Document Actions

IJEGE-11_BS-Staley-et-alii

background image
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
759
DOI: 10.4408/IJEGE.2011-03.B-083
OBSERVATIONS OF DEBRIS FLOWS AT CHALK CLIFFS, COLORADO,
USA: PART 2, CHANGES IN SURFACE MORPHOMETRY FROM TERRE-
STRIAL LASER SCANNING IN THE SUMMER OF 2009
Dennis M. STALEY
(*)
, Thad A. WASKLEWICZ
(**)
, Jeffrey A. COE
(*)
, J
ason
W. KEAN
(*)
,
Scott W. McCOY
(***)
& Greg E. TUCKER
(***)
(*) United States Geological Survey, Landslide Hazards Program, Golden, Colorado, USA
(**) East Carolina University, Department of Geography, Greenville, North Carolina, USA
(***) University of Colorado – Boulder, Department of Geology, Boulder, Colorado, USA
K
ey
words
: debris flow, terrestrial laser scanning, morpho-
metry, channel
INTRODUCTION
Runoff-generated debris flows are common in
sparsely vegetated or recently burned steeplands
throughout the world (b
eRti
et alii, 1999; C
annon
et alii, 2001a; C
annon
et alii, 2001b; C
annon
et alii,
2003; G
odt
& C
oe
, 2007; m
C
a
Rdell
et alii, 2007;
C
oe
et alii, 2008; s
anti
et alii, 2008). Relative to
flows that mobilize from a discrete landslide source,
the mechanisms that contribute to the initiation and
propagation of debris flows produced during runoff
events are less understood (C
annon
et alii, 20003;
b
eRti
& s
imoni
, 2005; C
oe
et alii, 2008; m
C
C
oy
et
alii, 2010). Runoff-generated debris flow occurs in
response to surface flow produced during a rainstorm.
Overland flow initiates as rainfall intensity exceeds in-
filtration capacity. Surface runoff then entrains materi-
al from the hillslopes forming rills and gullies (G
abet
,
2003; m
oody
& k
inneR
, 2006; s
Hakesby
& d
oeRR
,
2006; G
abet
& b
ookteR
, 2007). Although hillslopes
provide an important source of material, it has been
recognized that a significant portion of the debris-flow
volume is generated by erosion of channel fill (C
en
-
deRelli
& k
ite
, 1998; b
ovis
& J
akob
, 1999; J
akob
et
alii, 2005; s
anti
et alii, 2008). Mobilization of mate-
rial stored in the channel can be considered to be the
product of shear forces applied on the bed by the flow,
impulsive loading, liquefaction of channel fill, bank
ABSTRACT
High resolution topographic data that quantify
changes in channel form caused by sequential debris
flows in natural channels are rare at the reach scale.
Terrestrial laser scanning (TLS) techniques are utilized
to capture morphological changes brought about by a
high-frequency of debris-flow events at Chalk Cliffs,
Colorado. The purpose of this paper is to compare and
contrast the topographic response of a natural chan-
nel to the documented debris-flow events. TLS sur-
vey data allowed for the generation of high-resolution
(2-cm) digital terrain models (DTM) of the channel.
A robust network of twelve permanent control points
permitted repeat scanning sessions that provided mul-
tiple DTM to evaluate fine-scale topographic change
associated with three debris-flow events. Difference
surfaces from the DTM permit the interpretations of
spatial variations in channel morphometry and net
volume of material deposited and eroded within and
between a series of channel reaches. Each channel
reach experienced erosion, deposition, and both net
volumetric gains and losses were measured. Analysis
of potential relationships between erosion and depo-
sition magnitudes yielded no strong correlations with
measures of channel-reach morphometry, suggesting
that channel reach-specific predictions of potential
erosion or deposition locations or rates cannot be ad-
equately derived from statistical analyses of pre-event
channel-reach morphometry.
background image
D.M. STALEY, T.A. wASkLEwICZ, J.A. COE, J.w. kEAN, S.w. McCOY & G.E. TUCkER
760
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
volumetric changes from multiple debris flows. The
Chalk Cliffs have one of the highest debris-flow fre-
quencies in Colorado, where 1-4 debris flows are
typically recorded each summer (C
oe
et alii, 2008;
m
C
C
oy
et alii, 2010). Here, we present results from
four TLS surveys at Chalk Cliffs: the first was on 28
May 2009 and documented the channel configuration
prior to the onset of the summer debris flow season;
three additional surveys conducted during the summer
documented changes in channel form associated with
individual debris-flow events.
The purpose of this paper is to document the topo-
graphic changes that occurred in the stream channel as
a response to each debris-flow event. We compare and
contrast the morphologic response of the channel by
calculating 1) the net volume of material eroded from
the channel, 2) the net volume of material deposited in
the channel, and 3) the net overall volumetric change
that occurred in the channel reach. In addition, we an-
alyze the changes in mean channel-reach gradient and
surface roughness associated with each debris-flow
event, and attempt to predict net volumetric change as
a function of reach-scale morphometry.
STUDY AREA
The Chalk Cliffs study basin (Fig. 1) is a 0.3 km
2
watershed composed of a hydrothermally altered
quartz-monzonite (m
illeR
et alii, 1999). Primary geo-
morphic units in the basin include exposed bedrock
and sandy colluvium. Bedrock is exposed in approxi-
mately 60 percent of the basin, with gradients ranging
from approximately 80 percent to vertical. The sandy
colluvium occupies the remaining 40 percent of the
watershed, with gradients ranging between 50 and 100
percent (C
oe
et alii, 2008). Frequent rockfall and dry-
ravel processes contribute material to the channel. De-
bris flows initiate in both the east and west channels
(Fig. 1) in response to bursts of high intensity rainfall
typically associated with short-duration convective
rainstorms (C
oe
et alii, 2008). These storms have pro-
duced debris flows of varying magnitudes, with some
flows terminating a short distance from the confluence
of the east and west channels (m
C
C
oy
et alii, 2010),
and others traveling the entire 0.6- km length of the
channel and depositing as a fan at the confluence with
Chalk Creek (C
oe
et alii, 2008).
Here, we analyze a non-vegetated portion of the
channel beginning just below the lowermost cliff band
failure, and headward migration of knickpoints (b
ovis
& d
aGG
, 1992; e
GasHiRa
et alii, 2001; H
unGR
et alii,
2005). Channel deposition will occur when friction
increases along flow margins and internal pore-fluid
pressures diminish (m
aJoR
, 2000).
While there is a growing body of work on debris-
flow entrainment from single events, there has been
little research regarding the spatial patterns of both
channel erosion and deposition from multiple events.
Data related to channel erosion, deposition and volu-
metric growth of debris flows has traditionally been
collected from cross-sections (C
Hen
et alii, 2005;
s
anti
et alii, 2008), photogrammetry (C
oe
et alii,
1997; C
endeRelli
& k
ite
, 1998; v
eyRat
-C
HaRvillon
& m
eRnieR
, 2006), and more recently with airborne
LiDAR (Scheidl et al., 2008) and terrestrial laser
scanning (TLS) data (w
asklewiCz
& H
attanJi
, 2009).
Many of these studies do not consider data regard-
ing pre-event channel morphometry because a low
debris-flow frequency in most locations does not al-
low for the collection of pre- and post- event channel
morphometry data. Therefore, these studies rely upon
assumptions of the channel shape prior to the debris
flow that limit the broader applicability of the results.
These studies have been further limited by their ex-
amination of a single event, which likely is not rep-
resentative of the long-term debris flow record where
multiple debris flow events are interacting to produce
channel changes.
The Chalk Cliffs study basin near Buena Vista,
Colorado presents a unique opportunity to study the
patterns of channel erosion and deposition, and the
Fig. 1 - Chalk Cliffs study area. Contour interval is 40 m
background image
OBSERVATIONS OF DEBRIS FLOWS AT CHALK CLIFFS, COLORADO, USA: PART 2, CHANGES IN SURFACE MORPHOMETRY FROM
TERRESTRIAL LASER SCANNING IN THE SUMMER OF 2009
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
761
drilled into bedrock, or a single 50-cm length of rebar
driven into the ground and cemented in place. Leica
Geosystems High Definition Survey (HDS) targets
were set up on each monument. This network of con-
trol points served two purposes: first, they allowed for
the registration of scans from multiple instrument set-
ups with high accuracy and precision (to remove void
spaces resulting from shadowing); and second, they
allowed for the use of a common Cartesian coordinate
system between each of the four surveys. For each
survey, scans were recorded at 13 different locations
within the study reaches, producing a point cloud with
average point spacing in excess of 6000 points/m
2
.
Leica Geosystems CYCLONE and ESRI ArcGIS
programs were used to post-process the data, and
produce and analyze 2-cm digital elevation models
(DEMs) of pre- and post-event topography. A 25-cm
DEM was also generated. Post-processing was mini-
mal, as each point cloud was already “bare-earth” giv-
en a complete lack of vegetation in the analyzed chan-
nel. Anomalous points, such as dust particles, were
manually removed. The post-processed TLS data
were then brought into ArcGIS to generate DEMs and
analyze topographic change. A triangulated irregular
network (TIN) was created from the raw point cloud
and used to produce the DEMs. Topographic changes
between sequential scans were calculated by subtract-
ing the DEM of the post-event surface from that of
the pre-event DEM surface. Positive values indicated
areas of deposition, while negative values indicated
areas of erosion. Volumetric change, V (m
3
) was cal-
culated for each cell and totaled for the analysis area,
A (m
2
) as follows:
where ΔZ
i
= change in elevation at the pixel (m).
The morphometric parameters, gradient and
roughness, were calculated from the 2-cm DEMs in
ArcGIS. Gradient was calculated at the scale of the
channel reach as the mean of all 2-cm pixels located
along the channel centerline. Roughness was calcu-
lated at both particle and channel form scales. Rough-
ness was calculated as the standard deviation of gra-
dient within a 5x5 pixel neighborhood (f
Rankel
&
d
olan
, 2007). We used the 2-cm resolution DEM to
quantify roughness at the particle scale, and the 25-
cm resolution DEM to quantify roughness associated
in the west basin (red line in Fig. 1). We divided the
149 m surveyed length of the channel into five sepa-
rate reaches based on morphology (Fig. 1). Reach 1
is 27 m long, and occupies the lowermost portion of
the west channel. This reach is relatively steep (ap-
proximately 30 percent), and consists primarily of
exposed bedrock. What channel fill is there consists
of debris-flow deposits and colluvium that accumu-
lates between debris-flow events. Reach 2 occupies
the lowermost portion of the west channel. This is
the shortest (8.5 m) and steepest reach (55 percent
gradient), and is best characterized as a bedrock step.
Colluvium fills in the shallower-gradient portions of
the reach between debris-flow events. Reach 3 occurs
at the confluence of the east and west channels. The
bed material consists primarily of debris-flow depos-
its. This 12-m-long section of channel has the lowest
mean gradient of all the analyzed reaches (roughly 22
percent). Reach 4 occupies the main channel below
a small (roughly 2 m high) boulder step. This 52 m
length of channel is typically filled with debris-flow
deposits and colluvium, and no bedrock was exposed
in the upper two-thirds of the reach during the summer
of 2009. Debris flows periodically scour the channel
to bedrock in the lower third of the reach. The aver-
age gradient for Reach 4 is approximately 30 percent.
Reach 5 is the second longest (49 m) and has the sec-
ond shallowest gradient (roughly 28 percent) of the
analyzed reaches. The bed consists of debris-flow de-
posits and colluvium, with no exposed bedrock. Reach
5 is unique in that nearly the entire west bank consists
of a scree slope at the angle of repose, which provides
a continuous supply of material to the channel.
METHODS
Assessment of the magnitudes and locations of
erosion and deposition associated with each debris-
flow event relied upon repeat TLS surveys and geo-
graphic information systems (GIS) analysis of the
survey data. TLS surveys conducted with a Leica
Geosystems ScanStation 2. Accurate depiction of
fine-scale (<1 cm) changes in surface elevation be-
tween the surveys required a robust control network
and the reduction of data voids associated with sur-
face shadowing from topographic roughness elements
(e.g. steep slopes, overhangs, boulders). To accom-
plish this, twelve control points were established in
the study basin, consisting of either an expansion bolt
(1)
background image
D.M. STALEY, T.A. wASkLEwICZ, J.A. COE, J.w. kEAN, S.w. McCOY & G.E. TUCkER
762
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
sediment yield from the reach. The survey following
the 02 June event identified an area of 46.45 m
2
that
experienced erosion. This value was used for A (equa-
tion 4) in the calculation of volumetric error for eroded
material in reach 1 during the 02 June event. For all
events, volumetric errors are recorded as ± values (in
m) in Table 2. For easier reading, these values are not
included in the text. As there is a very small amount of
systematic error in any of the surveys, errors in volume
likely cancel each other out. As such, the calculated
errors represent a “worst-case scenario”, and are likely
much smaller than those reported.
The quantitative data presented here requires two
caveats. First, we have attempted to repeat the TLS
surveys as soon as possible after each debris-flow
event. In some cases, this represents a period of less
than 36 hours (e.g., the 02 June 2009 debris flow). In
other cases, several days had passed before the repeat
survey could be conducted. During this time, sediment
transported from rockfall and dry-ravel processes, and
not necessarily debris flow, may have increased the
amount of material in the channel. A moderate debris
flow that occurred on 06 September 2009 was not re-
corded and the results of the 29 September 2009 sur-
vey reflect the influence of two debris-flow events.
The second caveat is that the measurements of
topographic change do not, by themselves fully de-
scribe how the deposits were formed. For example,
during depositional events (e.g. 02 June 2009), moni-
toring station data (see m
C
C
oy
et alii, this volume)
measured more debris-flow surges than could be
identified in post-event topography because multiple
surges deposited materials on top of each other. In
addition, within single events, water-rich tails eroded
material initially deposited by the coarse-grained,
fluid-poor surge fronts. Given these two caveats,
our measurements represent an erosion minima and
depositional maxima associated with the debris-flow
events. Even with these limitations, the data present-
ed here represent the documentation of debris-flow
changes in channel morphometry in a natural moun-
tain stream channel at an unprecedented spatial and
temporal resolution, accuracy and spatial extent.
RESULTS
Four TLS surveys were conducted during the 2009
debris-flow season at the Chalk Cliffs study basin.
Here, we present measurements of sediment transport,
with channel form. The mean of each reach was then
calculated for both measures of roughness.
We assessed systematic errors, root mean square
error, and error propagation (Tab. 1) using errors as-
sociated with the registration data from the 187 con-
trol points common between individual scans and
surveys. Systematic error, SE was calculated to detect
any trends in the surveying and registration processes:
where n = number of control points and ΔZ
t
= eleva-
tion error at control point (m).
SE revealed virtually no bias in any of the four
surveys (Tab. 1). Given the independent, randomly
distributed target errors, and lack of systematic errors,
root mean square error (RMSE) values were calcu-
lated to obtain a global error for each survey:
During volume calculations, RMSE for each in-
cluded survey had to also be propogated to account
for errors in each included survey. For example, the
difference surface calculated after the 02 June event
included the errors associated with the pre-event sur-
face and the post-02 June surface. Following the law
of error propogation for independent, random errors
(t
ayloR
, 1982), we use the equation to calculate volu-
metric error, Ev (m3):
where E
v
= volumetric error (m
3
) and A = analysis area
(m
2
). Volumetric error was calculated on a reach-by-
reach basis for net volumetric change. For cells charac-
terized as either erosion or deposition, the total area of
each 0.004 m cell was summed to determine the area
of the individual sediment transport processes (either
erosion or deposition) per channel reach and that area
was used in the calculation of volumetric error. For
example, Reach 1 has a total area of 111.79 m2. This
value was used to calculate volumetric error for the net
(2)
(3)
(4)
background image
OBSERVATIONS OF DEBRIS FLOWS AT CHALK CLIFFS, COLORADO, USA: PART 2, CHANGES IN SURFACE MORPHOMETRY FROM
TERRESTRIAL LASER SCANNING IN THE SUMMER OF 2009
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
763
10-minute intensity of 9.1 mm/hr at 14:05, approxi-
mately 2 hours before the first debris flow surge re-
corded at the upper station (see m
C
C
oy
et alii, this
volume). Field reconnaissance suggests that the rain-
fall was produced by a small, isolated thunderstorm in
the steep upper basin, and the measured rainfall may
significantly underestimate the actual rainfall condi-
tions that initiated the flow.
A repeat TLS survey conducted on 04 June 2009
provided data for the quantification of topographic
change associated with the debris-flow event (Fig.
2a). The event was largely depositional. The surveyed
reaches experienced 12.8 m
3
of total erosion, 63.4 m
3
of total deposition, and a total sediment increase of
50.6 m
3
. Each reach experienced net aggradation. The
greatest amounts (by height and volume) of deposition
occurred in Reach 4 at the lower portion of the reach
below the middle station (Fig. 2a), and the upper por-
tion of Reach 5. Erosion was limited to two small ar-
eas of the channel within Reaches 4 and 5. Maximum
erosion of the pre-event surface occurred in Reach 4,
and was spatially coincident with the steepest portion
of the surveyed channel and a large bank and hillslope
failure. Calculations of net erosion and deposition for
each of the five analyzed reaches are shown in Figure
3 and Table 2.
GIS analysis revealed minor changes in the mor-
phometry of the channel following the debris flow
(Tab. 2). Mean channel gradient remained virtually
constant for Reaches 1, 4 and 5. Mean channel gra-
volumetric change, and subsequent changes in channel
morphometry for the five analyzed reaches organized
by storm date. We also present the net changes that oc-
curred throughout the entire summer season.
02 JUNE 2009
The debris flow event of 02 June 2009 occurred
in response to 9.9 mm of rainfall over 5 hours. A
rain gage at the upper station recorded a maximum
Tab. 1 - Summary of TLS Survey Parameters
Tab. 2 - By channel reach, the calculated volumetric
change and measured channel morphometry
Fig. 3 - Volumetric change calculated for each channel
reach by debris-flow event. A) 02 June 2009; B)
26 July 2009; C) 15 September 2009; D) Total
seasonal change
background image
D.M. STALEY, T.A. wASkLEwICZ, J.A. COE, J.w. kEAN, S.w. McCOY & G.E. TUCkER
764
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
dient decreased in Reach 2 and increased in Reach
3. Particle-size scale surface roughness increased in
Reaches 1, 2 and 3, and decreased in Reaches 4 and
5. Reach 4 experienced the greatest decrease in parti-
cle-scale roughness, reflecting the removal of coarse
clasts that were previously present in the eroded area.
Channel-scale roughness increased in Reaches 1, 2
and 4, and decreased in Reaches 3 and 4. The increase
noted in Reaches 1, 2 and 4 is attributed to the ob-
served deposition of terminal lobes and levees. The
decrease in roughness in Reach 3 is attributed to the
observed development of a smooth, U-shaped bed be-
tween levees. Channel-scale roughness decreases evi-
dent in Reach 4 are attributed to an observed general
smoothing of the channel in both the zones of erosion
and deposition.
26 JULY 2009
The passage of the first surge of debris-flow ma-
terial during the second documented event occurred
around 7:12AM on 26 July in response to 10.7 mm
of rainfall over 78 minutes (see m
C
C
oy
et alii, this
volume). The post-event survey was conducted on
03-04 August (Fig. 2b). This debris flow was almost
exclusively depositional (Figure 3 and Table 2), with
the channel experiencing a total of 51.0 m
3
of deposi-
tion and only 3.7 m
3
of erosion. The greatest amount
of deposition was located in the uppermost portion of
Reach 4, just below the boulder step (Fig. 2b). Erosion
was confined to two small areas in Reach 1.
Channel-bed gradient remained similar, with
Reaches 1, 2, 3 and 5 experiencing almost no changes
in gradient (Tab. 2). Reach 4 experienced a slight in-
crease in gradient, reflecting the influence of the steep
front of the terminal lobe that was deposited in the
uppermost section of the reach.
Particle-scale roughness decreased in all reaches
except Reach 1 (Tab. 2). The increase in Reach 1 is at-
tributed to the introduction of coarse clasts on the front
of the freshly-deposited terminal lobe. Deposition in
all other reaches created relatively smooth surfaces of
debris-flow terminal lobes. All reaches experienced a
decrease in channel-scale roughness, reflecting a gen-
eral smoothing of the surface as material filled depres-
sions produced during the previous debris flow.
06 SEPTEMBER 2009
No specific topographic information was col-
lected regarding the 06 September 2009 debris flow
event. This event occurred at 14:05 in response to 6.3
Fig. 2 - Changes in elevation
measured during TLS
surveys associated with
A) 02 June debris-flow
event; B) 26 July 2009
debris-flow event; C) 15
Sepetember 2009 debris-
flow event; C) entire 2009
debris-flow season. Con-
tour interval is 1 m. Flow
direction is from left to
right
background image
OBSERVATIONS OF DEBRIS FLOWS AT CHALK CLIFFS, COLORADO, USA: PART 2, CHANGES IN SURFACE MORPHOMETRY FROM
TERRESTRIAL LASER SCANNING IN THE SUMMER OF 2009
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
765
increased in all reaches. This reflects the removal of
the relatively smooth channel fill deposited during
the previous event, and the observed development of
knickpoints in Reaches 4 and 5, presumably during
the water-rich recessional flow of the debris flow.
SEASONAL CHANGES
Seasonal changes were quantified by comparing
topographic data from the 29 May and 29 September
TLS surveys (Fig. 2d). Overall, the analyzed channel
lost 138.4 m3 of material over the season. Of the five
reaches, only Reach 3 experienced net deposition (Fig.
3 and Tab. 2). The greatest positive changes in surface
elevation were identified in Reach 3, where the chan-
nel adjusted to account for the development of the
boulder step during the previous summer. Negative
topographic change was at its greatest at the steepest
section of Reach 4 and the lower half of Reach 5. In
both cases, these areas were immediately adjacent to
large failures of bank and hillslope material.
Mean channel bed gradients were similar between
pre- and post-season surveys (Tab. 2). Reaches 1, 4 and 5
maintained a fairly constant mean gradient, while Reach
3 experienced an increase in gradient during each debris
flow. This increase can be attributed to the channel ad-
justing to the development of the 2-m-high boulder step
during a debris flow in early October of 2008.
Particle-scale roughness increased in Reach 1, de-
creased in Reaches 4 and 5, and remained relatively
unchanged in Reaches 2 and 3. The increase in parti-
cle-scale roughness in Reach 1 reflects the removal of
fine-grained colluvial fill that was present at the onset
of the debris flow season. Decreases in particle-scale
roughness in Reaches 4 and 5 reflect the removal of
coarse material in the lower portion of Reach 4 and
upper portion of Reach 5.
DISCUSSION
Analysis of high-resolution topographic data
derived from multiple TLS surveys quantified how
sediment is transported during both individual debris
flow events and in response to a season of debris-flow
activity that included four debris-flow events. The
survey data clearly demonstrate that the locations and
magnitudes of channel erosion and deposition vary
from event to event. These findings are in agreement
with C
Hen
et alii (2005) who identified channel lo-
cations that experienced both erosion and deposition
mm of rainfall over approximately 45 minutes.
Flow stage recorded at channel monitoring sta-
tions suggested a moderately-sized debris flow, with
peak stages of 0.53 meters at the upper station, and
0.80 meters at the middle station. Video camera foot-
age was not available for the upper reaches, and video
of the lower reach suggested a modest amount of dep-
osition. No field reconnaissance was conducted before
the next debris flow occurred on 15 September 2009.
15 SEPTEMBER 2009
The 15 September 2009 debris flows were the
most significant event of the summer in terms of pre-
cipitation, flow stage and bed stress (m
C
C
oy
et alii,
this volume). This debris flow event initiated after
24.6 mm of rainfall over a period of 115 minutes. A
TLS survey of the post-event topography was con-
ducted on 29 September 2009 (Fig. 2c). The great-
est amount of topographic change for the summer
was recorded during this inter-survey period, which
included both the 06 and 15 September debris flows.
These flows resulted in significant erosion of the chan-
nel bed, with 248.5 m3 of sediment removed from the
channel, and only 12.3 m3 of deposition. All channel
reaches experienced net erosion with Reach 5 experi-
ence the greatest volumetric loss (Fig. 3 and Tab. 2).
The largest amount of erosion relative to the prior
survey occurred in Reach 4, in the same general loca-
tion of the area of greatest erosion during the 02 June
event. In both cases, this was spatially coincident with
the steepest channel gradient and adjacent to a large
failure of bank and hillslope material. Nearly the en-
tire length of Reach 1 was scoured to bedrock, though
some channel loading had occurred prior to the TLS
survey from rockfall and dry-ravel processes.
This event produced minor changes in mean chan-
nel gradient (Tab. 2). Reaches 2 and 3 both experi-
enced an increase in channel gradient, while gradient
remained virtually unchanged for Reaches 1, 4, and 5.
Gradient increases in Reach 2 reflect the removal of
fill on the bedrock step. Reach 3 increases in gradient
reflect the observed channel adjustment as material at
the top of the boulder step was removed.
Increases in roughness at the particle-scale were
identified for all reaches with the exception of Reach
3 (Tab. 2). Here, the relatively smooth pre-event sur-
faces of debris-flow deposits were eroded, resulting in
a surface of coarser clasts. Channel-scale roughness
background image
D.M. STALEY, T.A. wASkLEwICZ, J.A. COE, J.w. kEAN, S.w. McCOY & G.E. TUCkER
766
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
The spatial patterns of erosion, deposition, and net
volumetric change that we measured/quantified could
not have been consistently predicted from pre-debris
flow measurements of channel gradients or roughness.
Linear regression identified no significant relation (r2
< 0.017 for all three parameters). Although changes in
roughness and particle size impose new boundary con-
ditions on subsequent debris flows (Benda and Dunne,
1997), it is apparent that other factors, such as water
content, flow mechanics and rheology, also exert a
significant influence on the ultimate response of the
channel to the debris flow. As such, further research re-
garding the spatial and temporal patterns of debris-flow
erosion and deposition must incorporate both morpho-
metric data and information regarding flow properties.
CONCLUSIONS
We have documented the morphometric changes
of a natural channel in response to multiple debris
flow events during the summer of 2009. High resolu-
tion terrestrial laser scanning data permits the evalu-
ation of these changes with very high accuracy and
precision. Each monitored channel reach experienced
erosion, deposition, and both net volumetric gains
and losses during the summer debris flows. Overall,
the channel experienced a net erosion of 276.8 m
3
of
material in response to the four events of the season.
Local variations in channel morphometry were related
to the selective movement of particular grain sizes, the
occurrence of channel bank failures, and the develop-
ment of knick points. Mean channel-reach gradient
stayed constant throughout the season, indicating a
geomorphic equilibrium with the high sediment-trans-
port rates and frequent debris flows that occur in the
study basin. Examination of relationships between the
magnitude of erosion and deposition yielded no strong
correlations with channel reach gradients of roughness
measures, indicating that point-specific predictions of
locations or magnitudes of erosion or deposition can-
not be adequately derived from statistical analyses of
pre-event channel gradient or topographic roughness
alone. Instead, a combination of high-resolution topo-
graphic changes and process information from in-situ
measurements of flow dynamics may be necessary to
better understand the relationships between debris-
flow process mechanics, changes in surface form,
and debris-flow magnitude. At Chalk Cliffs, we are
building an extensive dataset of both process-oriented
during multiple debris flows in China. In our study,
Reaches 4 and 5 exhibited markedly different re-
sponses during the three documented events. The up-
per portion of Reach 4 (just below the boulder step)
experienced deposition during the first two events,
and erosion in response to the larger 14 September
debris flow. Material was eroded from the steep mid-
dle portion of Reach 4 during the 02 June and 15
September debris flows, and deposited during the 26
July event. Large bank failures were observed in this
reach following the June and September events, but
not following the July event. Reach 5 experienced net
deposition during the first two debris-flow events and
a substantial amount of erosion during the 15 Septem-
ber event. These findings highlight the importance of
smaller events as sources of sediment for subsequent
larger debris flows. For the entire season, the greatest
net erosion occurred in areas where material was de-
posited during the smaller events, such as in Reach 4
below the boulder step and below the middle channel
monitoring station, and in Reach 5.
Analyses of morphometric data reveal unique
insights into the evolution of channel gradient and
roughness in response to multiple debris flow events.
While there were some event-to-event fluctuations
in gradient within the study reaches, analysis of sea-
sonal changes show that mean channel-reach gradient
remained fairly constant over this longer time period.
The exception, however, occurred in Reach 3, which
exhibited a seasonal decrease in channel gradient. This
exception is attributed to channel response to the ob-
served development of a boulder step during the fall
of 2008. Because the gradients of the other reaches
(both colluvial and bedrock) remained fairly constant
over the season despite significant amounts of erosion,
we conclude that channel gradients are adjusted to the
high sediment-transport rates and debris flows in the
study basin, at least during the period of record.
Measurements of roughness showed more varia-
bility than did those of channel gradient, with changes
occurring at both particle- and channel-scales in re-
sponse to different flow events. In general, however,
roughness at both scales increased between May and
September, although different responses were exhib-
ited to individual events. The relative magnitudes of
the debris flows as well as the amount of erosion or
deposition influence both the particle- and channel-
scale roughness measures.
background image
OBSERVATIONS OF DEBRIS FLOWS AT CHALK CLIFFS, COLORADO, USA: PART 2, CHANGES IN SURFACE MORPHOMETRY FROM
TERRESTRIAL LASER SCANNING IN THE SUMMER OF 2009
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
767
the National Science Foundation under Grant No.
02-39749, Grant No. 0934131 and Grant No. EAR-
0643240. Any opinions, finding, or conclusions or
recommendations expressed in this material do not
necessarily reflect the views of the National Science
Foundation. The authors gratefully acknowledge Su-
san Cannon, Ryan Gold and the anonymous review-
ers for insightful reviews that have helped improved
this manuscript. We also would like to acknowledge
Cal Scheinert and Joseph Gartner for field assistance.
measurements and changes in topography associ-
ated with numerous debris-flow events (McCoy et al.,
2010, McCoy et al., this volume). Further analysis of
this unique combination of both datasets presents an
opportunity to ultimately quantify the link between
debris-flow processes and observed changes in the
morphometry of natural channels.
ACKNOWLEDGMENTS
This paper is based upon work supported by
REFERENCES
b
eRti
m., G
enevoi
R., s
imoni
a. & t
eCCa
P.R. (1999) - Field observations of a debris flow event in the Dolomites. Geomorphol-
ogy, 29: 265-274.
b
eRti
m. & s
imoni
a. (2005) - Experimental evidences and numerical modeling of debris flow initiated by channel runoff.
Landslides, 2: 171-182.
b
ovis
m.J. & d
aGG
b.R. (1992) - Debris flow triggering by impulsive loading: mechanical modeling and case studies. Canadian
Geotechnical Journal, 29: 345-352.
b
ovis
m.J. & J
akob
m. (1999) - The role of debris supply conditions in predicting debris flow activity. Earth Surface Processes
and Landforms, 24: 1039-1054.
C
annon
s.H., b
iGio
e.R. & m
ine
e. (2001a) - A process for fire-related debris flow initiation, Cerro Grande fire, New Mexico.
Hydrological Processes, 15: 3011-3023.
C
annon
s.H., k
iRkHam
R.m. & P
aRise
m. (2001b) - wildfire-related debris flow initiation processes, Storm king Mountain,
Colorado. Geomorphology, 39: 171-188.
C
annon
s.H., G
aRtneR
J.e., P
aRRett
C. & P
aRise
m. (2003) - wildfire related debris-flow generation through episodic progres-
sive sediment-bulking processes, western USA. In: R
iCkemann
d & C
Hen
C.l.
eds
. Debris-Flow Hazards Mitigation – Me-
chanics, Prediction and Assessment, Proceedings of the 3
rd
International Conference on Debris-Flow Hazards Mitigation,
Davos, Switzerland, 10-12 September 2003: A.A. Balkema, Rotterdam: 71-82.
C
endeRelli
d.a. & k
ite
J.s. (1998) - Geomorphic effects of large debris flows on channel morphology at North Fork Mountain,
eastern west Virginia, USA. Earth Surface Processes and Landforms, 23: 1229.
C
Hen
J., H
e
y.P. & w
ei
f.Q. (2005) - Debris flow erosion and deposition in Jiangjia Gully, Yunnan, China. Environmental Geol-
ogy, 48: 77-1-777.
C
oe
J.a., k
inneR
d.a. & G
odt
, J.w. (2008) - Initiation conditions for debris flows generated by runoff at Chalk Cliffs, central
Colorado. Geomorphology, 96: 270-297
e
GasHiRa
s., H
onda
n. & i
toH
t. (2001) - Experimental study on the entrainment of bed material into debris flow. Phyics and
Chemistry of the Earth (C), 26 (9): 645-650.
f
Rankel
k.l. & d
olan
J.f. (2007) - Characterizing arid region alluvial fan surface roughness with airborne laser swath map-
ping digital topographic data. Journal of Geophysical Research, 112: F02025.
G
abet
e.J. (2003) - Post-fire thin debris flows: sediment transport and numerical modeling. Earth Surface Processes and Land-
forms, 28: 1341-1348.
G
abet
e.J. & b
ookteR
a. (2007) - A morphometric analysis of gullies scoured by post-fire progressively-bulked debris flows in
southwest Montana, USA. Geomorphology, 96: 298-309.
G
odt
J.w & C
oe
J.a. (2007) Alpine debris flows triggered by a 28 July 1999 thunderstorm in the central Front Range, Colorado.
Geomorphology, 84: 80-97.
H
unGR
o., m
C
d
ouGall
s. & b
ovis
m. (2005) - Entrainment of material by debris flows. In: Jakob M. & Hungr O. eds. Debris-
flow hazards and related phenomena: Springer Berlin Heidelberg: 135-158.
J
akob
m., b
ovis
m. & o
den
m. (2005) - The significance of channel recharge rates for estimating debris-flow magnitude and
frequency. Earth Surface Processes and Landforms, 30: 755-766.
background image
D.M. STALEY, T.A. wASkLEwICZ, J.A. COE, J.w. kEAN, S.w. McCOY & G.E. TUCkER
768
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
m
aJoR
J. (2000) - Gravity driven consolidation of granular slurries – implications for debris-flow deposition and deposit char-
acteristics. Journal of Sedimentary Research, 70 (1): 64-83.
m
illeR
m.G. (1999) - Active breaching of a geometric segment boundary in the Sawatch Range normal fault, Colorado, USA.
Journal of Structural Geology, 21 (7): 769-776.
m
C
C
aRdell
b.w., b
aRtelt
P. & k
owalski
J. (2007) - Field observations of basal forces and fluid pore pressure in a debris flow.
Geophysical Research Letters, 34: L07406.
m
C
C
oy
s.w., k
ean
J.w., C
oe
J.a., s
taley
d.m., w
asklewiCz
t.a. & t
uCkeR
G.e. (2010) - Evolution of a natural debris flow:
In situe measurements of flow dynamics, video imagery and terrestrial laser scanning. Geology, 38 (8): 735-738.
m
C
C
oy
s.w., C
oe
J.a., k
ean
J.w., s
taley
d.m., w
asklewiCz
t.a. & t
uCkeR
G.e. (this volume) - Observations of debris flows
at chalk cliffs, Colorado, USA: part 1, in-situ measurements of flow dynamics, tracer particle movement and video imagery
from the summer of 2009.
In: Debris-flow hazards mitigation: Mechanics, prediction and assessmentflows hazard mitiga-
tion,
this volume.
m
oody
J.a. & k
inneR
d.a. (2006) Spatial structures of stream and hillslope drainage networks following gully erosion after
wildfire. Earth Surface Processes and Landforms, 31: 319-337.
s
anti
P.m.,
de
w
olfe
v.G., H
iGGins
J.d., C
annon
s.H. & G
aRtneR
J.e. (2008) - Sources of debris flow material in burned areas.
Geomorphology, 96: 310-321.
s
Hakesby
R.a. & d
oeRR
s.H. (2006) - wildfire as a hydrological and geomorphological agent. Earth-Science Reviews, 74:
269-307.
t
ayloR
J.R. (1982) - An introduction to error analysis. University Science Books: Mill Valley, CA, USA. 270pp.
v
eyRat
-C
HaRvillon
s. & m
eRnieR
m. (2006) - Stereophotogrammetry of archive data and topographic approaches to debris-
flow torrent measurements: calculation of channel-sediment states and a partial sediment budget for Manival torrent (Isere,
France)
. Earth Surface Processes and Landforms, 31: 201-219.
w
asklewiCz
t.a. & H
attanJi
t. (2009) - High-resolution analysis of debris-flow induced channel changes in a headwater
stream, Ashio Mountains, Japan. The Professional Geographer, 61: 231-249.
Statistics