Geomorphology of the Burnt River, eastern Oregon, USA: Topographic adjustments to tectonic and dynamic deformation

    Geomorphology of the Burnt River, eastern Oregon, USA: Topographic adjustments to tectonic and dynamic deformation Matthew Connor Morriss, Karl W. Wegmann PII: DOI: Reference:

S0169-555X(16)30827-3 doi: 10.1016/j.geomorph.2016.09.015 GEOMOR 5762

To appear in:

Geomorphology

Received date: Revised date: Accepted date:

5 December 2015 25 August 2016 6 September 2016

Please cite this article as: Morriss, Matthew Connor, Wegmann, Karl W., Geomorphology of the Burnt River, eastern Oregon, USA: Topographic adjustments to tectonic and dynamic deformation, Geomorphology (2016), doi: 10.1016/j.geomorph.2016.09.015

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

PT

Geomorphology of the Burnt River, eastern Oregon, USA: Topographic adjustments to tectonic and dynamic deformation

RI

Matthew Connor Morriss1,2 *, Karl W. Wegmann1 $

$

[email protected]

NU

[email protected]

MA

*

SC

1) Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University, Campus Box 8208, Raleigh, NC 27695-8208 2) ‘Present Address’ Department of Geological Sciences, University of Oregon, 100 Cascade, 1275 E 13th Avenue, Eugene, Or 97403-1272

D

Abstract

TE

Eastern Oregon contains the deepest gorge in North America, where the Snake River cuts vertically down 2300 m. This deep gorge is known as Hells Canyon. A

AC CE P

landscape containing such a topographic feature is likely undergoing relatively recent deformation. Study of the Burnt River, a tributary to the Snake River at the upstream end of Hells Canyon, yields data on active river incision in eastern Oregon, indicating that Quaternary faults are a first order control on regional landscape development. Through 1:24,000-scale geologic mapping, a 500,000-year record of fluvial incision along the Burnt River was constructed and is chronologically anchored by optically stimulated luminescence dating and tephrochronology analyses. A conceptual model of fluvial terrace formation was developed using these ages and likely applies to other nonglaciated catchments in eastern Oregon. Mapped terraces, inferred to have formed during glacial-interglacial cycles, provide constraints on rates of incision of the Burnt River. Incision through these terraces indicates that the Burnt River is down-cutting at 0.15 to

ACCEPTED MANUSCRIPT 0.57 m kyr-1. This incision appears to reflect a combination of local base-level adjustments tied to movement along the newly mapped Durkee fault and regional base-

PT

level control imposed by the downcutting of the Snake River. Deformation of terraces as

RI

young as 38.7 ± 5.1 ka indicates Quaternary activity along the Durkee fault, and when combined with topographic metrics (slope, relief, hypsometry, and stream-steepness),

SC

reveals a landscape in disequilibrium. Longer wavelength lithospheric dynamics

NU

(delamination and crustal foundering) that initiated in the Miocene may also be

MA

responsible for continued regional deformation of the Earth’s surface.

1. Introduction

TE

D

Key Words: strath terrace; fluvial incision; optically stimulated luminescence (OSL); tephrochronology; Quaternary stratigraphy; Inland Northwest

AC CE P

The ~2,850 km2 Burnt River catchment contains normal faults, Neogene volcanics and Quaternary fluvial terraces, all of which indicate a dynamic late Cenozoic landscape. The Burnt River drains into the Snake River at the upstream end of Hells Canyon. The Snake River sets the regional base level; any fluvial signal of the carving of Hells Canyon may have been communicated to tributary systems such as the Burnt River. The goal of this investigation is to improve our understanding of Neogene surface deformation in eastern Oregon through the development of a geo-chronologically-anchored model of fluvial dynamics and terrace development for the Inland Northwest (Fig. 1).

ACCEPTED MANUSCRIPT Detailed mapping of Quaternary deposits along the Burnt River constrains the rate and timing of fluvial incision in this catchment and reveals a ~500 ka record of incision in the

PT

lower and upper Burnt River canyons (Fig. 2A). Our mapping indicates that the Durkee

RI

fault, an active normal fault defining the Durkee Basin (Fig. 2A), controls the base level

NU

fluvial incision through this high-relief landscape.

SC

for the upper Burnt River catchment. Local faulting is believed to drive much of the

MA

Neogene incision in the lower Burnt River canyon is interpreted as a combination of base-level adjustments tied to the Snake River and footwall uplift along the Durkee fault

D

where the Burnt River exits the Durkee Basin (Fig. 2). The trend and size of this fault is

TE

inconsistent with Neogene extensional tectonics (i.e. Basin & Range) in the western U.S

AC CE P

(McCaffrey et al., 2013). It is possible that the Durkee fault formed due to a combination of rotational extension and long wavelength lithospheric flexure of in the Inland Northwest (McCaffrey et al., 2013). Geomorphic research of the Salmon River catchment indicates that this fluvial system is adjusting to 1 to 1.5 km of base-level drop that initiated between 8 and 10 Ma (Vogl et al., 2014; Larimer, 2015). Fluvial knickzones have migrated 250 km upstream from the Snake-Salmon confluence and have likely propagated through Hells Canyon into other tributaries of the Snake River, including the Burnt River.

2. Regional Geologic Setting The Burnt River drains portions of the Blue Mountain and Payette sections of the Columbia Plateau physiographic province (Fenneman and Johnson, 1946). Beginning at

ACCEPTED MANUSCRIPT 2,400 m in the Blue Mountains, the river drops nearly 1,800 m over a distance of 160 km before entering the Snake River at the upstream end of Hells Canyon. Two discrete

PT

canyon sections, the upper and lower Burnt River canyons, dominate the landscape of the

RI

lower 75 km of the river. The upper Burnt River canyon (Fig. 2A) is cut through the Triassic Burnt River Schist of the Baker Terrane (Ashley, 1995). The mouth of the upper

SC

canyon is coincident with the footwall trace of the Durkee fault, which defines the

NU

structurally-controlled Durkee Basin that is filled with Miocene to Pliocene lake sediments (Van Tassell et al., 2001). The river transits the Durkee basin, cutting across

MA

the Durkee fault for a second time as it enters the lower Burnt River canyon (Fig. 2A). It then encounters the terrane-bounding Connor Creek fault, transitioning from the Baker

TE

D

Terrane into the Jurassic Izee Terrane (Fig 3A; Dorsey and Lamaskin, 2008). Bedrock in this reach is dominated by steeply dipping shales of the Weatherby Formation (Dorsey

AC CE P

and Lamaskin, 2008). The Burnt River crosses into the Permo-Triassic volcanic and volcaniclastic rocks of the Olds Ferry Terrane and Huntington Arc, just before it enters the Snake River at the Brownlee Reservoir (Silberling, 1987).

A select series of events in the mid-to-late Cenozoic are intimately connected with the development of topography in eastern Oregon. Both the Sevier and Laramide orogenies resulted in high standing topography to the east and south of Oregon. Between 48 and 39 Ma, areas 50 to 100 km due east of Hells Canyon may have stood at an elevation of 3,700 m above sea level, as recorded in the δ18O of hydrous authigenic minerals in paleosols, whereas present elevations in this region are between 2,000 and 2,500 m (Mix et al., 2011). Rocks of the Precambrian Belt Supergroup formed the drainage divides between

ACCEPTED MANUSCRIPT major east and west-flowing Eocene rivers (Allen, 1991; Sears, 2014). Fluvial channelbed conglomerates composed of Proterozoic quartzite and Eocene Challis volcanic clasts

PT

occur at elevations of ~3,000 m in eastern Oregon mountain ranges (Allen, 1991). These

RI

isolated outcrops are remnants of former west-flowing Eocene channels (Cowan &

NU

SC

Reiners, 2004).

Miocene volcanism in southeast Oregon initiated at 16.8 Ma with the eruption of the

MA

McDermitt volcanic field and Steens Basalts. Both were associated with the interaction of the Yellowstone mantle plume with the base of continental lithosphere (Hooper et al.,

D

2007; Ferns and McClaughry, 2013). After these initial eruptions, the locus of basaltic

TE

volcanism propagated to the north, paralleling the edge of cratonic North America

AC CE P

(Hooper et al., 2002; Rodriguez and Sen, 2013; Ferns and McClaughry, 2013).

Extensional basins began opening in eastern Oregon and western Idaho as volcanoes were erupting. Normal faulting initiated on the north-south trending Oregon-Idaho graben (OIG) at 15.3 Ma and continued until 12.6 Ma (Supplementary Figs. 1S, 2S; Cummings et al., 2000). By 11 Ma, an internally-drained basin occupying the western Snake River Plain was present (Wood and Clemens, 2002). Sedimentary deposits associated with Lake Idaho in this basin are herein called Lake Idaho deposits (Wood and Clemens, 2002). Lake Idaho persisted for nearly 8 million years (10 to 1.7 Ma), was the size of modern Lake Huron, and may have had a surface water connection to other basins in northeastern Oregon (Van Tassell et al., 2001). The modern Snake River upstream of

ACCEPTED MANUSCRIPT Hells Canyon drains the area once inundated by Lake Idaho. Several other extensional basins in northeastern Oregon began to open by 9 Ma (Baker Basin, Grande Ronde

PT

Valley, Pine Valley, and perhaps the Durkee Basin; Supplementary Fig. 2S; Van Tassell

RI

et al., 2001). The geologic forces active in the region throughout the Neogene, including dynamic topography, tectonics, and volcanism have all left different marks on the

MA

NU

processes in the Burnt River catchment (Fig. 2).

SC

landscape. This study is focused on the fluvial geomorphic indicators of these active

3. Methods

D

A combination of morphometric analysis of a 10 m digital elevation model (DEM) and

TE

detailed fieldwork were used to investigate the development of topography in the Burnt

AC CE P

River catchment. Preliminary topographic analysis revealed a long wavelength convexity in the river longitudinal profile (Fig. 3 inset) that suggested a portion of the basin was in disequilibrium due to either short-wavelength tectonic or longer wavelength dynamic influences. We sought to deconvolve these two possibilities.

Field analysis focused on the upper and lower Burnt River canyons (Fig. 2A). Specifically, the spatial distribution and elevation of bedrock surfaces (straths) beneath river terrace gravels represent the primary field data collected in this study. Fluvial incision rates within both canyons were measured by comparing the modern river to strath elevations (e.g. Lavé and Avouac, 2000). Surficial travertine deposits, landslides, alluvial fans, pediments, alluvial units, and anthropogenically-modified sediments were

ACCEPTED MANUSCRIPT also mapped to better understand the late Quaternary evolution of the Burnt River

PT

catchment (Morriss, 2015).

SC

RI

3.1. Terraces

Fluvial terraces and their related deposits integrate tectonic, climatic, and geomorphic

NU

processes at the watershed scale (Bull, 1990; Pazzaglia, 2013). With constant incision, streams will establish channel gradients that allow for movement of all bedload supplied

MA

to the system (Bull, 1990, 1991). When the rate of horizontal river incision (meandering and valley widening) is greater than the vertical incision, the lateral beveling of a bedrock

TE

D

strath surface occurs (Pazzaglia et al., 1998; Wegmann and Pazzaglia, 2002; Montgomery, 2004). The lateral to vertical rate of incision is thought to be driven by both internal and

AC CE P

external controls (Hancock and Anderson, 2002; Wegmann and Pazzaglia, 2009). Bedload sediments are the tools of the stream that are needed to bevel a strath. These beveling processes continue until tectonic or climatic forces change stream power (discharge) or sediment supply (Bull, 1990). The strath surface may also be buried under alluvial fill and unmodified for an uncertain amount of time as described below; however with renewed down-cutting, a river abandons the strath and alluvial deposits in the landscape as a marker of where the river channel used to be. Many authors interpret river terrace sequences as resulting from primarily external forcing (climate, tectonics, etc); although it is important to note that internal (autocyclic) mechanisms have been invoked as possible driving mechanisms for the formation of strath terraces (e.g., meander migration, meander cutoff, and landslides; e.g., Finnegan and Dietrich, 2011).

ACCEPTED MANUSCRIPT

Fill terraces form when hillslope sediment influx is greater than stream capacity, resulting

PT

in aggradation. Channel aggradation buries the previously cut strath, creating a new,

RI

higher tread surface. Fill terraces are often interpreted to represent times of climatic

SC

change when hillslope sediment flux outpaces the channels ability to transport this sediment downstream. Tectonic activity expressed as regional uplift, propagation of base

NU

level fall accommodated through migrating knickpoints, or an increase in stream power

MA

due to climatic change can tip a stream from aggradation to degradation and cutting (Bull, 1991; Zaprowski et al., 2001; Cook et al., 2013). With an increase in power, streams will

D

incise through the thick accumulation of alluvial fill, preserving fill terraces in the

AC CE P

TE

landscape (Bull, 1990).

It is commonly assumed in tectonic geomorphology studies that the gradient (river longitudinal profile) of a paleo-strath surface (strath terrace) is similar to the modern channel profile for bedrock, or mixed bedrock alluvial rivers (e.g. Pazzaglia and Brandon, 2001). Before any correlations between straths and the modern channel are made, however, the modern channel is examined under the assumption that it currently represents the long-term equilibrium profile of the river in question (Wegmann and Pazzaglia, 2002). A vertical incision rate can then be calculated from the height of a strath terrace above the modern river as long as the age of the terrace deposits are known or can be estimated. For the two canyons of the Burnt River, terrace strath surfaces and deposits are well exposed by the hydraulic mining activities of late 19th century gold

ACCEPTED MANUSCRIPT miners. However, aggradation of the modern channel bed of the Burnt River in response to the influx of terrace and hillslope sediment liberated by these mining activities means

PT

that incision rate calculations along the Burnt River are minima, as bedrock is not

RI

exposed anywhere along the channel today. Thus the height differential between a terrace strath and the modern strath is unconstrained. Geotechnical and water wells were

SC

exploited in the lower Burnt River canyon to approximate the position of the modern

MA

NU

strath.

Recent models of terrace development (e.g. Pederson et al., 2006; 2013; Gallen et al.,

D

2015) imply that incision rates calculated between straths, rather than between a strath

TE

and the modern channel, represent a more precise estimate of long-term fluvial incision

AC CE P

rates. Rates determined between a strath and the modern channel may overestimate Quaternary incision rates because the position of a river channel is not a fixed datum with respect to the geoid – a requirement for the proper measurement of rate-dependent processes such as river incision (Gallen et al., 2015). Here we report incision rate calculations between terraces, and when possible, between terraces and the modern channel strath as constrained by data from water and geotechnical borings.

3.2. Field Methods This study focused on the landforms and geomorphology of the upper and lower Burnt River canyons (Fig. 2A). These canyons coincide with the two steeper segments of the Burnt River longitudinal profile and with steeper tributary streams (Fig. 3 inset, 4).

ACCEPTED MANUSCRIPT Terraces were located and strath surface elevations measured. Radiocarbon, tephrochronology, and optically-stimulated luminescence (OSL) samples were collected

PT

opportunistically from terrace alluvium, and basalt-cobble weathering rind thicknesses

RI

were noted (Fig. 5A). The degree of development of soil Bt and Bk horizons within the terrace deposits at different elevations above the modern channel were noted in addition

NU

SC

to the overall deposit thickness.

MA

Historic and recent hydraulic mining activities in and along the Burnt River exposed not only terrace deposits but also often the strath surface (Fig. 5B). Key outcrops exposed

D

along the lower Burnt River canyon were visible in road cuts (Fig. 5C). Strath surface

TE

elevations were measured using a Trimble GeoXH 2008-series differential GPS, with

AC CE P

locations resolvable to ± 0.15 m in both the horizontal and vertical (Supplementary Table 1S). At locations where the strath was not exposed, either a minimum or maximum elevation estimate was measured to provide reference to terraces up and downstream. A TruePulse laser range finder was used to measure the height of each strath above the modern Burnt River to within ± 1 m.

3.3. Tephrochronology Terrace deposits exposed by recent mining activity in the upper Burnt River canyon contained an intact tephra (Fig. 5E). A sample of this tephra was collected and analyzed at the Peter Hooper GeoAnalytical Lab at Washington State University (WSU) by an automated Siemens X-Ray powder diffractometer for major oxide constituents, (SiO2,

ACCEPTED MANUSCRIPT Al2O3, Fe2O3, TiO2, K2O, MgO, CaO, and Cl). The elemental profile from the Burnt River sample was compared for chemical similarity to the WSU tephra database (n =

PT

1716; Supplementary Table 2S). For a detailed description of the WSU

RI

tephrochronology and geochemistry procedures, see Johnson et al. (1999). Additional

3.4. Optically Stimulated Luminescence

NU

SC

research yielded an age of 201 ± 45 ka for this tephra (Kuehn and Negrini, 2010).

MA

Samples of fine-grained sand (100 – 250 µm) were collected from fluvial terrace deposits for OSL dating. Standard OSL sampling techniques were used. Samples were processed

TE

D

at the University of Georgia Luminescence Dating Laboratory. For a detailed description of lab methods, see http://osl.uga.edu/technique.html and Srivastava et al. (2005). OSL

AC CE P

measurements were made using an automated Risø TL/OSL-DA-15 reader, with a mounted 90Sr/90Y source and a dose rate of 0.085133 Gy/s (Markey et al., 1997). The quartz dose rate was measured following the Single Aliquot Regenerative protocol outlined in Murray and Wintle (2000). The in-situ dose rate was calculated using additional samples from each site. Dose rate calculations were made using a Daybreak alpha counting system to estimate U and Th concentrations. Sample water content was assumed to be ~8 ± 4%. The cosmic ray dose rate was estimated using sample elevation, depth, and latitude (Prescott and Hutton, 1994).

3.5. Remote Sensing DEM analysis of regional topography was conducted prior to field work to select priority

ACCEPTED MANUSCRIPT sites. The longitudinal profile of the Burnt River was extracted from the 10 m resolution National Elevation Dataset. TopoToolbox 2, a MATLAB native script package, was

PT

used almost exclusively for all topographic analyses (Schwanghart and Scherler, 2014).

RI

Topographic data were used to calculate normalized stream steepness (ksn). This normalized value refers to the steepness of a certain fluvial reach, normalized for the

SC

upstream drainage area (Kirby and Whipple, 2001; Kirby et al., 2003). Ksn is calculated

NU

using the following equation:

(eq. 1)

MA

Ksn = ksA(θref – θ)

where ks is the channel steepness index, A is the median upstream area from an examined

D

stream segment, and Θref is a reference channel concavity, usually the mean for channels

TE

in the region without knickpoints or tectonic disturbances, or in most cases Θref is

AC CE P

calculated over the entire catchment of interest (Wobus et al., 2006). Θref generally ranges from 0.35 to 0.65 (Kirby and Whipple, 2001; Wobus et al., 2006). Θ is the channel concavity as calculated from regression through a log-log plot of local slope versus contribution drainage area. A linear regression through the log-log slope and area plot provided an intercept (KS) of 22.73 and a slope (Θ) of 0.409. Ksn values were calculated for basins that drain areas ≥ 12 km2 and with channel lengths ≥ 2 km. The Ksn data were then exported and smoothed in ArcGIS (Fig. 4) with a moving mean window size of 100 m, similar in approach to the analyses of Miller et al. (2013).

ArcGIS was also used to create a slope raster with built-in DEM processing routines (Fig 2B) and to measure relief within the Burnt River catchment (Fig. 2C) within a circular

ACCEPTED MANUSCRIPT 100 m moving window. The possibility of a transient disequilibrium signal or knickpoint traveling up the Burnt River catchment was investigated further using slope-area plots

PT

and hypsometric curves (generated with the use of TopoToolbox 2) from seven select

RI

tributaries (Fig. 6). Hypsometry is the area distribution of elevations throughout a basin of interest (Strahler, 1952). The hypsometric integral (HI) was calculated for each sub-

SC

catchment by:

NU

HI = (Zo – Zmin)/(Zmax – Zmin)

(eq. 2)

MA

where Zo, Zmax, Zmin are the mean, maximum, and minimum elevations within the basin of interest. The hypsometric integral is an easy-to-calculate stand-in for tectonic activity,

D

indicating when catchments are undergoing active incision or have eroded to equilibrium

TE

(Keller and Pinter, 2002). Higher HI values (≥ 0.5) are often interpreted as indicating

AC CE P

disequilibrium either from tectonic activity or a migrating transient imparted from one or more knickpoints responding to base-level fall (Pike and Wilson, 1971; Keller and Pinter, 2002).

3.6. Lake Idaho sediments as a paleotopographic datum The presence of presumed Lake Idaho deposits within the Burnt River catchment serve as a potential paleoaltimetric control. Lake Idaho occupied the Western Snake River Plain (WSRP) between c. 11 and 1.7 Ma (Wood and Clemens, 2002). Wood and Clemens (2002) estimated that the maximum lake surface-water elevation was reached sometime between 3.8 and 2 Ma, at which time it overtopped a low divide between Dead Indian and Slaughterhouse Ridges that is today at 1,100 m above sea level (Fig. 1). The highest

ACCEPTED MANUSCRIPT elevation of Lake Idaho sediments provides a datum from which a gross incision rate for the lower Burnt River canyon may be calculated. We interpret these incision rates as

PT

minima as there has likely been erosion of lake deposits, some tectonic activity in the

RI

intervening time, and lake sediments were likely deposited below the true lake surface

NU

SC

level.

On Dixie Creek (Fig. 2), a horizontal basalt flow caps Lake Idaho deposits. This basalt

MA

flow and its eruptive edifice are part of the 2.2 to 0.1 Ma Kivett sequence (Ferns and McClaughry, 2013). These Quaternary flows provide a chronologic constraint on the

D

incision of Dixie Creek and an absolute topographic constraint, but only a minimum age

TE

constraint. If the lakebeds are older than 2.2 Ma, the resulting incision rate would be

4. Results

AC CE P

slower.

4.1. Terrace Stratigraphy

Seven distinct strath terraces were identified along the upper Burnt River canyon, while five were mapped in the lower canyon (Figs. 3A, B). Terraces are denoted from oldestto-youngest as Qt1U to Qt7U in the upper Burnt River canyon and from Qt1L to Qt5L in the lower canyon (Fig. 3). The upper and lower Burnt River canyon terrace sequences are treated separately, as the Durkee Basin is the local base level for the upper Burnt River canyon while the Snake River sets base level for the lower canyon reach (Fig. 2A). Terrace alluvium fines upward from cobble to boulder-sized clasts at the strath – terrace

ACCEPTED MANUSCRIPT contact, to overbank deposits of sandy silt (Fig. 5A). Geochronologic constraints on the younger terraces (Qt5U, Qt3L, and Qt5L) are provided by a geochemically matched tephra

PT

(Fig. 5E), OSL ages (see Table 1), and correlation of soil profile development to a

SC

RI

regional chronosequence established for the Palouse Hills in the Columbia Basin (Fig. 7).

4.1.1. Upper Burnt River Canyon Terrace Sequence

NU

In the upper Burnt River canyon, five mapped terraces appear to correlate across a

MA

significant distance (Qt2U, Qt3U. Qt5U. Qt6U and Qt7U). The upper Burnt River canyon provided only one reliable geochronometric constraint: a tephra collected in the fine-

D

grained overbank facies deposits of Qt5U (Fig. 5E). A sample from the Qt5u tephra was

TE

identified as a geochemical match to the 201 ± 47 ka Paoha Island tephra, which originated from the Mono Craters in eastern California (Supplementary Table 1S;

AC CE P

Williams, 1994; Kuehn and Negrini, 2010). A sample of fine sand to silt-sized quartzbearing fluvial sediment was gathered for OSL analysis from the higher and older Qt4 terrace just upstream from the terrace with the preserved tephra (Fig. 3A). The University of Georgia OSL lab returned an age of 81.59 ± 9.61 ka (Table 1); however, this OSL determination is inconsistent with the tephrochronology of the stratigraphically inset and younger Qt5U terrace. Lab results for this OSL sample (BRS17) do not exhibit obvious discrepancies when compared to the other samples in this study (Table 1); however, it is possible that some of the quartz sand grains collected and analyzed in this sample were contaminated by more recently-dosed sediment through bioturbation (rodent or tree roots), or the infilling of small slope-parallel fractures. Furthermore, we observe no evidence of an aggradation event large enough to deposit younger fluvial sediment

ACCEPTED MANUSCRIPT above the Qt5U terrace. With a high level of confidence in the tephra identification (≥ 95%), we chose to exclude the Qt4U OSL sample for constraining incision rates. The

PT

calculated incision rate at river km 56 for the Qt5U terrace (29.6 ± 1.0 m above the

RI

modern river) using the Paoha Island tephra age of 201 ± 47 ka is 0.15 ± 0.03 m ka-1

SC

(Table 2).

NU

Reconstructed terrace longitudinal profiles from the upper Burnt River canyon are approximately parallel to the modern channel profile (Fig. 3A). The one exception is a

MA

large step in the Qt3U terrace sequence, which correlates with an incised cutoff meander bend in the upper canyon between valley km 68 and 66. The river gradient before the

TE

D

meander was cutoff is preserved in the three uppermost Qt3U terraces. The pre-cutoff river was at least 1.8 km longer and had a gradient of ~0.001 over a distance of 6 km, as

AC CE P

preserved by Qt3U terraces (Fig. 3A). The gradient of the modern river across the same reach is ~0.004. This disparity in gradient could also be the result of meander growth outstripping fluvial incision (e.g., Finnegan and Dietrich, 2011). The section of the Qt3U terraces with a lower reconstructed gradient appears to correlate with a more sinuous reach of the Burnt River between valley km 68 and 62 (Fig. 3A). Lateral channel migration may have outpaced incision, resulting in a lower gradient channel. There is also a possible difference in the erodibility between the Nelson Marble and Burnt River Schist; the river may decrease its gradient faster where it is incising through marble than where it flows across schist. The contact between the two units is unmapped, but it may be between valley kilometer 62 and 59 (Fig. 3A; Brooks, 1979a, b; Ashley, 1995). However, this alone does not explain the observed lower channel gradient as the younger

ACCEPTED MANUSCRIPT Qt4U and Qt5U terraces do not exhibit similarly low gradients across the 68 to 58 valley

PT

km reach.

4.1.2. Lower Burnt River Canyon Terrace Sequence

RI

Only three terraces (Qt3L to Qt5L) exhibit mappable continuity along the length of the

SC

lower Burnt River canyon (Fig. 3B). Two OSL samples provide a chronology of incision

NU

for the lower Burnt River canyon during the past 90 ka. The fine-grained deposits in a Qt3L terrace returned an age of 84.74 ± 13.93 ka (Fig. 3B; Table 1), while a sandy lens

MA

from a Qt5L terrace yielded an OSL age of 38.68 ± 5.13 ka (Table 1). The two OSL constraints provided a long-term incision rate for the lower Burnt River canyon (Table 3).

D

To account for possible channel aggradation caused by late 19th to early 20th century

TE

gold-mining activities, we first calculated the incision rate between the Qt3L and Qt5L

AC CE P

terraces as 0.57 ± 0.2 m ky-1 (Table 2). The rate of incision measured directly from the Qt5L terrace to the modern channel is 0.2 ± 0.06 m ky-1. However, water and geotechnical wells drilled in the lower Burnt River canyon constrain 3 to 5 m of historic aggradation above the modern bedrock strath (Table 3). The incision rate between the Qt5L terrace and the bedrock strath below the modern alluvium is 0.31 ± 0.06 m kyr-1, and is comparable to the incision rate calculated between the Qt3L and Qt5L terraces.

The gradient of reconstructed terrace profiles Qt5L (0.004), Qt4L (0.003), and Qt3L (0.005) along the lower 25 km of the Burnt River above the Snake River confluence are similar to that of the modern Burnt River channel (0.004) (Figs. 3B, 8). However, with increasing proximity to the Durkee Basin, the gradients of the Qt4L and Qt3L terrace profiles increase compared to the modern river channel. This deflection may be due to

ACCEPTED MANUSCRIPT footwall uplift on the Durkee fault (see Section 5) as the higher and older Qt3L records a steeper gradient, consistent with the accumulation of more throw across the Durkee fault

RI

PT

relative to the younger Qt4L terrace (Fig. 3B).

SC

4.2. Paleo-elevation Metrics

NU

The current altitude of Miocene-to-Pliocene lacustrine sediments in the Burnt River

MA

catchment provide minimum paleoelevation estimates for Lake Idaho (c. 10 to 1.7 Ma) in and around the WSRP, as well as a possible mechanism for measuring post-Lake Idaho

D

surface deformation. The 1,100 m highstand of Lake Idaho occurred at approximately 4

TE

Ma (Wood and Clemens, 2002). The modern landscape position of these sediments within the lower Burnt River canyon was used to measure long-term incision at two

AC CE P

locations. First, lacustrine deposits are exposed on either side of the Burnt River valley at an elevation of 850 m two km upstream from the Burnt-Snake River confluence (Morriss, 2015). The Burnt-Snake River confluence is at 634 m (prior to the flooding of the Brownlee Reservoir). Using these age-elevation constraints, a minimum average incision rate of 0.54 ± 0.01 m ky-1 spanning the early Pliocene to Holocene was calculated for the lowermost Burnt River (Table 2).

A second paleoelevation –incision rate estimate comes from the Table Rock basalt flow, which is part of the Pleistocene Kivett Basalt sequence ( Hooper et al., 2002) that unconformably overlie Lake Idaho sediments in the mid-reaches of Dixie Creek (Fig. 3). The modern elevation of Dixie Creek nearest to the uppermost basalt flow is 780 m.

ACCEPTED MANUSCRIPT Available geochronology for the Kivett basalt sequence is limited to three 40Ar/39Ar ages: 0.8 ± 0.7 Ma, 1.7 ± 0.1 Ma, and 1.9 ± 0.3 Ma (Hooper et al., 2002). These three ages

PT

were treated as a single possible range for the age of the Table Rock flow (~ 1.15 ± 1.05

RI

Ma). The lava flow adjacent to Dixie Creek has not been dated; therefore, it is possible that they could be older than ~2.2 Ma. Based on visual inspection of post-flow

SC

weathering, it is unlikely to be younger than 100 ka. The estimated time-integrated rate of

NU

incision for Dixie Creek since emplacement of the uppermost Table Rock basalt flow on

MA

top of Lake Idaho sediments is 0.4 ± 0.3 m ky-1 (Table 2).

D

4.3. Topographic Metrics

TE

4.3.1 Normalized Channel Steepness (Ksn)

AC CE P

The upper Burnt River canyon coincides with a zone of high tributary and main channel Ksn values (Fig. 4). Ksn is low upstream of the upper canyon except in the western-most margin of the basin where the active Unity Valley fault system is crossed by streams (Personius, 2002). Tributary streams and the main stem of the Burnt River also display elevated Ksn values downstream of the Durkee Basin, where the river crosses back from the hanging wall to footwall block of the Durkee Fault.

4.3.2 Hypsometric Indices Sub-basins in the Burnt River catchment were examined for topographic evidence of tectonic activity. The seven largest catchments feeding the Burnt River exhibit three different hypsometric and slope characters (Fig. 6). Catchments 1, 2 and 3 have slightly steepened to nearly flat hypsometric curves. Mean slopes are between 15 and 20° over

ACCEPTED MANUSCRIPT the majority of each of these sub-basins. Catchment 2 contains a large area of lake beds capped by pediment surfaces which are likely responsible for the high percentage of area

PT

with low elevations and low slope. The hypsometric integrals for these three sub-basins

RI

range from 0.31 to 0.44. Catchments 4, 5 and 6 display much steeper hypsometric curves. Catchment 5 is only 3 km from the Durkee Basin, and catchment 6 flows directly

SC

into the Durkee Basin. These three catchments maintain the highest hypsometric integral

NU

values observed across the entire Burnt River basin (0.55 to 0.59). Catchments 5 and 6 have two distinct elevation-area peaks, which correspond to zones of both high and low

MA

elevations with low-gradient slopes (10 to 15°). Catchments 3 and 4 join the Burnt River at the steepest section along its longitudinal profile (Fig. 3 inset). The hypsometric curve

D

for catchment 7 is most similar to basins 1, 2 and 3; however, this basin has a higher

TE

hypsometric integral (0.48) and maintains a larger area at higher slope gradients (≥ 20°)

AC CE P

than catchments 1, 2 and 3. Catchment 7 is also the largest sub-basin in the lower half of the Burnt River basin. This catchment contains diverse bedrock, including two small Quaternary basaltic shield volcanoes (Cinder Butte and Table Rock; Fig. 3).

4.4. Soil Geochronology Hydraulic gold mining removed most terrace deposits and their overlying soils in both the upper and lower Burnt River canyons. One exception is a 7-m thick section above the Qt3L strath that contains both modern and buried soils, each with its own pedogenic carbonate horizon (Fig. 7). The lower, older soil contains a Btkb horizon with stage 3plus carbonate accumulation, while the upper (modern soil) exhibits only stage 2 Bk horizon development (Gile et al., 1965; Birkeland, 1999). Two mature soils; the presence

ACCEPTED MANUSCRIPT of thick weathering rinds (≥ 5 cm) on basalt cobbles, and the complete grussification of granite clasts are indicative of an old terrace deposit, which has experienced several Late-

RI

PT

Quaternary climatic fluctuations (Fig. 5A; Colman, 1981).

There is a robust soil stratigraphy in the Palouse Hills of eastern Washington (200 km

SC

north) where ages have been assigned to five late-Quaternary soils with the help of

NU

tephrochronology, OSL and radiocarbon dating (see McDonald et al., 2012, and references therein). The Qt3L soil along the Burnt River provides an excellent

MA

opportunity to correlate an eastern Oregon soil to the well-dated Palouse soil stratigraphy. Btk soil in the Qt3L deposit may correlate to the Washtucna Soil of the Palouse

TE

D

chronosequence (MIS 2, ~24 ka). The buried soil that contains the stage 3-plus Btkb horizon may correlate to the Devils Canyon Soil of the Palouse (MIS 4, ~60 to 70 ka;

AC CE P

McDonald et al., 2012; M. Sweeney, personal correspondence, 2015). This proposed soil chronostratigraphy, the presence of highly weathered basalt clasts, and grussification of granite clasts supports the 84.7 ± 13.94 ka OSL age collected from an intact fine-grained sandy interval directly beneath the lower, buried soil (Table 1; Fig. 7). The soils above the Qt3L strath likely demarcate intervals of reduced atmospheric silt loading and relative landscape stability which has been observed at sites in the Palouse Hills (Busacca et al., 1992; McDonald et al., 2012).

4.5. Active Faulting in the Durkee Basin The Durkee Basin separates the upper and lower canyons. The basin is filled with Miocene to Pliocene lake deposits truncated by a formerly active pediment surface (Fig.

ACCEPTED MANUSCRIPT 9). The pediment surface is now incised by transverse streams. Ashley (1995) mapped a normal fault on the southwest margin of the Durkee Basin. Yet, several pieces of

PT

evidence point toward the presence of a Quaternary fault that bounds the entire south side

RI

of the Durkee Basin. The evidence for this fault includes the following observations: (1) Triangular facets exist along a NW-SE striking trend, demarcating a linear

SC

mountain front (Fig. 9 inset).

NU

(2) Eight springs are mapped along the mountain front – piedmont interface on USGS 7.5-minute topographic quadrangle maps.

MA

(3) Outcrops of travertine found in three locations along the mountain front, indicate the presence of long-lived springs (Fig. 9).

TE

D

(4) Tributary streams to the Burnt River are significantly steeper along the southern limit of the Durkee Basin, as they cross the fault (Fig. 4) and normalized

AC CE P

values of channel steepness (ksn) derived from these tributaries decrease with distance from the Durkee Basin and its bounding fault. The steep and shallow stream gradients are within the same bedrock lithology (Nelson Marble) so lithological variation alone cannot explain observed changes in stream steepness. However, a normal fault actively dropping base level within the Durkee Basin would account for steep stream reaches within several km of the fault that then decrease farther from the fault. These areas with lower stream gradients may represent pre-faulting topography. (5) In the lower Burnt River canyon the Qt4L and Qt3L terrace profiles steepen in an upstream direction with increasing proximity to the Durkee Basin (Figs. 3B, 8). These terraces are on the footwall of the Durkee fault. Footwall uplift

ACCEPTED MANUSCRIPT typically accounts for 20 percent of total throw on extensional faults (e.g., Stein et al., 1988), with each successive displacement of the Durkee fault progressively

RI

PT

tilting terraces relative to the Burnt River, maximized closest to the fault.

A topographic swath profile constructed along the mountain front in the Durkee Basin

SC

indicates a decrease in elevation toward the ends of the 16 km long fault, with 0.5 km of

NU

footwall displacement (Fig. 10). The highest topography along a recently active normal fault is often observed toward the fault mid-section as maximum displacement typically

MA

occurs farthest from the fault tips (e.g. Dawers et al., 1993). The swath profile in the Durkee Basin shows such a signature with the greatest mountain-front relief located mid-

TE

D

fault and with relief tapering toward the fault tips (Fig. 10).

AC CE P

Fault-scaling relationships can be used to calculate the maximum expected moment magnitude (Mw) of an earthquake given slip along the entire surface trace of the Durkee fault following the Wells and Coppersmith (1994) regression equation: Mw = 5.08 + 1.16log(SRL),

(3)

where SRL is the surface rupture length, estimated at a maximum of 16 km for the Durkee fault. Using equation 3, a 6.5 Mw is the maximum expected moment magnitude for an earthquake on the Durkee fault.

5. Discussion The field investigation of fluvial terraces along the Burnt River reveals a two-fold story of river cutting and active faulting within the catchment. Incision by the Burnt River is

ACCEPTED MANUSCRIPT driven by two processes: (1) movement along a previously unmapped fault in the Durkee Basin, and (2) base-level adjustments tied to the Snake River. Prior to the field-mapping

PT

component of this study, we hypothesized that incision of the Snake River resulted in the

RI

propagation of a transient topographic signal into the Burnt River catchment, manifested through the broad convexity in the Burnt River longitudinal profile (Fig. 3 inset). As the

SC

Burnt River is a tributary to the Snake River upstream of Hells Canyon, a kinematic wave

NU

of incision from the cutting of Hells Canyon was hypothesized to have migrated through the Burnt River catchment, much like the transients studied by Larimer (2015).

MA

Following geomorphic mapping within the catchment, the aforementioned evidence indicates that movement on the previously unmapped Durkee fault is likely responsible

TE

D

for the convexity observed in the modern Burnt River longitudinal profile and incision of the Burnt River through the upper canyon. The lower canyon likely formed as a

AC CE P

combination of incision through the uplifting footwall of the Durkee fault combined with lowering of the Snake River at the uppermost end of Hells Canyon.

5.1. Terrace Genesis Model

The following model of terrace formation integrates terrace geochronology from this study and approximate ages for terraces linked with late-Quaternary climatic fluctuations. The model also makes a comparison between this study and published terrace ages from the Middle Fork of the Salmon River in central Idaho and the Boise River in western Idaho. The three dated terraces in the Burnt River catchment – Qt5L, Qt3L, and Qt5U – are concurrent with interglacial periods on the oxygen isotope curve (Fig. 11; Lisiecki

ACCEPTED MANUSCRIPT and Raymo, 2005). Qt5L corresponds to the latter part of MIS 3; Qt3L to late MIS 5 to early MIS 4, and Qt5U to late MIS 7. The geochronologic samples from each of the three

PT

dated terraces were extracted from sediments less than 1 m above their underlying

RI

bedrock strath. As such, we assume that the deposition of these basal terrace sediments occurred contemporaneously with beveling of the straths. This assumption is based on

SC

the terrace formation models discussed in Pazzaglia and Brandon, (2001), Wegmann and

NU

Pazzaglia (2002), Wegmann and Pazzaglia (2009), Pederson et al. (2013) and Pazzaglia, (2013). In our conceptual model, strath surfaces form during interglacial intervals (MIS

MA

3, late 5 and late 7) with the Burnt River potentially aggrading to a limited extent during glacial times (MIS 2, 4 and 6). It is worth noting that the terrace chronology constructed

AC CE P

region.

TE

D

along the Burnt River is the highest resolution dataset on fluvial terraces that exists in this

We assume that external forcing controls terrace formation in the Burnt River catchment. The incision resulting in the preservation of the Burnt River strath terraces above the main channel requires either a decrease in sediment load with constant stream power, or an increase in stream power with constant sediment supply. Late Quaternary climatic cycles provide the necessary environmental controls on the Burnt River catchment, either by sapping the river of alluvium during times of strath formation and degradation, or flooding the channel with colluvium, leading to aggradation. Uplift across the Burnt River catchment and movement on the Durkee fault throughout these climatic shifts allowed for continued river incision, stranding terraces as high as 90 m above the active channel (Fig. 3B). Examination of interstadial (MIS 3, 5 and 7) climate in the Inland

ACCEPTED MANUSCRIPT Northwest (INW) provides an internally consistent model for terrace formation within the

PT

Burnt River catchment.

RI

Climate and vegetation reconstructions indicate that MIS 3 was cool with moderate

SC

precipitation, comparable to the modern climate of the INW. Late MIS 5 was warmer

NU

and moister than today, and MIS 7 was a period of warming similar to MIS 5e when sea surface temperatures rose off the coast of Oregon and Washington and terrestrial

MA

precipitation increased (Fig. 11; Whitlock and Bartlein, 1997; Lyle et al., 2001; Grigg and Whitlock, 2002; Herring and Gavin, 2015). Pollen records indicate that the INW was

D

dominated by a Picea (spruce) and Pinus (pine) forest during MIS 3 and 5a, whereas the

TE

colder and drier MIS 2 and 4 intervals were dominated by shrubs (Artemisia) and grasses

AC CE P

(Poaceae) (Whitlock and Bartlein, 1997; Herring and Gavin, 2015). In contrast, MIS 3, 5a and 7 were times of forest expansion in the INW. There may have been dense, deeprooted vegetation along the hill slopes above the Burnt River during these intervals. Tree cover in the Burnt River catchment during MIS 3, 5a and 7 in combination with increased effective annual precipitation likely favored the creation and retention of soil and colluvium on hillslopes (e.g., Bull, 1991). Decreased hillslope sediment supply to the Burnt River during interstadials is consistent with a river in contact with bedrock during these periods, moving only a thin (1 to 3 m) veneer of alluvium. A moist climate ensured ample stream power with which to carve the extensive strath surfaces.

ACCEPTED MANUSCRIPT Glacial times in the Burnt River catchment (MIS 2, 4 and 6) were likely dry and dominated by shallow-rooted grasses and shrubs. Climatically-driven transitions in flora

PT

likely resulted in decreased hillslope stability and increased transport of hillslope material

RI

to the river. A drier climate would also hinder soil formation and result in decreased canopy interception of rainfall. As a result, processes such as rainwash erosion and

SC

sheetflow likely increased during glacial intervals, driving material down hillslopes (e.g.,

NU

Bull, 1991). More sediment in the main channel increased the resistance to fluvial incision, and the Burnt River began to aggrade. Other catchments in the Pacific

MA

Northwest experienced large fluxes glacio-fluvial sediment during glacial intervals (see Pazzaglia and Brandon, 2001). Lack of glaciers in the headwaters of the Burnt River

TE

D

may explain the absence of valley fill deposits from stadial aggradation periods. It seems likely that the Burnt River was raised above its strath due to the influx of sediment

AC CE P

beyond the transport capacity of the river; however, in comparison to formerly glaciated basins, aggradation along the Burnt River channel was much less significant.

Based on existing geochronology and aforementioned paleoclimate constraints, straths appear to have formed during interglacial periods and were likely incised through and captured in the landscape as strath terraces during the transition into glacial intervals, before valley aggradation began. Within the parameters of our model, previously undated terraces can now be associated with global cycles in marine δ18O, and as such, we have tied the formation and abandonment of Qt4L, Qt2L, Qt3U, Qt1L, Qt2U, and Qt1U to glacialinterglacial cycles as displayed in Fig. 11. This is the best-fit landscape-evolution model for available geochronology and paleoclimate data. Furthermore, sections of both the

ACCEPTED MANUSCRIPT Grande Ronde River (100 km to the south) and the Payette River (160 km to the east) are currently bedrock channels, supporting our model of interglacial strath formation on

RI

PT

Inland Northwest rivers (Pierce et al., 2011).

SC

This model also correlates with other terraces observed to the south and east of the Burnt

NU

River (Figs. 1, 11). The Middle Fork of the Salmon River appears to share a similar terrace chronostratigraphy to the Burnt River (Fig. 11; Meyer and Leidecker, 1999).

MA

However, the large errors inherent in the basalt weathering rind technique employed by Meyer and Leidecker (1999) makes any direct correlation problematic. Dates on terraces

D

along the Boise River in southwestern Idaho roughly correlate with terrace ages in the

TE

Burnt River catchment (Fig. 11; Othberg, 1994); however, these data lack the temporal

AC CE P

resolution required to draw direct correlations to the precisely dated Burnt River terraces.

5.2. Durkee Fault Mechanics

Our model of terrace development allows a previously undated terrace (Qt4L) to be used to determine the amount of slip along the Durkee fault. The paleo-longitudinal profile of Qt4L is defined at 5 locations where the strath is exposed along a 10 km reach downstream from its projected intersection with the Durkee fault. The reconstructed longitudinal profile of the fault-proximal portion of the Qt4L terrace increases in gradient in comparison to the Qt4L terrace farther downstream in the lower Burnt River canyon (Figs. 3B, 8). Extending the deformed and undeformed Qt4L profiles until they intersect with the fault plane provides an estimate of the total vertical footwall displacement of the

ACCEPTED MANUSCRIPT Durkee fault since terrace formation (Fig. 8). The reconstructed position of the Qt4L strath is 46 ± 4.6 m higher than it would be without fault deformation. The error on this

PT

measurement is provided by allowing for a 10% variance in terrace gradient. The

RI

estimated age for Qt4L from our regional terrace genesis model is ~55 ± 5 ka (MIS 4 to 3 transition), which results in a total footwall displacement rate estimate of 0.8 ± 0.1 mm

NU

SC

yr-1 for the Durkee fault (Table 4).

MA

Movement of the Durkee fault may not be the only explanation for increased terrace gradients in the upper reach of the lower Burnt River canyon. The modern profile of the

D

Burnt River is also over-steepened by ~0.11° along the same reach (river km 27 to 37;

TE

Fig. 19). The Burnt River is ~19 ± 1.9 m higher when it reaches the fault plane than if

AC CE P

the downstream gradient of 0.004 was maintained throughout the canyon (Fig. 8), which represents a 150% increase in channel gradient compared to the downstream reference gradient. Earthquake-related footwall uplift would steepen the river through this section; however, this steeper section of the profile could also be controlled by lithologic variations in rock hardness manifested as variations in channel gradient (Fig. 8). After exiting the Durkee basin, the river first flows across the Nelson Marble for ~10 km before encountering slates of the Weatherby Formation. The 10 km reach underlain by the Nelson Marble is steeper and less sinuous than the remainder of the lower Burnt River canyon (Fig. 3B). The different rock types along the lower Burnt River canyon (marble versus slate) may offer a first-order control on channel steepness. There is currently no way of determining whether or not some component of the observed steeper gradient is tectonic; however, it seems likely that the Burnt River is at a mostly pre-faulting profile

ACCEPTED MANUSCRIPT as there are no signs of Holocene surface ruptures along the Durkee fault. Employing this operating assumption, the modern downriver profile gradient is used to normalize this 10

RI

PT

km reach for possible bedrock control.

SC

Removing the potential increase in modern channel gradient resulting from rock-type

NU

variations from the Qt4L terrace results in 27 ± 5 m of potential tectonic over-steepening of the channel profile at the fault (Table 4). Integrating the modeled terrace age (55 ± 5

MA

ka) yields a revised footwall uplift rate of 0.5 ± 0.2 mm y-1. Assuming that footwall uplift accounts for ~20% of total fault slip (e.g. Stein et al., 1998), the maximum estimate

D

of fault displacement from the Qt4L terrace within the past 55 ± 5 ka is 2.5 ± 1.0 mm y-1

TE

(Table 4). This rate is comparable to the 0.01 – 2.8 mm y-1 rate of slip measured on an

AC CE P

extensional fault zone near Halfway, Oregon 50 km to the northeast (Essman, 2003).

5.3. Burnt River

The lower Burnt River channel and the reconstructed terrace longitudinal profiles maintain a similar gradient (0.004) to the modern stream (Fig. 3B). Channel steepness (gradient) normalized to basin area (Ksn) is greater for the lower Burnt River canyon and adjacent tributary streams than the observed general trend of streams in the upper half of the Burnt River catchment (Fig. 4). Ksn values in the lower canyon tend to be >150, while channel reaches in the upper catchment are generally <150 (Fig. 4). These steeper streams are in the lower 32 km of the Burnt River catchment, between the Durkee Basin and the river’s mouth. Slopes in the lower Burnt River canyon are steeper and relief is

ACCEPTED MANUSCRIPT higher than in the majority of the upper Burnt River catchment (Figs. 2B, C). Footwall uplift along the Durkee fault alone cannot account for these observations. Footwall uplift

PT

of a normal fault typically accounts for ~5% of the total fault throw at a distance of 10

RI

km from the fault plane (e.g. Stein et al., 1988). Over a distance of 30 km, the effect of slip along the Durkee fault should be negligible. Streams are consistently steep

SC

throughout the lower Burnt River catchment, rather than only adjacent to the Durkee fault.

NU

Therefore, some process other than slip along this fault is responsible for incision of the main channel and its tributary streams as observed along the lower Burnt River canyon.

MA

Two explanations for these observations include: (1) a regional signal of uplift across the

TE

D

Burnt River catchment; or (2) a drop in base level tied to incision along the Snake River.

AC CE P

The first explanation, regional uplift, could account for several observations made within the Burnt River catchment. Streams flowing across the hanging-wall side of the Durkee fault have incised through basin sediments (Morriss, 2015). An older Quaternary pediment surface preserved above the hanging wall of the Durkee fault serves as evidence that streams were graded to a local base level of erosion, which was then lowered – perhaps through regional uplift, or down cutting of the Burnt River transmitted through the lower canyon into the Durkee Basin. Tributary streams responded by cutting through this pediment surface. One explanation for the observed incision on the hangingwall side of the Durkee fault is uplift at a wavelength longer than the entire Durkee Basin. Wood and Clemens (2002) estimated the maximum elevation of Lake Idaho at 1100 m; however, within the Burnt River catchment late Miocene to Pliocene lacustrine sediments that are tentatively correlated to Lake Idaho exist to an elevation of at least 1,484 m

ACCEPTED MANUSCRIPT (Robyn, 1977). Applying age constraints of the Lake Idaho high-stand, the difference in maximum lake sediment elevations throughout the region represents a long-term uplift

RI

PT

rate of 0.1 ± 0.02 m kyr-1 (Table 2).

SC

The mechanism responsible for this observed uplift could be the aforementioned

NU

geodynamic processes, still active below eastern Oregon (delamination and sinking of a plutonic root; Hales et al., 2005; Darold and Humphreys, 2013). This estimate is only

MA

that – an estimate. It likely represents a minimum rate as lacustrine sediments are easily eroded. The overall deformation due to uplift in the region might be greater; however,

D

the time-averaged uplift rates are unlikely to be higher than ~0.3 m kyr-1. This upper

TE

bound on uplift in the region would place lakebeds at ~2300 m, or the elevation of

AC CE P

modern drainage divides between the John Day and Burnt rivers.

The second explanation, base-level drop tied to incision of the Snake River, is coupled with regional uplift. Seismic studies (Hales et al., 2005; Darold and Humphreys, 2013), geomorphic investigations (Larimer, 2015), and thermochronometric studies (Vogl et al., 2014) indicate dynamic interactions between the lower crust and mantle across NE-OR in the Miocene. Rivers in the Salmon River catchment continue to adjust to a Miocene uplift event (Larimer, 2015). It is possible a similar transient signal is present in the lower Burnt River canyon and the river has yet to adjust to the new base-level conditions imposed by the Snake River.

ACCEPTED MANUSCRIPT 6. Conclusions This study of the Neogene to Quaternary landscapes of eastern Oregon leads to the

PT

following conclusions based on mapped landforms, fluvial processes, and existing

RI

geology. Active normal faults within the Burnt River catchment provide a local base

SC

level and drive fluvial incision of the upper Burnt River canyon, creating 700 m of local relief in the process. The calculated incision rate for the fault controlled, upper Burnt

NU

River canyon based on 1:24,000 scale mapping and tephrochronology is 0.15 ± 0.03 m

MA

kyr-1. The Durkee Basin formed as a result of down-to-the-northeast movement on a previously unmapped normal fault, herein called the Durkee fault. The Burnt River

D

crosses this fault twice, creating two characteristically different fluvial reaches (the upper

TE

and lower Burnt River canyons). Major automobile and train traffic thoroughfares and a liquefied natural gas pipeline cross the Durkee fault. These transit and energy conduits

AC CE P

are threatened by potential seismic activity along this normal fault. Footwall uplift along the Durkee fault is responsible for over-steepened terraces in the uppermost lower Burnt River canyon. Incision rates within the undeformed reach of the lower Burnt River canyon range from 0.57 ± 0.2 to 0.21 ± 0.16 m ky-1.

New OSL ages and tephra geochronology on mapped strath terrace deposits within the Burnt River catchment indicate that during interglacial periods, horizontal incision outpaced vertical incision. The observed strath terraces were likely captured in the landscape through incision at the transition between interglacial and glacial periods. The Burnt River was likely raised above its strath during glacial periods due to channel

ACCEPTED MANUSCRIPT aggradation. However, the Burnt River did not aggrade significantly as fill terraces do not exist in the catchment. Extensive fill terraces likely did not form in the Burnt River

PT

basin, as hillslopes, not glaciers, were the first order control on catchment sediment

RI

supply. This genetic model of terrace development in eastern Oregon is consistent with the sparse river terrace data available from other regional fluvial systems. Additional

SC

river terrace studies from this region are warranted for refining the initial Inland

MA

geomorphology of the Burnt River basin.

NU

Northwest terrace genesis model for non-glaciated catchments based upon the

D

We document that continued incision throughout the Burnt River catchment not

TE

associated with slip along the Durkee fault may be due in part to long wavelength

AC CE P

regional uplift at a rate of ~ 0.1 ± 0.02 m kyr-1. This rate was determined by the elevation of late Miocene to Pliocene lake beds that are higher than the regional Lake Idaho high-stand elevation of 1,100 m. This uplift signal is ~ 30% of measured stream incision. It represents only a minimum estimate of uplift since the lake beds were deposited. Additional incision in the lower Burnt River canyon is likely tied to the regional base level that is set by the Snake River, which has cut a significant river gorge (Hells Canyon) downstream since 8 to 10 Ma. Further research in the region is warranted, as many questions regarding uplift and fluvial incision remain unresolved. This study represents only a preliminary dataset on the geomorphic landforms, paleotopographic datums, and rates of surface elevation change for eastern Oregon.

ACCEPTED MANUSCRIPT Acknowledgments The authors appreciate financial support provided by the U.S. Geological Survey,

PT

National Cooperative Geologic Mapping program Award Number G14AC00111 for the

RI

period May 1, 2014 to April 30, 2015. The views and conclusions contained in this

SC

document are those of the authors and should not be interpreted as necessarily representing the official policies either expressed or implied, of the U.S. Government.

NU

We also would like to thank Claire Vezie for her great support in the field and the

MA

insights offered by Skye Cooley. Additionally, without the exposures provided by 19th and 20th century gold miners, this study would have been nearly untenable. We benefited

D

from their landscape-modifying search for wealth. This manuscript was improved thanks

AC CE P

TE

to the comments and reviews offered by Dr. Lisa Ely and Dr. Kurt Othberg.

ACCEPTED MANUSCRIPT REFERENCES Allen, J., 1991. The case of the inverted auriferous paleotorrent—exotic quartzite gravels on Wallowa Mountain peaks. Oregon Geology, 53(5), 104-107.

SC

RI

PT

Ashley, R., 1995. Petrology and deformation history of the Burnt River Schist and associated plutonic rocks in the Burnt River Canyon area, northeastern Oregon, Geology of the Blue Mountains region of Oregon, Idaho, and Washington: Petrology and tectonic evolution of pre-Tertiary rocks of the Blue Mountains region: US Geological Survey Professional Paper 1438, Washington, D.C., pp. 457-496.

NU

Birkeland, P., 1999. Soils and Geomorphology. Oxford University Press, New York. 423 pp.

MA

Brooks, H.C., 1979a, Geologic map of the Huntington and part of the Olds Ferry quadrangles, Baker and Malheur Counties, Oregon: Oregon Department of Geologic and Mineral Industries Geologic Map Series GMS-13, scale 1:62,500.

TE

D

Brooks, H.C., 1979b, Geologic map of the Oregon part of the Mineral Quadrangle: Oregon Department of Geologic and Mineral Industries Geologic Map Series GMS-12, scale 1:62,500. Bull, W.B., 1990. Stream-terrace genesis: implications for soil development. Geomorphology, 3(3-4), 351-367.

AC CE P

Bull, W.B., 1991. Geomorphic responses to climatic change. Oxford Univresity Press, New York. 326 pp. Busacca, A.J., Nelstead, K.T., McDonald, E.V., Purser, M.D., 1992. Correlation of distal tephra layers in loess in the Channeled Scabland and Palouse of Washington State. Quaternary Research, 37(3), 281-303. Camp, V.E., Hanan, B.B., 2008. A plume-triggered delamination origin for the Columbia River Basalt Group. Geosphere, 4(3), 480-495. Colman, S.M., 1981. Rock-weathering rates as functions of time. Quaternary Research, 15(3), 250-264. Cook, K.L., Turowski, J.M., Hovius, N., 2013. A demonstration of the importance of bedload transport for fluvial bedrock erosion and knickpoint propagation. Earth Surface Processes and Landforms, 38(7), 683-695. Cowan, D.S., Reiners, P., 2004. Age and provenance of lower Tertiary fluvial strata, Elkhorn Mountains, E. Oregon. Geological Society of America Abstracts with Programs, 36(4), 34.

ACCEPTED MANUSCRIPT

PT

Darold, A., Humphreys, E., 2013. Upper mantle seismic structure beneath the Pacific Northwest: A plume-triggered delamination origin for the Columbia River flood basalt eruptions. Earth and Planetary Science Letters, 365, 232-242. Dawers, N.H., Anders, M.H., Scholz, C.H., 1993. Growth of normal faults: Displacement-length scaling. Geology, 21(12), 1107-1110.

RI

Dorsey, R.J., Lamaskin, T.A., 2008. Mesozoic collision and accretion of oceanic terranes in the Blue Mountains Province of northeastern Oregon: New insights from the stratigraphic record. Arizona Geological Society Digest, 22, 325-332.

SC

Essman, J.E., 2003. The Case for NE-SW extension in northeast Oregon [M.S. Thesis]. Oregon State University, Corvallis, 75 pp.

MA

NU

Ferns, M.L., McClaughry, J.D., 2013. Stratigraphy and volcanic evolution of the middle Miocene to Pliocene La Grande–Owyhee eruptive axis in eastern Oregon. In: Reidel, S.P., Camp, V.E., Ross, M.E., Wolff, J.A., Martin, B.S., Tolan, T.L., Wells, R.E. (Eds.), The Columbia River Flood Basalt Province. Geological Society of America Special Paper 497, Boulder, pp 401-427.

D

Finnegan, N.J., Dietrich, W.E., 2011. Episodic bedrock strath terrace formation due to meander migration and cutoff. Geology, 39(2), 143-146.

AC CE P

TE

Gallen, S.F., Pazzaglia, F.J., Wegmann, K.W., Pederson, J.L., Gardner, T.W., 2015. The dynamic reference frame of rivers and apparent transience in incision rates. Geology, 43(7), 623-626. Gile, L.H., Peterson, F.F., Grossman, R.B., 1965. The K Horizon: A master soil horizon of carbonate accumulations. Soil Science, 99(2), 74–82. Grigg, L.D., Whitlock, C., 2002. Patterns and causes of millennial-scale climate change in the Pacific Northwest during Marine Isotope Stages 2 and 3. Quaternary Science Reviews, 21(18), 2067-2083. Hales, T.C., Abt, D.L., Humphreys, E., Roering, J., 2005. A lithospheric instability origin for Columbia River flood basalts and Wallowa Mountains uplift in northeast Oregon. Nature, 438(7069), 842-845. Hancock, G.S., Anderson, R.S., 2002. Numerical modeling of fluvial strath-terrace formation in response to oscillating climate. Geological Society of America Bulletin, 114(9), 1131-1142. Herring, E.M., Gavin, D.G., 2015. Climate and vegetation since the Last Interglacial (MIS 5e) in a putative glacial refugium, northern Idaho, USA. Quaternary Science Reviews, 117, 82-95. Hooper, P.R., Camp, V.E., Reidel, S.P., Ross, M.E., 2007. The origin of the Columbia River flood basalt province: Plume versus nonplume models. In: Foulger, G.R.,

ACCEPTED MANUSCRIPT and Jordy, P.M. (Eds), Plates, Plumes, and Planetary Processes. Geological Society of America Special Paper 430, Boulder, pp. 635-668.

PT

Johnson, D., Hooper, P., Conrey, R., 1999. XRF analysis of rocks and minerals for major and trace elements on a single low dilution Li-tetraborate fused bead. Advances in X-ray Analysis, 41, 843-867.

RI

Keller, E.A., Pinter, N., 2002. Active Tectonics: Earthquakes and Landscape. Prentice Hall, Upper Saddle River, 362 pp.

SC

Kirby, E., Whipple, K., 2001. Quantifying differential rock-uplift rates via stream profile analysis. Geology, 29(5), 415-418.

NU

Kirby, E., Whipple, K.X., Tang, W., Chen, Z., 2003. Distribution of active rock uplift along the eastern margin of the Tibetan Plateau: Inferences from bedrock channel longitudinal profiles. Journal of Geophysical Research: Solid Earth, 108(B4).

MA

Kuehn, S.C., Negrini, R.M., 2010. A 250 ky record of Cascade arc pyroclastic volcanism from late Pleistocene lacustrine sediments near Summer Lake, Oregon, USA. Geosphere, 6(4), 397-429.

TE

D

Larimer, J.E., 2015. Implications for the timing and driving mechanisms of topograhic evolution of the inland northwest, [M.S. Thesis]. University of Idaho, Moscow, 63 pp.

AC CE P

Lavé, J., Avouac, J.P., 2000. Active folding of fluvial terraces across the Siwaliks Hills, Himalayas of central Nepal. Journal of Geophysical Research, 105(B3), 57355770. Lisiecki, L.E., Raymo, M.E., 2005. A Pliocene‐Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography, 20(1), PA1003. Lyle, M., Heusser, L., Herbert, T., Mix, A., Barron, J., 2001. Interglacial theme and variations: 500 ky of orbital forcing and associated responses from the terrestrial and marine biosphere, US Pacific Northwest. Geology, 29(12), 1115-1118. Markey, B., Bøtter-Jensen, L., Duller, G., 1997. A new flexible system for measuring thermally and optically stimulated luminescence. Radiation Measurements, 27(2), 83-89. McCaffrey, R., King, R.W., Payne, S.J., Lancaster, M., 2013. Active tectonics of northwestern US inferred from GPS‐derived surface velocities. Journal of Geophysical Research: Solid Earth, 118(2), 709-723. McDonald, E.V., Sweeney, M.R., Busacca, A.J., 2012. Glacial outburst floods and loess sedimentation documented during Oxygen Isotope Stage 4 on the Columbia Plateau, Washington State. Quaternary Science Reviews, 45, 18-30.

ACCEPTED MANUSCRIPT

PT

Meyer, G.A., Leidecker, M.E., 1999. Fluvial terraces along the Middle Fork Salmon River, Idaho, and their relation to glaciation, landslide dams, and incision rates: A preliminary analysis and river-mile guide. In: Hughes S.S., Thackray, G.D. (Eds.), Guidebook to the Geology of Eastern Idaho. Idaho Museum of Natural History, Pocatello, pp. 219-235.

RI

Miller, S.R., Sak, P.B., Kirby, E., Bierman, P.R., 2013. Neogene rejuvenation of central Appalachian topography: Evidence for differential rock uplift from stream profiles and erosion rates. Earth and Planetary Science Letters, 369, 1-12.

SC

Mix, H.T., Mulch, A., Kent-Corson, M.L., Chamberlain, C.P., 2011. Cenozoic migration of topography in the North American Cordillera. Geology, 39(1), 87-90.

NU

Montgomery, D.R., 2004. Observations on the role of lithology in strath terrace formation and bedrock channel width. American Journal of Science, 304(5), 454476.

MA

Morriss, M.C., 2015. Dynamic and tectonic landscapes in eastern Oregon reveal Neogene to Quaternary rearrangement of topography [M.S. Thesis]. North Carolina State University, Raleigh, 109 pp.

TE

D

Murray, A.S., Wintle, A.G., 2000. Luminescence dating of quartz using an improved single-aliquot regenerative-dose protocol. Radiation Measurements, 32(1), 57-73.

AC CE P

Othberg, K.L., 1994. Geology and geomorphology of the Boise Valley and adjoining areas, western Snake River Plain, Idaho. Idaho Geological Survey Bulletin, 29, 54. Pazzaglia, F.J., Gardner, T.W., Merritts, D.J., 1998. Bedrock fluvial incision and longitudinal profile development over geologic time scales determined by fluvial terraces. In Wohl, E., Tinkler, K. (Eds.), Rivers Over Rock: Fluvial Processes in Bedrock Channels. American Geophysical Union, Geophysical monograph Series, Washington, D.C., pp. 207-236. Pazzaglia, F.J., Brandon, M.T., 2001. A fluvial record of long-term steady-state uplift and erosion across the Cascadia forearc high, western Washington State. American Journal of Science, 301(4-5), 385-431. Pazzaglia, F. 2013. Fluvial terraces. In: Shroder, J., Wohl, E. (Eds.), Treatise on Geomorphology, Volume 9, Fluvial Geomorphology: Academic Press, San Diego, pp. 379-412. Pederson, J.L., Anders, M.D., Rittenhour, T.M., Sharp, W.D., Gosse, J.C., Karlstrom, K.E., 2006. Using fill terraces to understand incision rates and evolution of the Colorado River in eastern Grand Canyon, Arizona. Journal of Geophysical Research: Earth Surface, 111(F2).

ACCEPTED MANUSCRIPT Pederson, J.L., Cragun, W.S., Hidy, A.J., Rittenour, T.M., Gosse, J.C., 2013. Colorado River chronostratigraphy at Lee’s Ferry, Arizona, and the Colorado Plateau bull’s-eye of incision. Geology, 41(4), 427-430.

PT

Personius, S.F. 2002. Faults near Unity Valley, in Quaternary fault and fold database of the United States: U.S. Geological Survey website, http://earthquake.usgs.gov/hazardes/qfaults, [accessed 11/05/2014].

SC

RI

Pierce, J.L, Meyer, G.A., Rittenour, T., 2010, The relation of Holocene fluvial terraces to changes in climate and sediment supply, South Fork Payette River, Idaho, Quaternary Science Reviews, 30(5-6), 1-18.

NU

Pike, R.J., Wilson, S.E., 1971. Elevation-relief ratio, hypsometric integral, and geomorphic area-altitude analysis. Geological Society of America Bulletin, 82(4), 1079-1084.

MA

Prescott, J.R., Hutton, J.T., 1994. Cosmic ray contributions to dose rates for luminescence and ESR dating: large depths and long-term time variations. Radiation Measurements, 23(2), 497-500.

D

Robyn, T.L., 1977. Geology and petrology of the Strawberry Volcanics, NE Oregon [Ph.D. Thesis]. University of Oregon, Eugene, 224 pp.

AC CE P

TE

Rodriguez, S., Sen, G., 2013. Eruption of the Grande Ronde Basalt lavas, Columbia River Basalt Group: Results of numerical modeling. In: Hooper, P.R., Camp, V.E., Reidel, S.P., Ross, M.E. (Eds.), The Columbia River Flood Basalt Province: Geological Society of America Special Paper 497, Geologic Society of America, Boulder, pp. 259-272. Schwanghart, W., Scherler, D., 2014. Short Communication: TopoToolbox 2-MATLABbased software for topographic analysis and modeling in Earth surface sciences. Earth Surface Dynamics, 2(1), 1-7. Sears, J.W., 2014. Tracking a big Miocene river across the Continental Divide at Monida Pass, Montana/Idaho. In: Shaw, C.A. Tikoff, B. (Eds.), Exploring the Northern Rocky Mountains: Geological Society of America Field Guide 37. Geological Society of America, Boulder. pp. 89–99. Silberling, N.J., Jones, D., Blake, M., Howell, D., 1987. Lithotectonic Terrane Map of the Western Conterminous United State: Miscellaneous Field Studies Map 1874-C, U.S. Geological Survey, Washington, D.C., scale 1:2,500,000. Stein, R., King, G.C., Rundle, J., 1988. The growth of geological structures by repeated earthquakes, 2: Field examples of continental dip-slip faults. Journal of Geophysical Research, 93(B11), 13,319-313,331. Strahler, A.N., 1952. Hypsometric (area-altitude) analysis of erosional topography. Geological Society of America Bulletin, 63(11), 1117-1142.

ACCEPTED MANUSCRIPT Srivastava, P., Brook, G.A., Marais, E., 2005. Depositional environment and luminescence chronology of the Hoarusib River Clay Castles sediments, northern Namib Desert, Namibia. Catena, 59(2), 187-204.

PT

Van Tassell, J., Ferns, M., McConnell, V., Smith, G., 2001. The mid-Pliocene Imbler fish fossils, Grande Ronde Valley, Union County, Oregon , and the connection between Lake Idaho and the Columbia River. Oregon Geology, 63, 77-96.

SC

RI

Vallier, T., 1998. Islands and Rapid. Confluence Press, Lewiston, Confluence Press, 151 pp.

NU

Vogl, J.J., Min, K., Carmenate, A., Foster, D.A., Marsellos, A., 2014. Miocene regional hotspot-related uplift, exhumation, and extension north of the Snake River Plain: Evidence from apatite (U-Th)/He thermochronology. Lithosphere, 6(2), 108-123.

MA

Wegmann, K.W., Pazzaglia, F.J., 2002. Holocene strath terraces, climate change, and active tectonics: The Clearwater River basin, Olympic Peninsula, Washington State. Geological Society of America Bulletin, 114(6), 731-744.

D

Wegmann, K.W., Pazzaglia, F.J., 2009. Late Quaternary fluvial terraces of the Romagna and Marche Apennines, Italy: climatic, lithologic, and tectonic controls on terrace genesis in an active orogen. Quaternary Science Reviews, 28(1), 137-165.

AC CE P

TE

Wells, D.L., Coppersmith, K.J., 1994. New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement. Bulletin of the seismological Society of America, 84(4), 974-1002. Wheeler, H.E., Cook, E.F., 1954. Structural and stratigraphic significance of the Snake River capture, Idaho-Oregon. The Journal of Geology, 525-536. Whitlock, C., Bartlein, P.J., 1997. Vegetation and climate change in northwest America during the past 125 kyr. Nature, 388(6637), 57-61. Williams, S., 1994. Late Cenozoic tephrostratigraphy of deep sediment cores from the Bonneville Basin, northwest Utah. Geological Society of America Bulletin, 106(12), 1517-1530. Wobus, C., Whipple, K.X., Kirby, E., Snyder, N., Johnson, J., Spyropolou, K., Crosby, B., Sheehan, D., 2006. Tectonics from topography: Procedures, promise, and pitfalls. Geological Society of America Special Papers, 398, 55-74. Wood, S.H., Clemens, D.M., 2002. Geologic and tectonic history of the western Snake River Plain, Idaho and Oregon. Tectonic and Magmatic Evolution of the Snake River Plain Volcanic Province. Idaho Geological Survey Bulletin, 30, 69-103. Zaprowski, B.J., Evenson, E.B., Pazzaglia, F.J., Epstein, J.B., 2001. Knickzone propagation in the Black Hills and northern High Plains: A different perspective on the late Cenozoic exhumation of the Laramide Rocky Mountains. Geology, 29(6), 547-550.

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Table 1. Optically stimulated luminescence (OSL) data and ages from Burnt River terrace deposits.

Th (p pm )

K ( w t. % )

D os e R at e ( G y ky

Ali quo De( t Gy) no. 2

PT

U (p pm )

RI

De pt h (m )

Wa ter Con tent (%)

O O SL IS Ag 4 e (k yr) 3

NU

SC

Sa Si Elev Gr mpl te ation ain e ID s (m siz asl) e (µ m)

-

MA

1 1

Qt 963 6U

-

-

-

Qt 1028 4U

15 018 0

16

1.7 9± 0.2 5

BR S8

Qt 732 5L

12 518 0

4

1.3 1± 0.1 6

BR S9* BR S12

Qt 771 4L Qt 773 3L

-

-

-

12 525 0

5

1.5 9± 0.2

AC CE P

TE

D

BR S14 * BR S17

-

-

) -

-

4.5 1. 1. 16 5 ± 2 89 0.8 ± 7 0. 16 2.4 1. 1. 12 1 ± 1 62 0.5 ± 9 0. 13 - -

-

-

-

-

154. 8 ± 0± 4 12.8

81. 5a 6± 9.6

62.7 8 ± ± 4 6.5

38. 3 7± 5.1

-

-

-

-

2.7 1. 1. 8 152. 8 ± 84. 5a 6± 2 8 3± 4 7± 0.7 ± 21.8 13. 3 0. 9 15 * BRS14 and BRS9 did not contain enough quarts to return an OSL age. 1 Background dose rate (De) in Grays per 1000 yrs. 2 Equivalent dose rate (De) in Grays. 3 Optically stimulated luminescence age with error in 1000s of years before present.

ACCEPTED MANUSCRIPT 4

Correlative Oxygen Isotope Stage

Incision rate (m/kyr)

201 ± 47

6

29.6 ± 2

0.15 ± 0.03

5a

26.2 ± 2

0.57 ± 0.2

84.7 ± 13.9

5a

44 ± 2

0.52 ± 0.1

MA

84.7 ± 13.9 – 38.7 ± 5.1

84.7 ± 13.9

5a

48 ± 2

0.57 ± 0.1

Qt5L – modern channel

38.7 ± 5.1

3

8±2

0.21 ± 0.06

38.7 ± 5.1

3

12 ± 2

0.31 ± 0.06

AC CE P

Qt5L – strath below mining sediments

TE

Qt3L – strath below mining sediments

D

Qt3L – Qt5L Qt3L – modern Channel

OISa

a

SC

Qt5U – modern channel

Age (ka)

Vertical Incision Distance(m)

NU

Incision rate calculation endpoints

RI

PT

Table 2. Incision and uplift rates calculated within the Burnt River catchment. The Qt3L terrace incision rate was calculated based upon the vertical separation between Qt5L and Qt3L, not between Qt3L and the modern channel. The uppermost Lake Idaho deposits have been used in two places to measure overall incision and uplift, based on the Wood and Clemens (2002) hypothesis for an 1,100 m lake high stand between 3.8 and 2 Ma.

Age Source Paoha Island tephrab

OSL OSL OSL OSL

OSL

Table Rock Basalt flow (Dixie Creek)

1500 ± 700

-

470 ± 12

0.4 ± 0.03

Ferns and McClaughry, (2013)

Lake Deposits at mouth of BR (L)

4000 ± 900

-

216 ± 12

0.05 ± 0.01

Wood and Clemens (2002)

Uplift rate (m/kyr)

Source

0.1 ± 0.02

Robyn, (1977)

Uplift Indicator

Age (ka)

OISa

Change in Lake Bed Elevations (m)

Mean change of Lake deposits

4000 ± 900

-

384 ± 15

OIS = Oxygen Isotope Stage. b For tephrochronology see Kuehn and Negrini (2010)

ACCEPTED MANUSCRIPT Table 3. Available well data from the lower Burnt River canyon, indicating aggradation of the main channel, potentially in response to 19th and early 20th century mining activities. An average alluvium depth of 3.9 ± 0.4 m (assuming a 10% error) was used to correct incision rate calculations (Table 2).

PT

Elevation (m) 720 661.8

RI

Longitude (°) -117.3423 -117.3033

Latitude (°) 44.48494 44.381505

Proximal river Elevation (m) 712.6 658.8

Estimated Alluvial fill (m) 4.8 3

SC

Oregon Well ID Bake 52289 Bake 52020

Depth to Bedrock (m) 12.2 6

0.8 ± 0.1

27 ± 5.2

55 ± 5

0.5 ± 0.21

-

2.0 ± 0.84

-

2.5 ± 1.0

-

D

55 ± 5

AC CE P

Hanging wall displacement (4 x footwall) Total Fault Displacement (5 x footwall)

Rate of displacement (mm y-1)

46 ± 4.6

TE

Footwall Displacement Normalized* Footwall offset

Age (ka)

MA

Offset (m)

NU

Table 4. Displacement of the Durkee fault as calculated by deformation of Qt4L terraces. The minimum age for the Durkee fault has been calculated using the estimated maximum displacement of the footwall (Fig. 8) and the slip rate reported below.

-

Maximum footwall Displacement Durkee fault (mm) Slip rate (mm/yr) Minimum Fault age (Myr) 500000 0.5 ± 0.2 1.0 ± 0.4 * Normalization was an attempt to subtract the effects of more resistant bedrock beneath the Burnt River. See Discussion section for details.

ACCEPTED MANUSCRIPT

PT

FIGURE CAPTIONS

NU

SC

RI

Figure 1. Inset. Geographic location of the main map in relation to the Cascadia Subduction Zone. Main Map. Burnt River catchment (bold polygon) in the context of regional topography and drainage networks. The thin red line approximates the boundary between the North American craton and accreted terranes demarcated by the 87Sr/86Sr 0.706 isopleth (Leeman et al., 1992). The bulls-eye of high-standing topography described in Hales et al. (2005) and by Darold and Humphreys (2013) is outlined by the gray dashed line. Abbreviations used: Durkee fault (DF); Clarks Creek Fault (CCF1); for the Conner Creek Fault (CCF2).

AC CE P

TE

D

MA

Figure 2. Topographic metrics of the Burnt River catchment A. Shaded relief map of the Burnt River catchment. The largest communities in the catchment (Unity, Durkee, and Huntington) are labeled. The Durkee Basin and its bounding fault (the Durkee fault) divide the upper and lower Burnt River canyons. B. Relief within the Burnt River catchment. Relief was calculated using a 100 m moving window with the focal range routine in ArcGIS. Similar to slopes, areas of high relief are isolated to the upper and lower Burnt River canyons. C. Catchment slopes. A large portion of the steeper slopes are located in the lower half of the basin with the steepest slopes in the upper Burnt River canyon, located just upstream from the Durkee fault. Elevation, slope, and relief were calculated from the 10 m resolution National Elevation Dataset.

Figure 3. Inset. Longitudinal profile of the Burnt River from its confluence with the Snake River (Brownlee Reservoir) to its headwaters. The upper Burnt River canyon is notable as the river steepens significantly in this reach. The lower Burnt River canyon exhibits a consistently steep gradient. Both the lower and upper Burnt River canyons are steeper than the section of river just above the upper canyon. Main Figure. Correlations between Burnt River terraces in the upper (A) and lower canyons (B). The solid black line is the longitudinal profile extracted from the 10 m resolution National Elevation Dataset that was smoothed in SigmaPlot using a 1st-order polynomial LOESS tricube weighting routine with a kernel 0.1 times the river length in order to remove artificial steps from the digital elevation dataset. Terrace locations are plotted using elevation data from differential GPS field measurements. The spatial location of different geologic units are denoted on the map-view of the Burnt River channel planform shown above the longitudinal profiles for both the upper and lower canyons.

Figure 4. Normalized channel steepness (Ksn) for the Burnt River catchment. Values were smoothed using a 100 m moving window following the method outlined in Miller et al. (2013). Warmer colors represent areas in which streams are steeper (normalized for drainage area). The “hot spot” of steeper channels coincident with the upper Burnt River canyon is likely due to the presence of the previously unmapped Durkee fault at the mouth of the canyon, providing a firstorder control on local base level. The Burnt River catchment downstream from the Durkee Basin

ACCEPTED MANUSCRIPT

PT

still contains steeper streams than exist in the upper half of the catchment, indicating that the Durkee fault is not the only control on fluvial channel evolution in the basin. Ksn values were calculated using the MATLAB-based TopoToolbox (Schwanghart and Scherler, 2014).

MA

NU

SC

RI

Figure 5. Terrace deposit photos. (A) Basalt weathering rinds provid a relative chronometer of terrace age in the lower Burnt River canyon. (B) Hydraulic gold mining often removed all terrace deposits and colluvial cover, leaving the strath surface exposed. The apparent tilt of the strath surface is due to the dip of the underlying limestone units. (C) Several I-84 road cuts expose strath surfaces, terrace deposits and colluvial cover, as depicted in this photograph near milepost 336. (D) This Qt5U terrace deposit was identified in a mining cut, ~5 m above the Qt5 strath and ~12 m above the Qt6U strath (both identified by the dashed red lines). The 201 ± 45 ka Paoha Island tephra was found in these terrace deposits. (E) The tephra is interbedded with fine-grained fluvial overbank deposits atop coarser, axial-channel gravels (not in image). It was identified by staff at the WSU tephrochronology laboratory as the Paoha Island tephra from Mono Lake, California. Subsequent research suggests that the Paoha Island tephra is 201 ± 47 ka (Kuehn and Negrini, 2010).

AC CE P

TE

D

Figure 6. Topographic metrics for tributary catchments along the length of the Burnt River, including normalized hypsometry, mean slope, normalized elevation, and hypsometric integral (HI). Plots illustrate different levels of landscape adjustment to a transient signal and tectonic activity. Hypsometry and hypsometric integral calculations indicate catchments likely in equilibrium (1 and 2) and catchments in disequilibrium (3 – 7). Elevation and slope plots help to identify high elevation areas of low-slope that have not yet adjusted to stream incision driven by either local or regional base level changes. Figure 7. An outcrop of the Qt3L terrace in the lower Burnt River canyon that contains two distinct soils above terrace gravels [location: N 44.4731°, W 117.3336°]. Post-gravel stratigraphy represents two episodes of loess accumulation and soil development. The lower stage 3 to 3+ soil carbonate (Btkb) horizon likely correlates to the Btk horizon of the Devils Canyon Soil (MIS 4, ~60 to 70 ka) in the Palouse Hills; while the upper Btk soil carbonate horizon is correlative to the Washtucna Soil Btk horizon (MIS 2, ~24 ka) of the Palouse soil chronostratigraphy (McDonald et al., 2012). The degree of soil development observed at this exposure is consistent with the OSL age of 84.7 ± 13.94 ka obtained from the flood over bank deposits just below the Btkb horizon.

Figure 8. Terrace and modern channel gradients for the lower Burnt River canyon. Terraces near the Durkee fault are over-steepened with respect to downstream counterparts. The deformed segment of the Qt4L terrace is ~ 46 m higher at the fault plane than the projection of the undeformed section of the same terrace. This deformation indicates that the Qt4L terraces have been over steepened by accumulated throw on the Durkee fault. Additionally, the upper 10 km of the modern Burn River is ~ 19 m higher at the Durkee fault plane than it would be if it maintained the same gradient as the downstream section. It is possible that some of this deflection is due to a rock type change from the Nelson Marble into the Weatherby Formation, or because both stream power and available bedload tools likely decrease as the Burnt River traverses the Durkee Basin.

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

Figure 9. Inset Photo. View to the south-southeast across the mountain front defined by the Durkee fault. An elongate outcrop of travertine now forms a ridge on the hanging wall side of the fault (photographed from N 44.571947°, W 117.529318°). The fluids that deposited the travertine ridge flowed through a paleochannel, likely draining the mountain front. With time and preferential erosion of less-resistant late Miocene-to-Pliocene lacustrine sediments, the former paleo-low is now a high-standing ridge. Main Figure. Geologic map of western edge of the Durkee Basin and lower three kilometers of the upper Burnt River canyon. Miocene to Pliocene lacustrine units were planed into a pediment surface, and cut again by modern channels. A ridge of travertine marks a paleo-channel that now stands as a topographic high due to differential erosion of the underlying lacustrine units. Location “A” marks the site where the inset photograph was taken.

AC CE P

TE

D

MA

Figure 10. (A) Hillshade image of the Durkee Basin with the location of a 10 km wide topographic swath profile that parallels the Durkee fault. The hillshade was constructed using the 10 m resolution National Elevation Dataset and tools native to ArcGIS. (B) Topographic swath data, showing minimum, mean, and maximum elevations along the 10 km wide section. The dashed line represents maximum elevations, a proxy for displacement along the Durkee fault, indicating a total footwall displacement of at least 0.5 km (indicated by the black vertical line). The thick black line is the approximate location of the Durkee fault, with a length of 16 km. The ratio between maximum displacement (Dmax) and fault length (L) is 0.031. Abbreviations include: TBG – True Blue Gulch; PC – Powell Creek; and HC – Hollowfield Canyon.

Figure 11. Inferred terrace forming intervals for the Burnt, Salmon, and Boise Rivers. These two rivers provide the only published constraints on terrace formation in eastern Oregon and westcentral Idaho. Strath formation appears to take place during warmer, moist interglacial intervals when stream behavior favors enhanced horizontal incision (valley widening and strath cutting) versus vertical incision, which either occurs at the transition from interglacial to glacial or during glacial periods. The oxygen isotope curve is from Lisiecki and Raymo (2005).

AC CE P

Fig. 1

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Fi. 2

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Fig. 3

AC CE P

Fig. 4

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Fig. 5

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Fig. 6

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

Fig. 7

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

Fig. 8

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Fig. 9

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Fig. 10

AC CE P

Fig. 11

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Highlights

TE

D

MA

NU

SC

RI

PT

7 new Late-Quaternary Terraces in NE Oregon were mapped Strath terrace deposits were dated using tephrochronology and OSL dating Landscape position of terraces indicate faulting at a rate of 2.5 ± 1.0 mm yr-1 Incision rates ranged from 0.57 ± 0.2 to 0.21 ± 0.16 m ky-1 Channel steepness appears largely controlled by faulting

AC CE P

    