Discovery of the Oregon Trail: Robert Stuart's Narratives of His
Overland Trip Eastward from Astoria in 1812­1813. Bison Books, University of
Nebraska Press, Lincoln. reprinting of original 1935 edition by Charles Scribner's Sons.
Secretary of War, 1836. Colonel Dodge's Journal. House Executive Document No. 181.
The Executive Documents of the House of Representatives for the First Session of
the Twenty-Fourth Congress. Washington, D.C., Blair and Rives Publishers.
Swinehart, J.B., Dreeszen, V.H., Richmond, G.M., Tipton, M.J., Bretz, R., and Steece, F.V.,
1994. Quaternary geologic map of the Platte River 4°×6° Quadrangle, United States.
United States Geological Survey Map I-1420.
Williams, G.P., 1978. The case of the shrinking channels: The North Platte and Platte
rivers in Nebraska. United States Geological Survey Circular, vol. 781, pp. 1­48.
Fig. 14. (a) Ratios of cumulative stabilized island area to channel area over time in study
stretches. (b) Changes in ratios of cumulative stabilized island area to channel area over
time, expressed as percentages relative to value from first aerial photograph year in
each stretch, either 1938 or 1941.
418 R.M. Joeckel, G.M. Henebry / Geomorphology 102 (2008) 407­418