Geomorphic evidence for post-10 Ma uplift of the western flank of the Central
Andes 18°30-22°S
1
2
3
4
5
6
7
8
9
10
11
12
13
Department of Earth and Atmospheric Sciences, Cornell University, Snee Hall, Ithaca NY
14853, U.S.A.
2
Servicio Nacional de Geologia y Minera, Av. Santa Maria 0104, Santiago, Chile
3
Department of Earth Atmospheric and Planetary Sciences, Massachusetts Institute of
Technology, 54-1117, Cambridge, MA 02139 U.S.A.
*
now at: Department of Earth and Environmental Sciences, University of Rochester, 227
Hutchinson Hall, Rochester, NY 14627 U.S.A.
14
15
Abstract
The western Andean mountain front forms the western edge of the Central Andean
16
Plateau. Between 18.5° and 22°S latitude, the mountain front has ~3000 m of relief over ~ 50 km
17
horizontal distance that has developed in the absence of major local Neogene deformation.
18
Models of the evolution of the plateau, as well as paleoaltimetry estimates, all call for continued
19
large magnitude uplift of the plateau surface into the late Miocene (i.e. younger than 10 Ma).
20
Longitudinal river profiles from 20 catchments that drain the western Andean mountain front
21
contain several streams with knickpoint-bounded segments that we use to reconstruct the history
22
of post 10 Ma surface uplift of the western flank of the Central Andean Plateau. The generation
23
of knickpoints is attributed to tectonic processes and are not a consequence of base level change
24
related to Pacific Ocean capture, eustatic change, or climate change as causes for creating the
25
knickpoint-bounded stream segments observed. Minor valley filling alluvial gravels intercalated
26
with the 5.4 Ma Carcote ignimbrite suggest uplift related river incision was well under way by
27
5.4 Ma. The maximum age of river incision is provided by the regionally extensive,
28
approximately 10 Ma El Diablo-Altos de Pica paleosurface. The river profiles reveal that relative
29
surface uplift of at least1 km occurred after 10 Ma.
30
31
Introduction
The processes and rates related to the formation of large continental plateaus remain
32
controversial. Many different mechanisms and ideas have been invoked to explain their
33
evolution, spanning climatic controls [e.g. Montgomery, et al., 2001], pure structural thickening
34
via horizontal shortening [e.g. McQuarrie, 2002], to viscous flow of lower crustal material
35
[Clark and Royden, 2000]. The outcomes of some models of plateau forming processes are much
Gregory D. Hoke1*, Bryan L. Isacks1, Teresa E. Jordan1, Nicolás Blanco2, Andrew J. Tomlinson2,
and Jahandar Ramezani3
1
1
2
36
easier to test with observation than others. The simplest metric of tectonic activity is upper
37
crustal deformation combined with chronologic constraints. In contrast, proposed mechanisms of
38
plateau formation not resulting in surface-breaking faults, are much more difficult to constrain.
39
The Andes of South America are home to the Central Andean Plateau, the second highest
40
continental plateau on Earth at ~4 km average elevation. The plateau is bounded on either side by
41
mountainous topography of the Eastern and Western Cordilleras (Figure 1). The eastern side of
42
the Andean orogen is characterized by a wide belt of Neogene supracrustal deformation
43
expressed by numerous surface-breaking faults and folds. The belt has varied tectonic style,
44
typified by basement uplifts in the Argentine and Peruvian forelands, as well as the spectacular
45
thin-skinned Sub-Andean fold-and-thrust belt in Bolivia. Structural studies indicate that as much
46
as 300 km of Neogene horizontal shortening [McQuarrie, et al., 2005] has occurred across the
47
entire plateau and postulate that it generated 3000-4000 m of relief between the plateau and the
48
undeformed foreland [Elger, et al., 2005]. By contrast, the western mountain front of the Central
49
Andes contains a similar amount of relief but only minor Neogene supracrustal deformation
50
(Figure 2). Several studies related to the development of the plateau or the Central Andes as a
51
whole [Garzione, et al., 2006; Gregory-Wodzicki, 2000; Isacks, 1988; Kay and Kay, 1993; Kay,
52
et al., 1994; Lamb and Hoke, 1997] postulate that uplift of the plateau surface occurred
53
throughout the Miocene. However, the western mountain front has provided little evidence in
54
support of the uplift history interpretations that were based on observations from other parts of
55
the plateau, due to the lack of well-defined geologic structures of appropriate age. Where
56
structural geology approaches cannot be applied, we must explore other methods to test the
57
various hypotheses for plateau formation. Here we examine the longitudinal stream profiles from
58
20 drainage networks (Figure 2) descending the western flank of the Central Andean Plateau in
59
an effort to detect and quantify post-10 Ma relief formation in the region. To do so we perform
60
three tasks. The first is the extraction of stream profiles from digital elevation data, the second is
61
demonstration of the ages of stream segments, and third is reconstruction of the surface relief at
62
the times when the dated stream segments were created.
63
64
65
66
Geologic setting
The Andes are the type example of a non-collisional orogen in a convergent margin
setting. Here the Nazca plate subducts below the overriding South American plate. Subduction
3
67
along this margin has been active since the Mesozoic, with the modern phase of mountain
68
building beginning in the latest Oligocene [Jordan, et al., 1997; Mpodozis and Ramos, 1989].
69
The Central Andean segment of the orogen is dominated by the Central Andean Plateau which
70
spans some ~1,500 km N-S and 300 km E-W at its widest point. It is bounded to the west by the
71
volcanic peaks of the Western Cordillera (the volcanic arc) and to the east by the Eastern
72
Cordillera (the hinterland of the modern Sub-Andean fold-thrust belt).
73
Our study area lies within the modern Andean forearc and forms the western flank of the
74
Central Andean Plateau. Several physiographic provinces exist in the study area (Figure 2). From
75
west to east, they are the: 1) Coastal Cordillera, 2) Central Depression (Pampa del Tamarugal),
76
3) western mountain front, 4) Western Cordillera and 5) Altiplano (Figure 2). In the study area
77
the Coastal Cordillera lies between the Pacific Ocean and the Central Depression and has peaks
78
as high as 2500 meters and an average elevation of ~1500 m that wanes towards the north, where
79
it is less than 1000 m at 18.5°S. The Coastal Cordillera is composed of Mesozoic volcanic arc
80
and plutonic rocks, and marine sedimentary rocks, with Oligocene to Recent alluvial and
81
colluvial sedimentary units filling intermontane basins [García, et al., 2004; Muzzio J., 1986,
82
1987; SERNAGEOMIN, 2002; Silva, 1977]. The Central Depression is a broad plain at ~ 1000 m
83
elevation that lies between the Coastal Cordillera and the western mountain front and is
84
underlain by Oligocene to Recent sedimentary fill. The western mountain front connects the high
85
elevations of the Western Cordillera and Altiplano to the Central Depression. The Western
86
Cordillera is comprised of rocks of the Miocene to Recent volcanic arc.
87
All parts of the study area below 3500 m elevation are subject to the hyper-arid
88
conditions of the Atacama Desert. The analysis of meteorological station data by Houston and
89
Hartley [2003] shows that rainfall is strongly dependent on elevation in this area. At 1000 m, in
90
the Central Depression, rainfall is <50 mm/yr, though at elevations > 3500 m, rainfall is ~ 150
91
mm/yr. The high peaks of the Western Cordillera (5000-6000 m) do not have station data, but are
92
believed to have annual precipitation in excess of 200 mm/yr. Several lines of evidence point to
93
the long-term stability of climate in this region [Alpers and Brimhall, 1988; Houston and
94
Hartley, 2003; Sillitoe and McKee, 1996].
95
The western mountain front in northernmost Chile is a steep (~3°) topographic front with
96
~3000 m of vertical relief occurring over ~50 km horizontal distance. The latest Oligocene-
97
middle Miocene sedimentary package exposed along the front is broadly deformed into a west-
4
98
dipping monoclinal form [Isacks, 1988; Mortimer, et al., 1974; Mortimer and Saric Rendic,
99
1975] (Figure 3), with several distinct sedimentary and volcanic units creating marker horizons
100
that are continuous from the crest of the mountain front to its toe. Locally, parts of this sequence
101
are folded and cut by steeply dipping west-vergent reverse faults that are estimated to have
102
produced 3-8 km of horizontal shortening [Digert, et al., 2003; Farias, et al., 2005; García,
103
2002; García, et al., 2004; Muñoz and Charrier, 1996; Pinto, et al., 2004; Victor, et al., 2004].
104
Currently, two competing models exist for the topographic relief development of the western
105
Andean mountain front. The Isacks [1988] model interprets the upper crust as a monoclinal
106
flexure that is an accommodation structure between the rising plateau and the static forearc
107
[Isacks, 1988]. An alternative contends that offset on upper crustal faults generates the full
108
topographic relief [Muñoz and Charrier, 1996; Victor, et al., 2004]. Although the two opposing
109
models lead to the same basic geometry and end result of the western monocline as a landform,
110
there are differences in both the mechanisms and timing of uplift. In Isacks’s [1988] model,
111
excluding the thermal effects, uplift is a result of isostatic adjustment to thickening in the plateau
112
caused by east-west shortening across the orogen. In this model, deformation along the western
113
mountain front, in particular the monoclinal flexure, does not play an important role in the
114
structural shortening but instead accommodates the difference in isostatic compensation between
115
the undeformed forearc and the deforming plateau, and as such, the flexure is a consequence
116
rather than a cause of uplift. Because the model links uplift to shortening history across the
117
orogen, it predicts that formation of the western mountain front occurred throughout the time
118
span of shortening, hence throughtout the Miocene to Recent [Allmendinger, et al., 1997; Isacks,
119
1988]. In the fault based model, the faulting and folding are the cause and not the consequence,
120
of uplift. Therefore the model predicts that relief generation is restricted to the timing of the local
121
faults and folds [Farias, et al., 2005; Muñoz and Charrier, 1996; Victor, et al., 2004].
122
The profound and long-term aridity of this region results in the extraordinary preservation
123
of strata on the steep western mountain front, as well as an extremely well preserved poly-phase
124
landscape [García and Hérail, 2005; Hoke, et al., 2004; Mortimer and Saric Rendic, 1975;
125
Tosdal, et al., 1984; Worner, et al., 2002] (Figure 3 and 4). The Neogene cover of the western
126
mountain front is comprised mostly of alluvial sedimentary rocks with intercalated tuffs ranging
127
from the latest Oligocene to Pliocene age. These units are: north of 19.75°S, the Oligo-Miocene
128
Oxaya Formation [García, 2002; Salas, et al., 1966] and overlying middle Miocene El Diablo
5
129
Formation; south of 19.75°S, the Oligo-Miocene Altos de Pica Formation [Galli Olivier and
130
Dingman, 1962] (Figure 4) and overlying uppermost Miocene- lower Pliocene strata. These units
131
and several of their exposed members create regionally extensive, recognizable surfaces that can
132
be used as timelines. Particularly useful in this study are the top surfaces of the El Diablo and
133
Altos de Pica Formations.
134
Based on different geochronologic constraints, we treat the top surfaces of the El Diablo
135
and Altos de Pica Formation as one continuous, temporally equivalent pediment surface.
136
Traditionally the pediment at the top of El Diablo Formation [Tobar, et al., 1968] has been
137
interpreted to be ~ 10 Ma, an age constrained by the ages of andesite clasts (youngest clast age of
138
11.9 Ma) within the top few meters of the unit [García, 2002; García and Hérail, 2005] and by
139
the 8-9 Ma K-Ar age of the Tana lava flow which overlies the El Diablo surface [Mortimer, et
140
al., 1974; Muñoz and Sepulveda, 1992; Pinto, et al., 2004] (Figure 4). Likewise, new
141
chronologic data (see supplementary data 1) for ignimbrite horizons broadly bracket the age of
142
the pediment surface at the top of the Altos de Pica Formation. The pediment is older than the
143
Carcote ignimbrite inset in river valleys incised into the pediment surface (hence older than 5.4
144
Ma) and younger than the youngest ignimbrite interbedded within the Altos de Pica Formation,
145
hence, younger than 13.03 Ma (Figure 4). Further to the south, the base of the Arcas fan overlies
146
the top surface of Altos de Pica and has an age of ~ 7 Ma [Kiefer, et al., 1997] Recent studies
147
reported in von Rotz et al. [2005] tentatively reassign the maximum age limit on portions of the
148
El Diablo surface to between 8.5 and 7.5 Ma on the basis of magnetostratigraphy from distal and
149
proximal sections located between 18.5°S and 19.75°S and a Rb-Sr age on an intercalated tuff
150
near the top of the formation. While there is no formal stratigraphic correlation between the El
151
Diablo and Altos de Pica formations, there time equivalence, has always resulted in their de-
152
facto correlation [Farias, et al., 2005; Mortimer, 1980; Pinto, et al., 2004]. The summary of
153
prior data, combined with the new data presented here reflect the current state of knowledge
154
regarding the age of the surface. 21NeCosomogenic nuclide studies attempting to date the
155
pediment surfaces have all yield ages that are outside of what is stratigraphically reasonable
156
[Dunai, et al., 2005; Evenstar, et al., 2005]. No data directly constrain the age of the El Diablo
157
Altos de Pica pediment surface.
158
159
Rivers of northernmost Chile
6
160
We consider rivers that drain the western mountain front of the central Andes between
161
18.5° and 22°S latitude (Figure 5) as potential recorders of surface uplift on the western slope of
162
the Central Andean plateau. Mortimer [1980] was the first to study and describe the drainages of
163
northern Chile. The age of the mountain front drainage systems are constrained by the fact that
164
all of the basins are incised into the 10 Ma Altos de Pica-El Diablo pediment surface and river
165
valleys between 20-22°S are partly filled by the 5.4 Ma Carcote ignimbrite (Figure 4; see
166
supplementary data 1) and therefore can be assigned a Late Miocene age.
167
Two general classes of river valleys exist in this region, those with outlets to the Pacific
168
Ocean (rivers between 18.5°-19.75°S and the Río Loa with its outlet at 21.5°S) and those that
169
drain into the effectively closed Pampa del Tamarugal basin (19.75°-22°S). All of the modern
170
drainage basins used in this study have their headwaters at elevations > 4000 m. Rivers that drain
171
to the Pacific Ocean have a maximum relief of ~ 6000 m, while those streams that drain into the
172
Pampa del Tamarugal Basin have ~ 4500 m of relief. Figure 5 shows the representative relief
173
structures for the two classes of drainage system.
174
North and south of the closed Pampa del Tamarugal Basin, rivers that drain the mountain
175
front traverse the Coastal Cordillera and drain to the Pacific (Figure 2). The least advanced
176
migration of the breakthrough knickpoints are in the Río Loa and the Quebrada Tiliviche. The
177
timing of this event for the Río Loa can be constrained by the cross-cutting relationships between
178
the valley and the local stratigraphy near the town of Quillagua [Sáez, et al., 1999]. Near the top
179
of the deposits cut by the Río Loa is a 6 Ma volcanic ash [Sáez, et al., 1999], which can be used
180
to infer that the initiation of incision occurred sometime after 6 Ma. Despite the fact that the Loa
181
has captured all of the basins that drain into the Salar de Llamara, the extremely slow
182
propagation of the knickpoint back into the Salar has thus far effectively isolated the streams
183
which drain the western Andean slope from their new base level.
184
The age of down cutting of the Quebrada Tiliviche across the Coastal Cordillera (Figure
185
2) is constrained by new geochronologic data presented here. A strath terrace cut into Mesozoic
186
bedrock occurs at an elevation of 950 m.s.l., a position that is 50 m below the top of the canyon
187
and ~550 m above the valley bottom (Figure 6). The position of the terrace high on the canyon
188
wall indicates it accumulated at an early stage in the incision of the canyon. U-Pb dating of
189
zircons in a volcanic ash intercalated in the river deposits overlying the strath indicates a
190
depositional age of 6.4 Ma. The post-terrace incision rate calculated using these values is ca. 90
7
191
m/my. This incision has two possible origins, one related to a breakthrough or spillover to the
192
Pacific Ocean and the other being related to surface uplift of the Coastal Cordillera. The
193
breakthrough or spillover hypothesis requires that the Central Depression did not drain to the
194
Pacific prior to the ca. 6.4 Ma incision, a situation that is not consistent with the dominance of
195
fluvial deposits and lack of evaporite deposits near the latitude of Quebrada Tiliviche. In either
196
case, despite the ca. 6.4 Ma age for initiation of down cutting, the knickpoint has only migrated
197
10 km into the Central Depression where river incision is minimal (see profiles in Mortimer,
198
1980; and in Hoke, 2006). In contrast, the knickpoint has already migrated through the canyons
199
farther north to the foot of the western slope. This may suggest that either the canyons to the
200
north have an older age than the 7.5 Ma age estimated by Kober et al. [2006], or that the smaller
201
Tiliviche valley has a much lower stream power for down cutting. Wörner et al. [2002] placed a
202
minimum age constraint on the development of the Lluta valley with a 2.7 Ma 40Ar/39Ar age on
203
the Lauca-Perez ignimbrite at the bottom of the valley near its outlet to the Pacific Ocean.
204
205
206
Stream profile reconstruction methodology
We use knickpoint-bounded segments of river profiles to reconstruct the downstream
207
paleo-profiles of rivers that drain the western flank of the central Andes. This approach relies on
208
the implicit assumption that the river profile segment being modeled existed under equilibrium
209
conditions and that knickpoints migrate upstream in erosional waves [Whipple and Tucker,
210
1999]. River profiles can serve as sensitive indicators of climate change [Zaprowski, et al.,
211
2005], base-level fluctuation and/or tectonic forcing [Kirby and Whipple, 2001; Snyder, et al.,
212
2000; Whipple and Tucker, 1999; Wobus, et al., 2003]. In this paper we follow the reasoning
213
discussed in Snyder et al. [2000] and Schoenbohm et al. [2004]. The stream power erosion model
214
for a detachment-limited bedrock stream predicts channel erosion as a function of drainage area
215
(A) and slope (S),
216
217
= KA m S n
(1)
218
where is the rate of bedrock channel erosion, K is a coefficient of erosion and the exponents m
219
and n are positive constants [Whipple and Tucker, 1999]. According to this model, slope must
220
decrease with increasing drainage area in order to keep the rate of channel bed erosion constant.
8
221
In tectonically active regions, river-bed elevation through time (dz/dt) is a balance between uplift
222
rate (U) and erosion rate ,
dz /dt = U = U KA m S n
223
(2).
224
225
226
In the steady-state (equilibrium) case where dz/dt = 0, (2) simplifies and can be
rearranged to yield
227
229
228
S = (U /K)1/ n Am / n
(3)
230
231
Substitution of Hack’s law, A = xh [Hack, 1957], where x is equal to distance
232
downstream in meters and h is a dimensionless constant with a value of ~ 1.6, for the area term
233
(A) in equation (3) and integration yields the simplified version of the long profile of a steady
234
state river given in Whipple and Tucker [1999]:
235
237
239
z(x) = z0 (U /K)1/ n (1
236
hm 1 1hm / n
) (x 0
x1hm / n )
n
(4)
240
where x0 and z0 are the position and elevation at the top of the profile, respectively. The power
241
law relationship in (3) is observed directly in stream profiles extracted from digital elevation
242
models, and the parameters m/n and (U/K) 1/n can be derived by a regression through slope-area
243
data. The ratio m/n (slope of the regression line) is commonly referred to as the concavity index
244
(), while (U/K) 1/n (slope intercept) is the steepness index (ks). The values of ks and are highly
245
correlated; therefore we report values for ks that are normalized (ksn) to a reference concavity =
246
0.4, a value in the range of those commonly reported [Snyder, et al., 2000]. Concavity values for
247
the rivers considered in this study range between 0.15 and 20, which is much greater than the
248
theoretical range of 0.3 to 0.6 [Hack, 1957; Snyder, et al., 2000 and references therein]. As
249
discussed in Whipple [2004], such very high concavities are (a) usually local to short reaches of
250
rivers and (b) often indicative of disequilibrium river profiles. An alternative explanation for this
251
concavity range could be the strong contrast between different rock types.
252
River long profiles are thought to represent the relief structure of the landscape [Whipple,
253
et al., 1999; Whipple and Tucker, 1999], therefore knickpoint-bounded profile segments are
254
transient pieces of a profile that represent some prior set of relief conditions. Whether relief is
9
255
generated through base level fall in the downstream end or uplift of the headwaters, a
256
perturbation in the relief structure is produced. In northern Chile, base level remained nearly
257
fixed during the period between 10 and 5.4 Ma, as evidenced by the incision of Quebrada
258
Tiliviche in the north and the closed/effectively closed Pampa del Tamarugal to the south, while
259
the total relief for the basins draining the western slope changed because of uplift. Each profile
260
segment reflects different relief conditions and therefore can be used to estimate the amount of
261
relief generated between the knickpoint bounded segments (i.e. Clark et al [2005]).
262
Knickpoints or broader knick zones are discontinuities in river profiles that can be large
263
or small and generally are introduced into a profile by a perturbation in either side of equation
264
(1). The perturbations may be either a change of erosion rate (through climate change), or a
265
change in downstream slope caused by adjustment to a new base level through capture or
266
breakthrough, or a change in upstream slope caused by tectonic uplift or subsidence.
267
Conceptually, tectonic uplift can be viewed as a relative base level fall. Irrespective of the
268
ultimate cause, knickpoints create anomalous zones of increased slope that result in focused
269
erosion. According to Whipple and Tucker [1999], knickpoints should migrate upstream as a
270
wave from their point of origin. The section of the stream that lies upstream from the knickpoint
271
operates under the same set of conditions as before, while points downstream of the knickpoint
272
have adjusted to a new equilibrium profile. In areas where the channel incsion rate () is low, the
273
propagation of the knickpoint wave upstream through the profile may be sufficiently slow that it
274
can be observed and measured. Here we surmise that the conditions in northern Chile favor the
275
slow upstream propagation of knickpoints.
276
The river profile data presented here were extracted from Shuttle Radar Topography
277
Mission (SRTM) digital topography corrected for data dropouts using a linear interpolation
278
scheme. Fortunately the SRTM data for this region do not contain large (> 0.5 km2) zones of data
279
dropout. Large flat areas constituting salars (salt pans) were removed from the topographic data
280
using the mapped distribution of salars from a 1:1,000,000 scale geologic map of Chile
281
[SERNAGEOMIN, 2002] prior to the extraction of directional and area accumulation grids.
282
Individual drainage basins were identified and extracted from the larger grids of area and
283
elevation for analysis. Because two different classes of drainage basins exist in our study area,
284
two separate sets of criteria were used to extract drainage basins from the topographic data. In
285
the region where rivers drain into the Pampa del Tamarugal Basin (19.5-22°), outlets were
10
286
picked at alluvial fan heads, which generally mark the transition between the alluvial
287
(depositional) and confined (net erosional) segments of the rivers. Drainage basins that reach the
288
Pacific Ocean were assigned outlets at sea level. River profiles were extracted from these data
289
and analyzed following the procedures outlined by Wobus et al. [2006]. Due to the medium
290
spatial resolution of the topographic data, the minimum area cutoff used for river profiles was
291
121 grid cells or 0.98 km2, the observed upper limit of the transition between colluvial and
292
alluvial processes [Montgomery and Foufoula-Georgiou, 1993]. The elevation data for each
293
profile were binned in 30-meter increments, mildly filtered with an algorithm that removes
294
artificial spikes in the data and smoothed with a 500 m (horizontal) moving window [Wobus, et
295
al., 2006]. In most cases, longitudinal profiles and plots of slope vs. area were examined for
296
several tributaries in each drainage basin. Knickpoints and other profile discontinuities were
297
mapped on the profiles. The parameters used to model individual stream segments, the steepness
298
index (ksn) and concavity index (), were derived from regression of channel slope versus
299
cumulative drainage area in log-log space. While we apply the methods presented above,
300
particularly equations (3) and (4) to project the stream segments, we find it necessary to point out
301
that the end result would likely vary very little from results obtained by the application of a
302
simple power law curve-fitting to the profile segments.
303
304
305
Results
The rivers that drain the western Andean mountain front exhibit a wide variety of profile
306
forms (e.g. Figure 5; Figure 7), but generally have at least 2 and often 3 distinct concave-up
307
knickpoint-bounded segments, an upper, middle and lower (Figure 5). The river segments
308
studied here are broadly similar in form to the of Schoenbohm et al. [2004] study of tributaries
309
draining into the Red River in China. Lower segments reflect either the modern transition
310
between alluvial fans and confined channels in the Pampa del Tamarugal (Figure 5b; Figure 7) or
311
the Pacific Ocean (Figure 5a). Middle segments are generally steeper and can contain linear
312
segments and broad convex knick zones marking their upper limits; the highest concavity index
313
and ks values are found in these segments. The upper segments almost universally fall within the
314
‘typical’ range of the concavity index (see discussion above) with values between 0.15 and 1.0.
315
316
For reasons detailed in the three paragraphs that follow, we attribute the generation of
knickpoints on the western slope to be the result of tectonic uplift of the western Andean
11
317
mountain front and adjacent Altiplano, rather than the result of climate change or sudden base
318
level fall related to breakthrough of streams to the Pacific Ocean (see below) or relative
319
subsidence of the Pampa del Tamarugal Basin. The generation of knickpoints remains
320
controversial, since it involves many factors related to lithology (K in eq. 1), climate (K in eq. 1)
321
and uplift (U in eq .1) and how these factors contribute to channel form. It is possible that
322
knickpoints or waves of erosion generated in part of the profile where the rock is easily eroded
323
could accumulate spatially on more resistant lithologies creating broader knick-zones.
324
Since the climate of Northern Chile has been generally arid to hyper-arid since the middle
325
Miocene [Alpers and Brimhall, 1988; Rech, et al., 2006; Sillitoe and McKee, 1996], we reason
326
that the episodic nature of precipitation that accompanies arid conditions has not changed in a
327
way that would result in significant profile modification (i.e. knickpoint generation), although it
328
is difficult to rule out entirely. While climatic fluctuations have been detected regionally
329
[Betancourt, et al., 2000; Bobst, et al., 2001; Latorre, et al., 2002; Rech, et al., 2002], Rech et al.
330
[2006] argue that time averaged precipitation did not exceed 200 mm/yr for the entire area.
331
In this part of northern Chile there are streams that outlet to the Pacific Ocean and those
332
that are effectively closed basins, yet both classes of streams contain prominent knickpoint
333
bounded segments and prominent knickzones (Figure 5). For many catchments that drain into the
334
Pacific Ocean, the wave of erosion has migrated back substantially in the trunk streams, but this
335
wave has not propagated through all of the tributaries (Figure 5a ‘cam2’ and ‘cam 3’ trunk
336
stream profiles, compared with the ‘cam1’ profile and the suca profiles).
337
Surface uplift of the western mountain front and Altiplano Plateau relative to the Central
338
Depression would not occur as a uniform step function all across the region being uplifted, but
339
rather as a ramp function or a sigmoidal function (Figure 8). This logic follows from the
340
common observation that crustal thickness varies laterally in a continuous manner, thus the
341
amount of uplift should decrease smoothly towards the unthickened crust of the forearc
342
potentially resulting in a ramped or sigmoidal shape. Either function would result in a rotational
343
steepening of the catchment. Despite this deviation from the step-function assumed by Whipple
344
and Tucker [1999] and Snyder et al. [2000], the end result of ‘relative base level fall’ related to
345
the uplift should be the same and induce the formation of a knickpoint that allows its catchment
346
to readjust to its new base level. Kirby and Whipple [2001] improved upon the step function
347
models initially presented in Whipple and Tucker [1999] and modeled the evolution of channel
12
348
profiles by using a simple power law uplift function. In the case where the exponent was
349
negative, and uplift decreased away from the headwaters, the concavity () values were high
350
(>1). If uplift increased away from headwater regions, the concavity values were low and even
351
negative (forming convexities) in some cases. In general, stream segments in the lower and
352
middle parts of the northern Chilean streams considered here have () values > 1, suggesting a
353
decreasing uplift rate to the west. The sigmoidal uplift function that is expected for the western
354
Andean mountain front (Figure 8) has a near constant amount of uplift in the higher regions
355
(east) that gently rolls over into a steep gradient in uplift rates towards the west before decreasing
356
to zero. A simple numerical experiment using a stream power based model for which uplift
357
conditions changed from an initial steady-state profile, produced with a uniform uplift pattern, to
358
a sigmoidal-shaped uplift pattern confirmed that a sigmoidal uplift pattern would indeed produce
359
a knickpoint. Knickpoint generation occurred at the base of the profile outlet and migrated
360
upstream producing a geometry similar to what we observe for the rivers of the western Andean
361
slope.
362
Assuming a tectonic origin for the knickpoint bounded segments, we modeled the
363
downstream projection of river longitudinal profiles using equation (4), as illustrated in Figure
364
7a, for a total of 34 knickpoint-bounded profile segments from 15 of the 20 catchments that drain
365
the western slope (Figure 9a). The parameters ( and ksn) used to project the profiles are from
366
linear regressions of the stream segments in slope-area plots (Table 1 and supplementary data 2).
367
The outlet elevations of the projected stream profiles (extended to the same horizontal position as
368
the modern outlet) are between 0 m and 3672 m, inclusive of the profiles projected for lower,
369
upper and middle stream segments (Figure 9a). Because the inferred uplift pattern for the
370
western slope results in westward rotational steepening of the catchments, our modeled outlet
371
elevations are minimum estimates.
372
An especially interesting measure is the relief between the modeled paleoprofile outlet
373
constructed for the middle segments of streams compared to the actual elevation of the modern
374
outlet (Figure 9a). If tectonic forces are responsible for the relief generation, as we postulate
375
here, then the difference between the outlets serves as a measure of uplift. When considering the
376
measured relief, rivers that presently drain into the Pacific Ocean contain ~1 km more relief than
377
the rivers that drain into the Pampa del Tamarugal Basin (difference in elevation between the
378
crosses and triangles in Figure 9a). This difference in relief invites two different interpretations:
13
379
either an additional kilometer of uplift has occurred in the area north of 19.75°S, or the projected
380
middle segments of the northern streams formed relative to a base-level equivalent to the
381
elevation of the modern Central Depression, i.e. prior to whatever event caused the incision of
382
the Coastal Cordillera and the Central Depression.
383
In order to isolate the impact of tectonic uplift of the western slope, we subtract the relief
384
generation caused by the incision event across the Coastal Cordillera. This correction for
385
“Coastal Cordillera incision” is quantified as follows. Profiles from particularly inactive
386
tributaries in two separate catchments that drain into the Pacific Ocean contain rivers with lower
387
profile segments having downstream projections that result in outlet elevations of 1070± 110 m
388
and 1200 ± 100 m. These elevations correspond with elevations of the El Diablo pediment
389
surface that forms the broad flat interfluves between the canyons. The 1070-1200 m elevation
390
range also corresponds to the outlet elevations for the modern streams that drain into the
391
effectively closed Pampa del Tamarugal Basin to the south. In three northern drainages
392
(Tiliviche, Suca-Camarones, and Vitor), we interpret these projections to represent the prior base
393
level of the mountain front drainage system, before the streams equilibrated to the modern ~
394
1000 m relief between the Pacific Ocean and the Central Depression. Based on the dated river
395
terrace in the Quebrada de Tiliviche (see above), where it currently incises across the Coastal
396
Cordillera, we interpret that the rivers to the north were graded to the Pampa del Tamarugal base
397
level prior to ~ 6.4 Ma. The Río Azapa valley, which outlets where there is no Coastal
398
Cordillera, was also graded to the 1000 m high El Diablo surface, so we can apply a similar
399
correction to its projected stream segments. For basins where lower profile stream segments
400
project near the closed basin paleo-base level, that value was applied to correct for “Coastal
401
Cordillera incision”, otherwise the stream base level is set to 1000 m. Such a correction is not
402
necessary for the closed basins south of the Quebrada Tiliviche, where no ambiguity exists
403
regarding the lower river segment base level history.
404
Applying the base level correction to the streams north of the modern Pampa del
405
Tamarugal closed basin, the 21 modeled profiles that represent the middle segments have an
406
average relief between the modeled paleo-profile outlet compared to the actual/corrected
407
elevation of the modern outlet of 1080 ± 230 m (Figure 9b). We interpret the consistent
408
clustering of relief over a 350 km long stretch of the western mountain front to represent
14
409
tectonically generated relief related to the relative uplift of the western mountain front and
410
adjacent Central Andean Plateau with respect to the Central Depression.
411
In the preceding we have assumed, for the sake of simplicity, that the elevations of the
412
Coastal Cordillera and Central Depression have remained similar to their present values
413
throughout the time of stream profile preservation. However, there are no constraints. While this
414
does not affect the relative uplift interpretations for the modeled stream segments, it does leave
415
alternative interpretations for the lower stream segments in the northern streams. Specifically, we
416
can not differentiate between a model where base-level fell through drainage breakthrough of the
417
Coastal Cordillera and a model where the northern streams persistently outlet to the ocean and
418
~1 km of wholesale uplift of the Coastal Cordillera and Central Depression occurred following
419
6.4 Ma [García and Hérail, 2005]. The lack of accommodation structures in the seismic data
420
between the Central Depression and the western mountain front [Digert, et al., 2003], implies
421
that this uplift may also affect the elevations on the plateau.
422
The modeled projections of the upper profile segments (11 profiles) have a wide range of
423
projected outlet elevations, between -10 and 3675 m.s.l. (Figure 9a). This great scatter highlights
424
the difficulty in deciphering the evolution of these highest stream segments with respect to the
425
rest of the catchment. These parts of the stream network reach high in the Western Cordillera and
426
Altiplano, where a long history of volcanic edifice construction and early Cenozoic deformation
427
complicate the set of controls. There was a pervasive influence of middle Miocene to Pliocene
428
volcanic activity as far south as 20°S, as well as the existence of prior mountainous relief related
429
to the Incaic (Eocene-Oligocene) deformation in areas south of 20°S, each of which makes it
430
difficult to ascertain whether or not the highest reaches of the profiles were always linked to their
431
current trunk streams or are the result of more recent captures. Because it is uncertain that these
432
upper stream segments evolved as part of their current catchment basin, we do not consider these
433
upper river profile projections to be an accurate indicator of tectonic relief creation.
434
The timing of uplift recorded in the middle stream segments can be inferred from the age
435
of the stream segments modeled. As all middle segments dissect the 10 Ma (or 8.5-7.5 Ma
436
proposed by von Rotz et al., 2005) pediment surfaces that cap the El Diablo and Altos de Pica
437
Formations, the modeled segments and their uplift bet younger than the pediment surface. The
438
middle profile segments of suc003 and tili01a (table 5.1) also deeply dissect (>500 m) the 8-9
439
Ma Tana lava [Pinto, et al., 2004] and underlying units, indicating that these modeled stream
15
440
segments are younger than 8-9 Ma. Lastly, the middle profile segments of sup04, tamb03 and
441
arc01 transect, with only 10’s of meters of incision, valleys that contain terraces built of alluvial
442
gravels and the intercalated 5.4 Ma Carcote ignimbrite (Figure 4). In downstream areas of the
443
same middle profile segments, the valleys contain strath terraces capped by Carcote ignimbrite
444
high on the canyon walls [Tomlinson, et al., 2001]. We use the projection of the river profiles
445
and the average elevations of areas mapped as Carcote in Tomlinson et al. [2001] as a means of
446
evaluating the age of incision of the rivers (Figure 10). In the Quebrada Sipuca (Figure 10a), the
447
elevation of the Carcote exposures corresponds well with the actual middle stream segment for
448
sup04 and modeled downstream projection of the middle segment between 40 and 33 km along
449
the profile, then the Carcote terraces decrease sharply in elevation (33 to 26 km in figure 10a
450
inset). We interpret the Carcote Ignimbrite capped terraces between ~36 km and 26 km to reflect
451
the paleo-stream profile at 5.4 Ma. Clearly, incision of the middle segment was well under way
452
in the Quebrada Sipuca by 5.4 Ma and suggests that surface uplift had to have occurred before
453
deposition of the Carcote ignimbrite. The correspondence between the elevation of the Carcote
454
remnants and the modeled middle segment are consistent with an interpretation in which the
455
lower profile knickpoint erosional wave has deeply dissected the lower profile segments and is
456
currently working on the downstream limit of the uplifted middle profile segments. Furthermore
457
the correspondence of the modeled stream profile with the now stranded exposures of Carcote
458
Ignimbrite confirm the validity of the profile modeling approach applied here. Despite similar
459
occurrences of the Carcote in the Quebrada Tambillo (tamb03) and the Quebrada Arcas (arc01)
460
profiles, not enough of the downstream extent of the ignimbrite terraces is preserved to allow for
461
similar observations (Figure 10 b,c). The timing of uplift is similar in age, and perhaps
462
contemporaneous with the 6.4Ma age of down cutting of the northern rivers to the ocean deduced
463
from relations in the Quebrada Tiliviche. As the persistence of profile segments depends heavily
464
on the rate of headward propagation of the knick zones, the suggestion is that the rate of
465
propagation must have been slow, as evidenced by the slight advance of the Quebrada Tiliviche
466
knick zone over the last 6.4 Ma [Hoke, 2006] and the 3 km of retreat since 5.4 Ma in the
467
Quebrada Sipuca.
468
469
Discussion
16
470
Our data show clear evidence for at least 1 km of surface uplift of the western Andean
471
mountain front with respect to the Central Depression during the last 10 Ma. This result is
472
unambiguous between 19.75° - 22°S where the streams drain to the modern closed Pampa del
473
Tamarugal basin. In the area spanning 18.5° - 19.75°S we reach this conclusion after subtracting
474
~ 1000 m for the base level change between the El Diablo/Altos de Pica surface and the modern
475
river outlets to the Pacific Ocean. This is an important new constraint that tests Isacks [1988]
476
hypothesis of passive surface uplift of the western Andean slope and Muñoz and Charrier [1996]
477
and Victor et al.’s [2004] model of upper crustal fault control on the western mountain front. Our
478
data can only track the creation of increased relief, and not absolute elevation change, because
479
the absolute elevation of our reference datum, the Central Depression, is not constrained. The
480
100s of meters of incision of the El Diablo surface in the Lluta and Azapa valleys cross the
481
Central Depression would suggest that some uplift of the coastal region had to have occurred
482
within the last 10 Ma.
483
The dynamics of the uplift model proposed by Isacks [1988] call for the formation of the
484
western mountain front to be controlled by thermal and mechanical weakening of the South
485
American lithosphere, due to subduction zone processes. In the model, the location of the
486
mountain front coincides with the boundary between strong and weak lithosphere, which in turn
487
coincides with the western tip of the asthenospheric wedge under the South American
488
lithosphere [Isacks, 1988]. The model proposes a first phase of uplift to be the consequence of
489
widespread horizontal shortening and crustal thickening during the ‘Quechua’ phase of
490
deformation. In the second stage of the model, when upper crustal deformation shifts eastward
491
and the eastern foreland fold-and-thrust belts develop, the plateau grows higher in response to
492
lower crustal ductile shortening and thickening driven by addition of mass from the
493
underthrusting of the Bolivian foreland [Isacks, 1988]. The timing of the initiation of the second
494
stage of deformation was refined by Allmendinger et al. [1997] to be 12 to 6 Ma. The cessation
495
of deformation in the Eastern Cordillera was at ~ 10 Ma with earliest deformation in the Sub-
496
Andean fold-and-thrust belt occurring as early as 9 Ma [Echavarria, et al., 2003]. The Isacks
497
[1988] model treats the western Andean mountain front as a passive consequence of uplift and
498
monoclinal tilting, driven by the tectonics within and east of the plateau.
499
500
In contrast, the faulting models of Muñoz and Charrier [1996] and Victor et al. [2004]
suggest that the uplift of the western mountain front can be explained solely by displacements on
17
501
a west-verging fault system that is exposed discontinuously from 18°30’S to ca. 21°30’S [Digert,
502
et al., 2003; García, 2002; García, et al., 2004; Muñoz and Charrier, 1996; Victor, et al., 2004].
503
Throughout the system of faults and fault-related folds, the ages of unconformities, growth folds
504
and syntectonic conglomerates indicate deformation is pre-6 Ma, with the greatest shortening
505
rates constrained to be older than 10 Ma [García, et al., 2004; García and Hérail, 2005; Pinto, et
506
al., 2004; Victor, et al., 2004]. Only minor post-6 Ma faulting and folding (< 100 m
507
displacement) occurs [García, et al., 2004; Pinto, et al., 2004]. For example, the modeled middle
508
segments of stream profiles suc003 and tili01a (Table 1) lie east of the Moquella flexure, a 4 km
509
wide growth fold [Pinto, et al., 2004]. Pinto et al. [2004] document ca. 650 m of uplift across the
510
Moquella flexure between 25 Ma and the deposition of the 8-9 Ma Tana Lava, but only 50 m of
511
uplift younger than 8-9 Ma. The modeled stream segments crosscut, and therefore, postdate the
512
Tana Lava, but the 50 m of uplift by post-Tana flexural folding is insufficient to explain the ca. 1
513
km of uplift indicated by the projected stream profile outlets.
514
Shortening between 25 and 10 Ma along the western mountain front [García, et al., 2004;
515
García and Hérail, 2005; Pinto, et al., 2004; Victor, et al., 2004]is compatible with both Isacks’s
516
[1988] first stage of plateau growth and the faulting models of Muñoz and Charrier [1996] and
517
Victor et al. [2004], but the two models depart in regards to their predictions for the post-10 Ma
518
uplift history. With the paucity of post-10 Ma fault activity on the west flank of the plateau
519
[Elger, et al., 2005; Victor, et al., 2004], the upper crustal faulting model implies that significant
520
uplift of the western edge of the plateau did not continue into the late Miocene. This prediction
521
is contradicted by paleoelevation estimates from paleobotanical and stable isotope studies of
522
paleosols which indicate large amounts of young uplift of the plateau, with more than 50% of the
523
plateau elevation being attained in the last 10 Ma [Garzione, et al., 2006; Ghosh, et al., 2006;
524
Gregory-Wodzicki, 2000; Rech, et al., 2006]. The early-uplift requisite of the faulting model is
525
also incompatible with the geomorphic results presented herein and other geomorphic studies
526
which likewise indicate important post-10 Ma uplift [Barke and Lamb, 2006; Gubbels, et al.,
527
1993; Hoke, et al., 2004; Kennan, et al., 1997; Schildgen, et al., in press; Worner, et al., 2002]
528
Another important distinction between Isacks [1988] model and the Muñoz and Charrier
529
[1996] and Victor et al [2004] model is their degree of consistency with isostatic compensation.
530
Since the ratio of positive topography to continental root is ~ 1:5 [Lowrie, 1997], the 3-3.5 km
531
of Neogene relief generation postulated for the Central Andean Plateau requires ~ 20 km of
18
532
Neogene crustal thickening. This value is in agreement with Eocene and current crustal thickness
533
estimates under the western mountain front. Geochemical proxy data suggest that following
534
Eocene magmatism and the Incaic shortening event, crustal thicknesses were ~ 45 km at 35 Ma
535
[Haschke and Gunther, 2003; Haschke, et al., 2002], whereas geophysical data determine its
536
current thickness to be 65-70 km [Beck, et al., 1996; Wigger, et al., 1994]. Shortening estimates
537
from balanced cross sections along the western mountain front and western edge of the plateau
538
(Western Cordillera) indicate 3 to 8 km of horizontal shortening that produces ~3-7 km of crustal
539
thickening, far less than is required to explain the thickened Neogene crustal root. Since most of
540
the shortening along the western mountain front occurred between 25 and 10 Ma, it would have
541
been an important contribution to topographic development along the western edge of the
542
Central Andean Plateau during the early and middle Miocene, in accordance with both stage 1 of
543
Isacks’s model and the fault uplift model. However, the timing relationships clearly reveal that
544
significant uplift of the Central Andean Plateau younger than 10 Ma cannot be caused by the
545
west-vergent faults in northern Chile.
546
We clearly document that uplift persisted into the latest Miocene. In this respect, our
547
results provide evidence in support of the Isacks [1988] model for the uplift of the Altiplano.
548
Specifically, we demonstrate that uplift of the western flank of the plateau continued as the
549
eastern extent of Andean crustal shortening shifted eastward into what is now the Sub-Andean
550
fold-thrust belt. The paucity of significant post-10 Ma faulting activity in our study area suggests
551
that ~1000 m of late Miocene uplift accompanied a lithospheric phenomenon that is not
552
associated with local faulting and folding. Two likely candidates are lower crustal flow of
553
material from the east in response to upper crustal shortening in the Sub-Andean fold-thrust belt
554
and underthrusting of the Brazilian shield [Isacks, 1988], or removal of a dense lower crust
555
and/or lithospheric mantle [Garzione, et al., 2006; Kay and Kay, 1993]. Mounting evidence
556
exists in support of both of these mechanisms in the generation of continental plateaus [e.g.
557
Clark and Royden, 2000; Kay, et al., 1994]. Our results are consistent with elevation increase
558
since 10 Ma, and so are consistent with the paleoelevation estimates of Gregory-Wodzocki
559
[2000] and Rech et al. [2006]. Our evidence for post-10 Ma uplift is similar to the timing of
560
uplift suggested by Garzione et al. [2006] and Ghosh et al. [2006]. Furthermore our results lend
561
support to other geomorphic evidence that suggests uplift more recent than 10 Ma, most notably
562
the incision of the San Juan del Oro surface [Barke and Lamb, 2006; Gubbels, et al., 1993;
19
563
Kennan, et al., 1997; Schildgen, et al., in press], and the tilting and dissection of the ~10 Ma
564
Atacama Gravels south of 25°S [Mortimer, 1973; Riquelme, et al., 2003].
565
566
567
Conclusions
Downstream modeling of knickpoint bounded stream segments on the western Andean
568
mountain front reveal an ~1 km increase in relief, and landform relationships to dated strata
569
reveal that the timing of relative uplift was between within the last 10 Ma. Previous models for
570
northern Chile calling upon only involvement of local deformation predicted the uplift of the
571
plateau to have been nearly complete by the end of the middle Miocene. The late Miocene relief
572
documented herein was created in the absence of significant deformation (i.e. faulting or folding)
573
in the area, suggesting that processes occurring in the lower crust (i.e. lower crustal flow) or
574
upper mantle (lithospheric delamination) are responsible for the growth of the Central Andean
575
Plateau during this period. Our results confirm the presence and timing of the second stage of
576
plateau growth advocated by Isacks [1988].
577
578
Acknowledgements
579
This material is based upon work supported by the National Science Foundation under grant No.
580
0208130 to TEJ and BLI, and supported by NASA Headquarters under Earth System Science
581
Fellowship Grant NGT5-30461 to GDH. We are indebted to Carlos Pérez de Arce for analysis of
582
the biotites. Kelin Whipple and his research group helped acquaint GDH with the stream profile
583
analysis methods. Discussions with Carmala Garzione and Noah Finnegan were very helpful and
584
have improved the ideas presented here.
585
586
587
588
589
590
591
592
593
594
595
596
References
Allmendinger, R. W., et al. (1997), The evolution of the Altiplano-Puna plateau of the Central
Andes, Annu Rev Earth Pl Sc, 25, 139-174.
Alpers, C. N., and G. H. Brimhall (1988), Middle Miocene Climatic-Change in the Atacama
Desert, Northern Chile - Evidence from Supergene Mineralization at La-Escondida, Geol Soc Am
Bull, 100, 1640-1656.
Barke, R., and S. Lamb (2006), Late Cenozoic uplift of the Eastern Cordillera, Bolivian Andes,
Earth Planet Sc Lett, 249, 350-367.
Beck, S. L., et al. (1996), Crustal-thickness variations in the central Andes, Geology, 24, 407410.
20
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
Betancourt, J. L., et al. (2000), A 22,000-year record of monsoonal precipitation from Northern
Chile's Atacama Desert, Science, 289, 1542-1546.
Bobst, A. L., et al. (2001), A 106 ka paleoclimate record from drill core of the Salar de Atacama,
northern Chile, Palaeogeogr Palaeocl, 173, 21-42.
Clark, M. K., et al. (2005), The non-equilibrium landscape of the souther Sierra Nevada,
California, GSA Today, 15, 4-9.
Clark, M. K., and L. H. Royden (2000), Topographic ooze: Building the eastern margin of Tibet
by lower crustal flow, Geology, 28, 703-706.
Digert, F. E., et al. (2003), Subsurface Stratigraphy of the Neogene Tamarugal Basin, Northern
Chile, paper presented at X Congreso Geológico Chileno, Concepción, Chile.
Dunai, T. J., et al. (2005), Oligocene-Miocene age of aridity in the Atacama Desert revealed by
exposure dating of erosion-sensitive landforms, Geology, 33, 321-324.
Echavarria, L., et al. (2003), Subandean thrust and fold belt of northwestern Argentina; geometry
and timing of the Andean evolution, AAPG Bulletin, 87, 965-985.
Elger, K., et al. (2005), Plateau-style accumulation of deformation: Southern Altiplano,
Tectonics, 24, -.
Evenstar, L., et al. (2005), Miocene-Pliocene climate change in the Peru-Chile Desert, paper
presented at 6th International Symposium on Andean Geodynamics, Institut de Recherche pour
le développment, Barcelona, Spain, September 12-14, 2005.
Farias, M., et al. (2005), Late Cenozoic deformation and uplift of the western flank of the
Altiplano: Evidence from the depositional, tectonic, and geomorphologic evolution and shallow
seismic activity (northern Chile at 19 degrees 30 ' S), Tectonics, 24, -.
Galli Olivier, C., and R. J. Dingman (1962), Cuadrangulos Pica, Alca, Matilla y Chacarilla, con
un estudio sobre los recursos de agua subterranea, provincia de Tarapaca (Carta geologica de
Chile v. 3, no. 2-5), 125 pp.
García, M. (2002), Evolution Oligo-Miocene de L'Altiplano Occidental (Arc et Avant-Arc du
Nord Chili, Arica), 118 pp, Grenoble, France.
García, M., et al. (2004), Hoja Arica, Región de Tarapacá, Servicio Nacional de Geología y
Minería, Santiago, Chile.
García, M., and G. Hérail (2005), Fault-related folding, drainage network evolution and valley
incision during the Neogene in the Andean Precordillera of Northern Chile, Geomorphology, 65,
279-300.
Garzione, C. N., et al. (2006), Rapid late Miocene rise of the Bolivian Altiplano: Evidence for
removal of mantle lithosphere, Earth Planet Sc Lett, 241, 543-556.
Ghosh, P., et al. (2006), Rapid uplift of the Altiplano revealed through C-13-O-18 bonds in
paleosol carbonates, Science, 311, 511-515.
Gregory-Wodzicki, K. M. (2000), Uplift history of the Central and Northern Andes: A review,
Geol Soc Am Bull, 112, 1091-1105.
Gubbels, T. L., et al. (1993), High-Level Surfaces, Plateau Uplift, and Foreland Development,
Bolivian Central Andes, Geology, 21, 695-698.
Hack, J. T. (1957), Studies of longitudinal stream profiles in Virginia and Maryland, U. S.
Geological Survey, Reston, VA.
Haschke, M., and a. Gunther (2003), Balancing crustal thickening in arcs by tectonic vs.
magmatic means, Geology, 31, 933-936.
Haschke, M., et al. (2002), Repeated crustal thickening and recycling during the Andean orogeny
in north Chile (21 degrees-26 degrees S), J Geophys Res-Sol Ea, 107, -.
21
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
Hoke, G. D. (2006), The influence of Climate and Tectonics on the Geomorphology of the
Western Slope of the Central Andes, Chile and Peru, Doctoral thesis, 283 pp, Cornell University,
Ithaca, NY.
Hoke, G. D., et al. (2004), Groundwater-sapping origin for the giant quebradas of northern Chile,
Geology, 32, 605-608.
Houston, J., and A. J. Hartley (2003), The central andean west-slope rainshadow and its potential
contribution to the origin of HYPER-ARIDITY in the Atacama desert, Int J Climatol, 23, 14531464.
Isacks, B. L. (1988), Uplift of the Central Andean Plateau and Bending of the Bolivian Orocline,
J Geophys Res-Solid, 93, 3211-3231.
Jordan, T. E., et al. (1997), Variability in Age and Initial Shortening and Uplift in the Central
Andes, in Tectonic Uplift and Climate Change, edited by W. F. Ruddiman, pp. 41-61, Plenum
Press, New York.
Kay, R. W., and S. M. Kay (1993), Delamination and Delamination Magmatism,
Tectonophysics, 219, 177-189.
Kay, S. M., et al. (1994), Young Mafic Back Arc Volcanic-Rocks as Indicators of Continental
Lithospheric Delamination beneath the Argentine Puna Plateau, Central Andes, J Geophys ResSol Ea, 99, 24323-24339.
Kennan, L., et al. (1997), High-altitdue paleosurfaces in the Bolivian Andes: evidence for late
Cenozoic surface uplift, in Paleosurfaces: Recognition, Reconstruction and Paleoenvironmental
Interpretation, edited by M. Widdowson, pp. 307-323, Geological Society of London, London.
Kiefer, E., et al. (1997), Gravity-based mass balance of an alluvial fan giant: the Arcas Fan,
Pampa del Tamarugal, Northern Chile, Rev Geol Chile, 24, 165-185.
Kirby, E., and K. Whipple (2001), Quantifying differential rock-uplift rates via stream profile
analysis, Geology, 29, 415-418.
Kober, F., et al. (2006), Surface Uplift and Climate Change: The geomorphic evolution of the
Western Escaprment of the Andes of Northern Chile between the Miocene and Present, in
Tectonics, Climate and Landscape Evolution, edited by S. D. Willett, et al., pp. 75-86,
Geological Society of America, Denver, Colorado.
Lamb, S., and L. Hoke (1997), Origin of the high plateau in the Central Andes, Bolivia, South
America, Tectonics, 16, 623-649.
Latorre, C., et al. (2002), Vegetation invasions into absolute desert: A 45 000 yr rodent midden
record from the Calama-Salar de Atacama basins, northern Chile (lat 22 degrees-24 degrees S),
Geol Soc Am Bull, 114, 349-366.
Lowrie, W. (1997), Fundamentals of Geophysics, 1st ed., 354 pp., Cambridge University press,
Cambridge, United Kingdom.
McQuarrie, N. (2002), The kinematic history of the central Andean fold-thrust belt, Bolivia:
Implications for building a high plateau, Geol Soc Am Bull, 114, 950-963.
McQuarrie, N., et al. (2005), Lithospheric evolution of the Andean fold-thrust belt, Bolivia, and
the origin of the central Andean plateau, Tectonophysics, 399, 15-37.
Montgomery, D. R., et al. (2001), Climate, tectonics, and the morphology of the Andes, Geology,
29, 579-582.
Montgomery, D. R., and E. Foufoula-Georgiou (1993), Channel network source representation
using digital elevation models, Water Resour Res, 29, 3925-3934.
Mortimer, C. (1973), The Cenozoic history of the southern Atacama Desert, Chile, Journal of the
Geological Society of London, 129, 505-526.
22
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
Mortimer, C. (1980), Drainage evolution in the Atacama Desert of northernmost Chile, Rev Geol
Chile, 11, 3-28.
Mortimer, C., et al. (1974), K-Ar ages from Tertiary lavas of the northernmost Chilean Andes,
Geol Rundsch, 63, 484-490.
Mortimer, C., and N. Saric Rendic (1975), Cenozoic studies in northernmost Chile, Geol
Rundsch, 64, 395-420.
Mpodozis, C., and V. Ramos (1989), The Andes of Chile and Argentina, in Geology of the Andes
and its relation to hydrocarbon and mineral resources, edited by G. E. Ericksen, et al., pp. 5690, Circum-Pacific Council for Energy and Mineral Resources, Houston, Texas.
Muzzio J., G. (1986), Geologia de los cuadrángulos Caleta Camarones, Cuya, Punta Gorda, y
Cerro Atajaña, I region, Chile: Informe de Avance, SERNAGEOMIN, Santiago.
Muzzio J., G. (1987), Geologia de los cuadrángulos Pisagua, Zapiga, Caleta Buena y Huara:
informe de Advance Proyecto hoja Pisagua, 2 maps 1:100,000, SERNAGEOMIN, Santiago de
Chile, 1987.
Muñoz, N., and R. Charrier (1996), Uplift of the western border of the Altiplano on a westvergent thrust system, Northern Chile, J S Am Earth Sci, 9, 171-181.
Muñoz, N., and P. Sepulveda (1992), Estructuras compresivas con vergencia al oeste en el borde
oriental de la Depresion Central, norte de Chile (19 degrees 15'S), Rev Geol Chile, 19, 241-247.
Pinto, L., et al. (2004), Sedimentación sintectónica asociada a las estructuras néogeneas en la
Precordillera de la zona de Moquella, Tarapacá (19°15'S, norte de Chile), Rev Geol Chile, 31,
19-44.
Rech, J. A., et al. (2006), Neogene Climate Change and Uplift in the Atacama Desert, Chile,
Geology, 34, 761-764.
Rech, J. A., et al. (2002), Late Quaternary paleohydrology of the central Atacama Desert (lat 22
degrees-24 degrees S), Chile, Geol Soc Am Bull, 114, 334-348.
Riquelme, R., et al. (2003), A geomorphological approach to determining the Neogene to Recent
tectonic deformation in the Coastal Cordillera of northern Chile (Atacama), Tectonophysics, 361,
255-275.
Salas, R., et al. (1966), Geologa y recursos minerales del departamento de Arica, provincia de
Tarapaca, Servicio Nacional de Geologia y Mineria, Santiago, Chile.
Schildgen, T., et al. (in press), Uplift of the western margin of the Andean Plateau revealed from
canyon incision history, southern Peru, Geology.
Schoenbohm, L. M., et al. (2004), Geomorphic constraints on surface uplift, exhumation, and
plateau growth in the Red River region, Yunnan Province, China, Geol Soc Am Bull, 116, 895909.
SERNAGEOMIN (2002), Mapa Geológico de Chile, 3 sheets, Servicio Nacional de Geología y
Minería, Santiago, Chile.
Sillitoe, R. H., and E. H. McKee (1996), Age of supergene oxidation and enrichment in the
Chilean porphyry copper province, Econ Geol Bull Soc, 91, 164-179.
Silva, L. I. (1977), Hojas Pisagua y Zapiga, I Region, Instituto de Investigacines Geologicas,
Santiago de Chile.
Snyder, N. P., et al. (2000), Landscape response to tectonic forcing: Digital elevation model
analysis of stream profiles in the Mendocino triple junction region, northern California, Geol Soc
Am Bull, 112, 1250-1263.
Sáez, A., et al. (1999), Late Neogene lacustrine record and palaeogeography in the QuillaguaLlamara basin, Central Andean fore-arc (northern Chile), Palaeogeogr Palaeocl, 151, 5-37.
23
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
Tobar, A., et al. (1968), Cuadrángulos Camaraca y Azapa. Provincia de Tarapaca, 20 pp and 22
maps 21:50,000 scale, Instituto de Investigaciones Geologicas, Santiago, Chile, 1968.
Tomlinson, A. J., et al. (2001), Geologia de la Precordillera Andina de Quebrada Blanca Chuquicamata, Regiones I y II [20°30'-22°30'], 444 and 420 Maps pp, Servicio Nacional de
Geología y Minería, Santiago, Chile.
Tosdal, R. M., et al. (1984), Cenozoic Polyphase Landscape and Tectonic Evolution of the
Cordillera Occidental, Southernmost Peru, Geol Soc Am Bull, 95, 1318-1332.
Victor, P., et al. (2004), Uplift of the western Altiplano plateau: Evidence from the Precordillera
between 20 degrees and 21 degrees S (northern Chile), Tectonics, 23, -.
von Rotz, R., et al. (2005), Assessing the age of relief growth in the Andes of northern Chile:
Magneto-polarity chronologies from Neogene continental sections, Terra Nova, 17, 462-471.
Whipple, K. X. (2004), Bedrock rivers and the geomorphology of active orogens, Annu Rev
Earth Pl Sc, 32, 151-185.
Whipple, K. X., et al. (1999), Geomorphic limits to climate-induced increases in topographic
relief, Nature, 401, 39-43.
Whipple, K. X., and G. E. Tucker (1999), Dynamics of the stream-power river incision model:
Implications for height limits of mountain ranges, landscape response timescales, and research
needs, J Geophys Res-Sol Ea, 104, 17661-17674.
Wigger, P., et al. (1994), Variation in Crustal Structure of the Southern Central Andes Deduced
from Seismic Refraction Investigations, in Tectonics of the Southern Cenrtal Andes, edited by
K.-J. Reutter, et al., pp. 23-48, Springer-Verlag, Berlin, Germany.
Wobus, C., et al. (2006), Tectonics from topography: proceedures, promise and pitfalls, in
Tectonics, Climate and Landscape Evolution, edited by S. D. Willett, et al., pp. 55-74,
Geological Society of America, Denver, Colorado.
Wobus, C. W., et al. (2003), Has focused denudation sustained active thrusting at the Himalayan
topographic front? Geology, 31, 861-864.
Worner, G., et al. (2002), Evolution of the West Andean Escarpment at 18 degrees S (N. Chile)
during the last 25 Ma: uplift, erosion and collapse through time, Tectonophysics, 345, 183-198.
Zaprowski, B. J., et al. (2005), Climatic influences on profile concavity and river incision,
Journal of Geophysical Research, 110, 19.
24
767
768
769
Figure 1: Shaded relief topographic image with the locations of the major physiographic
770
elements of the western Andes of northernmost Chile and names of the 20 basins considered in
771
this study. Country boundaries are shown with light gray lines. The towns of Iquique, Pica and
772
Arica are labeled for reference. Between the Quebrada de Tiliviche and Río Loa, all of the
773
drainage basins are closed or effectively closed. The location of the Río Loa is labeled in red.
774
Inset shows the location of the study area, political boundaries, country names and the central
775
Andean plateau delineated by the 3000 meter elevation contour (thick line).
776
777
Figure 2: Geologic map modified from the 1:1,000,000 geologic map of Chile
778
(SERNAGEOMIN, 2002) of the area between Quebradas Chacarilla and Maní draped over the
779
shaded relief topography of the SRTM 90 m DEM. The elevation of the SW corner of the map is
780
~ 1200 m, while the northeast corner is at 4000 m. The 500 m contours on the map show the
781
monoclinal form of the western mountain front and the relatively un-dissected nature of the
782
upper member of the Altos de Pica Formation as it extends from the lower elevations of the
783
Pampa del Tamarugal to higher elevations to the east. The highest islands of Altos de Pica occur
784
at elevations near 4000 m.
785
786
Figure 3: Generalized stratigraphy of the western Andean mountain front between 18°30 and
787
22°S. Age constraints given in the figure are from Mortimer et al. [1974] (Tana Lava), García et
788
al. [2002] (andesite clasts), Worner et al. [20002] (Lauca-Peréz Ignimbrite), Kiefer et al [1997]
789
(Arcas Fan) and this paper (Quebrada Tiliviche, Carcote Ignimbrite, and the 13 Ma ash horizon
790
in Altos de Pica Formation).
791
792
Figure 4: The distribution of the Carcote Ignimbrite in the catchments (heavy black lines) that
793
drain the southern end of the study area (gray polygons). Topographic contours (gray lines) are
794
derived from 90 m SRTM data and have a 500 meter spacing.
795
796
Figure 5: Shaded relief topography, drainage networks, and swath topographic profiles showing
797
the relief structure of basins that drain the western mountain front into A) the Pacific Ocean and
25
798
B) the effectively closed Pampa del Tamarugal Basin. Knickpoints are marked by yellow
799
triangles. A) The Río Camarones-Quebrada Suca catchment with trunk stream and tributaries
800
(colored profiles) from the Río Camarones (‘cam’) and the Quebrada Suca (‘suc’) plotted with
801
the maximum and minimum topography for a 70 km wide swath (black profiles) running along
802
the basin axis. B) The Quebrada Tarapacá catchment with trunk and tributaries (colored
803
profiles) plotted with the maximum and minimum topographic envelopes (black profiles) for a
804
47 km wide swath. All river profiles were projected into the line used to create the swath
805
profiles. Minimum and maximum topographic profiles are determined by selecting the maximum
806
and minimum cell values for each 1 km step length in the swath profiles. Note that the Tarapacá
807
profiles end in the Central Depression and do not outlet to sea level.
808
809
Figure 6: Aerial photo showing the location of the strath terrace preserved in the Quebrada
810
Tiliviche canyon, downstream from the confluence of the Quebradas Camiña and Rematilla. U-
811
Pb dating of zircons in an ash preserved on the terrace yield an age of ~6.4 Ma (see
812
supplementary data 1).
813
814
Figure 7: An example of a stream long profile from the Tarapacá basin with three distinct
815
concave profile segments that are bounded by zones of strong convexities. A. The stream long
816
profile with the downstream projection of the middle stream segment; the different stream
817
segments are labeled accordingly. B. Graph showing cumulative drainage area with distance
818
downstream. C. Plot of stream gradient versus cumulative drainage area for the same stream. The
819
concavity and ks values for the modeled stream segment in A. were derived by the linear
820
regression of slope-area data that correspond to that stretch of the stream (dark blue line). Purple
821
crosses show the slope data corresponding to 30 m elevation bins while the red boxes show a
822
log-binning of the same data.
823
824
825
Figure 8: Different uplift models (dashed lines) plotted against the maximum topography of the
826
Tarapacá Basin (solid line). The uplift models represent the total uplift of the plateau relative to
827
the Central Depression (~25 Ma to Present). Since so little erosion has occurred on the western
26
828
Andean mountain front over the last 10 My, the present topography serves as an approximation
829
for uplift as a function of distance from the Central Depression, U(x).
830
831
Figure 9: A) Projected tributary outlet elevations for 34 streams in 15 catchments that drain the
832
western slope and have well defined knickpoint-bounded segments. Open circles correspond to
833
upper knickpoint-bounded segments, filled trianges middle knickpoint-bouned segments and
834
filled diamonds are lower profile segment projections as described in Figure 7. The elevations of
835
the modern outlets are marked by crosses, which are very similar in elevation to the lower
836
segment projections for the rivers north of the effectively closed Tamarugal Basin. B) Relief
837
between modern river outlets and the elevation of the projected river segment outlets for 21
838
streams that have well defined middle elevation knickpoint bounded segments. Closed basin and
839
corrected open basin relief values have an average of 1080± 230 m, which is interpreted here to
840
represent tectonic uplift.
841
842
Figure 10: Relationship between the 5.4 Ma Carcote ignimbrite, the modeled middle river
843
segments and the actual river profile for sup04, tamb03 and arc01. The profiles, modeled and
844
actual, and the average elevation of mapped outcrops of Carcote ignimbrite were projected into
845
one straight profile line for each catchment. A) The sup04 river profile shows the Carcote
846
ignimbrite is at similar elevations to the knickpoint-bounded middle segment and its modeled
847
projection until ~33 km along the profile where it drops in elevation to a position between the
848
modern profile and the modeled profile. B and C). The tamb03 and arc01 modern and projected
849
river profiles and the average elevations of the mapped exposures of Carcote ignimbrite. Less
850
extensive deposition or exposure of the ignimbrite makes the interpretation more difficult, but
851
the initial modeled segments and the mapped distribution of the Carcote correspond well with
852
one another.
853
27
853
Supplementary data:
854
1. Geochronology
855
We report three radiometric age determinations on volcanic horizons found in the upper
856
Miocene-Pliocene sedimentary sequences at the foot of the western mountain front and in the
857
Coastal Cordillera (Tables 1 and 2).
858
Sample GH-03-02 is from a volcanic ash horizon interbedded in fine grained deposits
859
preserved on top of a strath terrace ~ 5 km from Quebrada de Tiliviche’s outlet to the Pacific
860
Ocean (Figure 7 in paper). Zircons separated from the tuff are clear, elongate and doubly
861
terminated prisms ranging 150-200 microns in length and 20-30 microns in width. All analyses
862
were based on single zircon grains that were carefully hand-selected and treated using the
863
chemical abrasion technique (CA-TIMS) of Mattinson (2005) prior to complete dissolution and
864
analysis by the U-Pb isotope dilution method. A full description of U-Pb analytical protocols at
865
MIT is reported in Schoene et al. (2006). Accordingly, zircons were annealed inside a furnace at
866
900°C for 60 hours and subsequently subjected to partial dissolution by concentrated HF for 12
867
hours. The latter step was carried out at 180°C for analyses z1 to z5 and at 210°C for analyses z6
868
to z9 in order to ensure effective removal of high-U zones susceptible to Pb loss.
869
U-Pb zircon geochronology of zircons younger than 10 Ma can be challenging due to
870
small amounts of radiogenic Pb in the zircon and large uncertainties associated with the
871
207
872
common Pb (normally 0.3 to 0.4 picograms) is assumed to originate from laboratory blank. The
873
data exhibit a spread along the concordia curve which we interpret to reflect a combination of
874
subtle inheritance from slightly older zircon, as well as possible magmatic residence time.
875
Because of the slightly greater ionic radius of Th relative to U, the intermediate daughter 230Th is
876
normally preferentially excluded from the octahedral site in zircon, resulting in 230Th, and thus
877
206
878
dramatic for zircon younger than 50 Ma. Disequilibrium between minerals and magma may be
879
quantified and 230Th deficiencies corrected with a standard methodology utilizing 232Th/238U
880
ratios (cf. Schaerer 1984). Evaluation of the disequilibrium partitioning is mainly limited by our
881
knowledge of the Th/U of the magma coexisting with the crystal during growth which for most
882
zircon-bearing eruptive systems is quite limited (e.g., Schmitz and Bowring, 2001). We have
883
corrected the 206Pb/238U dates assuming a magmatic Th/U ratio of 4 and zircon Th/U ratios which
Pb/206Pb dates. In this study, radiogenic Pb concentrations vary from 1 to 2 picograms and all
Pb, deficiencies that give rise to anomalously young 206Pb/238U dates. This effect is most
28
884
885
are estimated from the 208Pb/206Pb of the analyzed zircon.
We interpret the weighted mean 206Pb/238U date of the three youngest analyses, that is
886
6.433 ± 0.047 Ma (MSWD=0.34) as the best estimate of the eruption age of this rock. The
887
reported 2 error includes analytical uncertainties, as well as systematic errors associated
888
with tracer calibration and U decay constants. Given the young age of the zircons and the
889
aggressive treatment by CA-TIMS, we are confident that the youngest age cluster does not
890
represent Pb-loss from an older population.
891
Sample IB-7 is an ignimbritic ash intercalated in the upper member of the Altos de Pica
892
Formation in the Quebrada Guatacondo valley. This sample yielded an 40Ar/39Ar plateau of 13.03
893
± 0.17 Ma on biotite (figure 2, table 2) and thus indicates that the top surface of the Altos de Pica
894
Formation must be younger than 13 Ma.
895
Sample IB-9 is an ignimbrite intercalated in gravel deposits that overlie and are slightly
896
discordant with the upper member of the Altos de Pica Formation in Quebrada Guatacondo. The
897
ignimbrite was identified tentatively in the field to be the Carcote Ignimbrite, known elsewhere
898
to be ca. 5.8 Ma (K-Ar) (Tomlinson et al., 2001). This sample has a well-defined 40Ar/39Ar
899
plateau age of 5.38 ±0.09 Ma on biotite and a concordant inverse isochron age of 5.34 ±0.14 Ma
900
(figure 3; table 2). While it is younger than the previously published age of the Carcote
901
Ignimbrite, because of the higher precision of the40Ar/39Ar dating method, it is likely that this
902
sample has a more accurate age determination for the ignimbrite than the K-Ar data. These IB-9
903
data indicate that the top surface of the Altos de Pica Formation must be older than 5.3±0.14 Ma.
904
905
Figure 1: U-Pb concordia diagram for single zircon analyses from sample GH03-02. See Table
906
5.2 for detailed analytical results. The location of the sample is shown in figure 7. Analyses were
907
preformed at the Massachusetts Institute of Technology U-Pb geochronology laboratory.
908
909
Figure 2: 40Ar/39Ar age spectra for step-heating experiments on biotites separated from volcanic
910
ash samples IB-7 (a) and IB-9 (b). See Table 3 for details pertaining to the analyses. Analyses
911
were performed at the Ar geochronology lab of the Servicio Nacional de Geología y Minería,
912
Santiago, Chile.
913
914
References
29
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
MATTINSON, J.M., 2005, Zircon U/Pb chemical abrasion (CA-TIMS) method; combined
annealing and multi-step partial dissolution analysis for improved precision and accuracy of
zircon ages: Chemical Geology, v. 220, p. 47-66.
SCHÄRER, U., 1984, The effect of initial 230Th disequilibrium on young U-Pb ages; the
Makalu case, Himalaya: Earth and Planetary Science Letters, v. 67, p. 191-204.
SCHMITZ, M.D., and BOWRING, S.A., 2001, U-Pb zircon and titanite systematics of the Fish
Canyon Tuff; an assessment of high-precision U-Pb geochronology and its application to young
volcanic rocks: Geochimica et Cosmochimica Acta, v. 65, p. 2571-2587.
SCHOENE, B., CROWLEY, J.L., CONDON, D.J., SCHMITZ, M.D., and BOWRING, S.A.,
2006, Reassessing the uranium decay constants for geochronology using ID-TIMS U-Pb data:
Geochimica et Cosmochimica Acta, v. 70, p. 426-445.
930
931
932
2. Modeled stream segments
All stream profiles with outlet projections shown in Figure 9 are shown in the following set
933
of figures. The figures include a shaded relief map of each drainage basin, longitudinal profiles
934
of all the streams examined in each basin and the longitudinal profiles and modeled segments of
935
the streams in Figure 9. The basin maps and profiles are in order from north to south with the
936
outlet locations given in table 1. On each map knickpoints are shown by open triangles, the
937
stream network for the basin is shown in colors that reflect increasing drainage area. Individual
938
streams that were examined within a basin are color coded by number (blue = 1; green = 2, red =
939
3, cyan = 4, magenta = 5, yellow = 6 and black = 7). The breaks in the color-coded lines
940
separate the parts of each stream for which and Ksn values were derived. Not all basins or
941
parameter fit profile segments were used to create downstream profile projections. The same
942
color scheme is used for the grouped basin wide long profiles, mapped knickpoints and steep
943
segments are shown as ‘+’ symbols. For each modeled segment, we show a profile of the stream,
944
the downstream projection of the modeled segment, the area accumulation with distance, and the
945
slope-area plot from which the and Ksn values were derived.
946
Table 1. Projected stream outlet elevations and base levels
stream
outlet
segment concavity
ksn
θ =-0.4
segment
latitude
type
(θ)
aza003
-18.47
m
-0.70
53.0
aza006
-18.47
m
-0.58
74.0
vit201-1
-18.76
u
0.42
31.1
vit201-2
-18.76
l
2.10
111.0
vit202
-18.76
u
0.58
57.7
cam001
-19.19
u
0.39
46.9
cam003
-19.19
u
0.54
42.5
suc001-1
-19.19
u
0.25
53.2
suc001-2
-19.19
m
2.50
94.6
suc003
-19.19
m
1.16
72.0
tili01-1
-19.55
m
10.30
212.0
tili01-2
-19.55
u
21.00
184.0
tili01-3
-19.55
l
5.10
77.2
tara02-1
-20.09
m
13.80
164.0
tara02-2
-20.09
m
3.60
192.0
tara03
-20.09
m
8.40
133.0
tara04
-20.09
m
0.66
44.0
quip03
-20.25
u
0.37
53.8
jmor01
-20.52
m
1.00
43.3
jmor04
-20.52
m
0.45
89.4
chac01
-20.80
m
1.00
56.3
chac07
-20.80
u
0.83
16.2
guat01
-20.99
m
1.20
87.6
mani01
-21.10
m
0.41
39.5
sup02a-1
-21.23
m
1.60
52.0
sup02a-2
-21.23
m
0.68
44.0
sup04
-21.23
m
0.76
64.5
viej01
-21.31
m
0.99
35.2
tamb01
-21.43
u
0.79
45.3
tamb02
-21.43
u
2.30
68.7
tamb03
-21.43
m
0.15
26.4
tamb04
-21.43
u
0.57
68.8
tamb04-2
-21.43
m
0.38
34.9
arc01
-21.70
m
0.65
24.9
r2
0.84
0.90
0.79
0.82
0.77
0.81
0.82
0.75
0.75
0.71
0.79
0.75
0.83
0.76
0.47
0.78
0.53
0.86
0.92
0.72
0.70
0.82
0.70
0.78
0.60
0.82
0.82
0.90
0.98
0.60
0.56
0.49
0.85
0.94
outlet
error base level
elevation (m) (m)
(m)
2262
130
1000
2011
188
1000
3645
65
1211
1211
101
1211
1891
320
1211
1369
264
1000
2798
194
1000
-10
383
0
1931
104
1000
2062
119
1000
1987
111
1067
3158
48
1067
1067
110
1067
2194
76
1219
2216
280
1219
2574
72
1219
2277
276
1219
2231
209
1337
2414
123
1263
1778
378
1263
2465
101
1353
3672
32
1353
2755
165
1295
2392
121
1309
2566
149
1376
2337
106
1376
2854
109
1376
3191
63
1900
3251
72
1307
3206
98
1307
2105
289
1307
1864
267
1307
2157
97
1307
2759
70
1661
70°W
69°W
70°W
65°W
Peru
15°S
Bolivia
20°S
W
18°S
e
C
Pacific h
i
Ocean l
e
Rio Lluta
25°S
s
t
n
e r
s t
W e
Arica
Rio Azapa
A
19°S
l
n
t
i
C
o
M
l
a i n
n t
o u
Tarapacá
r
Pamp
Aroma
p
Q. Tiliviche
a
d
Q. Quipisca
20°S
n
i
Juan de Morales
o
l
Quisma
Pica
l
Chacarilla
Ramada
e
r
arugal
l Ta m
a de
Chipana
F r o n t
21°S
Tambillo
oa
oL
r a
Ri
Guatacondo
Maní
Supica
Viejo
Cerillos
a
l e
C o r d i l
Iquique
21°S
r
Suca-Camarones
20°S
30°S
e
t a l
a s
C o
Q. Vitor
19°S
Argentina
Arcas
elevation (m)
22°S
6600
22°S
0
Kilometers
0
70°W
20 40
69°W
Hoke et al. Figure 1
80
69°10'W
69°0'W
68°50'W
AdP2
AdP2
20°40'S
20°40'S
00
40
Q . Ch acar
illa
45
00
00
40
3500
3000
2500
2000
1500
20°50'S
45
00
20°50'S
3500
do
4000
at
Q. Gu
aco
n
21°S
21°S
0
2
4
8
0
00
3
Kilometers
69°10'W
69°W
68°50'W
Q. Aluvium
M.-Plio. sediments
Altos de Pica 2
pre Olig. sediments
Q Eolian
Altos de Pica 5
Altos de Pica 1-3
pC-Ord. metamorphic
P-Pl slides
Altos de Pica 4
pre- Olig. igneous
Hoke et al. Figure 2
Time (Ma)
S
Arcas Fan
deposition
begins
Ash
Horizon
13 Ma
Northern Chile
0
22°S - 19.75°S
5
Carcote ign.
10
15
20
25
N
19.75°S - 18°S
Lauca-Peréz ing.
diment
El Diablo/Altos de Pica pe
5
Altos
de 4
3
Pica
2
1
??
El Diablo Fm
Oxaya Fm
Azapa Fm
Hoke et al. Figure 4
Q. Tiliviche
downcutting
~6.4 Ma
8-9 Ma Tana Lava
Andesite
clasts 12 Ma
69°W
3000
3500
21°S
21°S
00
45
Mani
2000
1500
Supica
0
400
Viejo
Cerillos
21°30'S
21°30’S
Tambillo
2500
5
10
20
4000
0
Kilometers
69°W
Hoke et al. Figure 4
3500
Arcas
70°W
19°S
19°S
70°W
Suca - Camarones drainage basin topography
6000
elevation (m)
5000
4000
3000
2000
1000
0
0
20
40
basin max
basin min
suca 1
suca 1k
60
80
distance from mouth (km)
suca 2
suca 2k
suca 3
suca 3k
100
suca 4
suca 4k
cam 1
cam 1k
Hoke et al. Figure 5a
120
cam 2
cam 2k
cam 3
cam 3k
140
69°W
20°S
20°S
69°W
Tarapaca drainage basin topography
6000
5500
elevation (m)
5000
4500
4000
3500
3000
2500
2000
1500
1000
0
10
20
30
40
50
60
distance from mouth (km)
basin max
basin min
tara 1
tara 1k
tara 2
tara 2k
tara 3
tara 3k
Hoke et al. Figure 5a
70
tara 4
tara 4k
80
90
100
tara03
elevation (m)
5000
A
Upper (Altiplano) segment
modeled profile
actual profile
4000
middle segment
3000
θ = 8.39 ± 2.42
ksn =144
2000
1000
120
lower segment
100
80
60
40
Distance from mouth (km)
20
0
100
80
60
40
Distance from mouth (km)
20
0
cumulative drainage area (m2)
10
10
B
5
10
120
10
0
θ=8.4 ± 2.4 ksn =144
gradient
C
10
10
10
−1
−2
θref = −0.4
−3
10
5
10
6
10
7
10
8
2
drainage area (m )
Hoke et al. Figure 7
10
9
10
10
6000
5500
Tarapaca Basin maximum topography
simple ramp uplift
5000
sigmoidal uplift
Relative uplift
elevation (m)
4500
4000
3500
3000
2500
2000
1500
1000
0
10
20
30
40
50
60
Distance (km)
Hoke et al. Figure 8
70
80
90
100
0
4000
3500
closed basin
open basin
2500
2000
1500
1000
upper segments
middle segments
500
lower segments
modern outlet elevation
18
18.5
19
19.5
20
20.5
21
21.5
Elevation (m)
3000
22
0
Latitude (°S)
B.
2000
1800
closed basin
1400
1200
1080 ± 230 m
1000
800
600
400
200
18
18.5
19
19.5
20
20.5
Latitude °S
Hoke et al. figure 9
21
21.5
22
0
Relief (m)
1600
open basin
Sipuca profile 4 and Carcote
4500
A
elevation (m)
4000
3500
approx. 5.4 Ma
knickpoint location
pre- 5.4 Ma
incision
post- 5.4 Ma
incision
3000
2500
30
2000
35
projected profile
avg. Carcote elevation
actual profile
1500
1000
0
5
10
15
20
25
distance (km)
30
35
40
45
Tambillo profile 3 and Carcote ignimbrite
4500
B
4000
3500
elevation (m)
3000
2500
2000
projected profile
actual profile
avg. Carcote elevation
1500
1000
0
5
10
15
20
25
30
distance (km)
35
40
45
50
Quebrada Arcas profile 1 and mapped Carcote ignimbrite
4000
C
elevation (m)
3500
3000
2500
actual profile
projected profile
avg. Carcote elevation
2000
1500
0
5
10
15
20
25
distance (km)
Hoke et al. figure 10
30
35
40
45