T33D-2086
Numerical modeling of the late Cenozoic geomorphic evolution of Grand Canyon, Arizona
The late Cenozoic geomorphic evolution of Grand Canyon has been influenced by three primary tectonic and drainage adjustment events. First, incision into the Paleozoic strata of the southwestern margin of the Colorado Plateau began at 16 Ma in response to relief production along the Grand Wash Fault. Second, the ancestral Upper Colorado River reversed drainage and became integrated with the Lower Colorado River basin through Grand Canyon between 5.5 and 6 Ma. Third, the Colorado River was influenced by Plio- Quaternary normal faulting along the Hurricane and Toroweap Faults. Despite the relatively firm constraints available on the timing of these events, the geomorphic evolution of Grand Canyon is still not well constrained and many questions remain. For example, was there a deeply-incised gorge in western Grand Canyon before Colorado River integration? How and where was the Colorado River integrated? How have incision rates varied in space and time? In this paper, I describe the results of a numerical modeling study designed to address these questions. The model integrates the stream power model for bedrock channel erosion with cliff retreat and the flexural-isostatic response to erosion. The model honors the structural geology of the Grand Canyon region, including the variable erodibility of rocks in the Colorado Plateau and the occurrence of Plio-Quaternary normal faulting along the Hurricane-Toroweap Fault system. We present the results of two models designed to bracket the possible drainage architectures of the southwestern margin of the Colorado Plateau in Miocene time. In the first model, we assume a 13,000 km2 drainage basin primarily sourced from the Hualapai and Coconino Plateaux. The results of this model indicate that relief production along the Grand Wash fault initiated the formation of a large (700 m) knickpoint that migrated headward at a rate of 15 km/Myr prior to drainage integration at 6 Ma to form a deep gorge in western Grand Canyon. This model also illustrates that integration of the Colorado River increased the rate of knickpoint migration to 60 km/Myr, resulting in rapid incision of eastern Grand, Marble, and Glen Canyons down to the level of the Redwall Limestone from 6-4 Ma. Widening of Grand Canyon by cliff retreat triggered flexural- isostatic rebound and renewed river incision of up to 400 m in Plio-Quaternary time. Plio-Quaternary normal faulting significantly dampened incision rates in western Grand Canyon relative to eastern Grand Canyon. As an alternative, we also consider the results of a model in which no incision in western Grand Canyon is assumed prior to 6 Ma. In that model, headward erosion prior to 6 Ma was not significant (by assumption), but the remaining results of the model are similar to that of the first model for the post-6 Ma period, illustrating the robustness of the post-integration behavior of the model with respect to pre-integration drainage scenarios. The results of the first model illustrate that headward erosion could be sufficient to capture the ancestral Upper Colorado River east of the Shiwitz Plateau, but the limited volume of Miocene clastic debris in the Grand Wash Trough and adjacent basins requires that slow rates of cliff widening and/or significant sediment storage in western Grand Canyon be invoked in order for this model to be consistent with the stratigraphic record.
T33D-2087
What Can Modern River Profiles Tell Us about Orogenic Processes and Orogen Evolution?
Numerous lines of evidence from theory, numerical simulations, and physical experiments suggest that orogen evolution is strongly coupled to atmospheric processes through the interrelationships among climate, topography, and erosion rate. In terms of orogenic processes and orogen evolution, these relationships are most important at the regional scale (mean topographic gradient, mean relief above surrounding plains) largely because crustal deformation is most sensitive to erosional unloading averaged over sufficiently long wavelengths. For this reason, and because above moderate erosion rates (> 0.2 mm/yr) hillslope form becomes decoupled from erosion rate, attention has focused on the river network, and even on particularly large rivers. We now have data that demonstrates a monotonic relationship between erosion rate and the channel steepness index (slope normalized for differences in drainage area) in a variety of field settings. Consequently, study of modern river profiles can yield useful information on recent and on-going patterns of rock uplift. It is not yet possible, however, to quantitatively isolate expected climatic and lithologic influences on this relationship. A combination of field studies and theoretical analyses are beginning to reveal the timescale of landscape response, and thus the topographic memory of past conditions. At orogen scale, river profile response to a change in rock uplift rate is on the order of 1-10 Myr. Because of these long response times, the modern profiles of large rivers and their major tributaries can potentially preserve an interpretable record of rock uplift rates since the Miocene and are insensitive to short-term climatic fluctuations. Only significant increases in rock uplift rate, however, are likely to leave a clear topographic signature. Strategies have been developed to differentiate between temporal and spatial (tectonic, climatic, or lithologic) influences on channel profile form, especially where spatially distributed data on recent incision rates is available. A more difficult question is one of cause and effect. Only in some circumstances is it possible to determine whether rivers are steep in response to localized rock uplift or whether localized rock uplift occurs in response to rapidly incising steep rivers.
T33D-2088
Universality and variability in the 2D form of drainage basins: a study of active orogens and passive margins
Identifying the controls on basin morphology is needed to understand how landscapes evolve under changing conditions. Some important morphological and topological relationships such as Hack's and Horton/Tokunagas" laws suggest that river basins, irrespective of the climatic, tectonic and/or biogeographic conditions under which they developed, are likely to have statistically predictable planform morphologies. Indeed empirical studies of the spacing between basins that drain the main divide of some active and inactive orogens suggest that this may be the case. In such situations major basins typically have an outlet spacing that is approximately half of the distance between the outlet and the main divide. However, notable deviations from this outlet spacing relationship have also been documented. We therefore address the question: what controls the planform morphology or "packing arrangement" of river basins (of all scales) on the flanks of orogens? To answer this question we conducted a study of basin morphology from both active mountain belts (e.g. Himalaya and Andes) and from morphological similar, but tectonically very different, passive margins (e.g. southeast Africa). We specifically investigate the relationship between basin outlet spacing and other two dimensional geometric properties such as basin length, main valley length and basin area. We show that if the rate of relative displacement between the main divide and outlet is slower than the rate of river erosion then, river basins, irrespective of size, tend towards an approximately diamond shaped packing arrangement in which the basin width is approximately half that of basin length. However, where relative displacement between the main divide and outlet has exceeds the rate of river erosion, such as in the western Himalaya, rivers are forced into an alternative, and possibly unstable packing arrangement.
T33D-2089
Landscape stability and instability in an experimental mountain
We have designed a series of physical experiments to explore the interrelationships between drainage network organization and topographic development in high-relief landscapes. Our experiments reveal striking internal variability in catchment-scale sediment efflux and drainage network evolution under constant rainfall and rates of base-level fall. We capture this internal dynamic using a unique set of measurement systems that allow us to relate the 3D evolution of topography to sediment flux from the model-orogen. We also observed remarkable stability off the main catchment drainage divides, suggesting that catchment structure is more robust than recent studies propose. We infer that a transition in the mix of dominant processes at the scale of a catchment generates the observed scale dependant autogenic dynamics. Our experimental apparatus is an erosion box in which two opposing panels slide downwards, so simulating base-level fall across emerging topography. Rainfall is generated by an ultra-fine misting apparatus.
T33D-2090
A Generalized Kolmogorov-Smirnov Statistic for Detrital Zircon Analysis of Modern Rivers
The Kolmogorov-Smirnov (K-S) statistic is widely used to test the null hypothesis (i.e., are two distributions drawn from the same population?). In detrital zircon provenance analysis of river systems, it is useful to have an equivalent statistical measure for comparisons involving composite samples that collectively represent regionally extensive river systems. We present a generalized K-S statistic that depends on the proportional contribution and sample size of individual distributions as well as the correlation between the respective individual populations. Analytical expressions for the significance function were obtained for end member cases in which the individual populations are either completely independent or identical. Although intermediate cases must still be tested by numerical analysis, we provide examples involving the modern Colorado River (Southwest North America) and Yalu- Brahmaputra (Southeast Asia) river systems that demonstrate how useful statistical comparisons may still be obtained using the end member analytical expressions.
T33D-2091
Gangdese Arc Detritus in the Eastern Himalayan Foreland Basin: Implications for the Evolution of the Yalu-Brahmaputra River System.
The tectonic evolution of the Himalayan orogen is closely intertwined with the development of its major river systems. The Yalu-Brahmaputra River system is one of the most interesting, yet poorly understood of the Himalayan rivers. An important clue to unraveling the history of the Yalu-Brahmaputra River is given by the presence of extraregional sediment eroded from the Gangdese batholith within the eastern Himalayan foreland. We have used U-Pb and Lu-Hf analyses of Neogene foreland sediments to assess the spatial and temporal extent of material derived from the Gangdese batholith within the foreland. Our results show that deposition of this material occurred throughout late Miocene and Pliocene time and was limited to the Arunachal Himalaya (92 E to 95 E). Additional study of modern sand from the Yalu-Brahmaputra River and its major tributaries indicate that the extraregional sediment was sourced from north of the Himalaya and was most likely deposited by a south-flowing tranverse river. Our preferred model to explain the data invokes capture of the Yalu River by the Subansiri River at approximately 10 Ma, followed by capture of the Yalu River by the Siang River at 3-4 Ma.
T33D-2092
Cosmogenically-derived Erosion Rates From the Eastern Tibetan Plateau Between Lhasa, Tibet and the Namche Barwa Massif, Eastern Tibet
In this study, we examine the processes driving the development of recent relief along the southeastern edge of the Tibetan Plateau within tributaries to the Yarlung Tsangpo. We determined erosion rates by measuring cosmogenic 10-Beryllium in river sands collected at point bars to infer basin averaged erosion rates of twenty-seven tributaries situated along two east-west transects across the northern and southern forks of the Yarlung between Lhasa, Tibet, and the Namche Barwa massif to determine the relationship between erosion rates, relief, the tectonics of the Himalayan eastern syntaxis, and climate. In the north, the highest erosion rates occur along the middle of the transect, with erosion rates decreasing towards the headwaters of the Nyang River and its confluence with the Yarlung. These high erosion rates correlate with a transition from low topographic relief in the west to high relief in the east, and lie at the upstream most extent of lake deposits observed along the river. This trend appears to reflect the migration a knickpoint through the Nyang River, with high erosion rates at the knickpoint's current location and low erosion rates upstream and downstream of the knickpoint, which may have been triggered by breaching of an ice dam at the Yarlung river gorge or the northward projection of E-W extensional deformation in the area. Interestingly, large increases in recent erosion rates are not observed upstream of the Nyang River's confluence with the Yarlung, suggesting that the enhanced erosion that we expect to be associated with doming of the Namche Barwa syntaxis does not reach significantly beyond the high-relief massif. Future work will involve quantifying rainfall across the region and correlating it with the distribution of erosion rates to discern the relationship between precipitation, relief, erosion, and the river incision history of this region.
T33D-2093
River Mediated Development of an Active Extensional Culmination: the Yulong Mountains, Southwestern China
The Yulong Xueshan (Jade Dragon Snow Mountains) are an isolated range of anomalously high and steep mountains bisected by the Jinsha (Yangzi) river in Western Yunnan province, in the eastern Tibetan borderlands. The range is defined and bounded by a closed network of active transtensional faults, whose most recent major rupture was a M 7.0 quake in 1997. Some of the deepest regional structural and stratigraphic levels are exposed in a domal culmination in the footwalls of these faults and at the bottom of the nearly 4km deep gorge cut by the Jinsha river. Within the gorge, the Jinsha river becomes extremely narrow and steep, suggesting adjustment to anomalously high rates of surface uplift. Geomorphic arguments suggest erosion rates of between 1 to 5 mm/yr. These rates are consistent with independently derived estimates of ~3 mm/yr based on stratigraphic and structural evidence. We argue that the anomalous exhumation and uplift rates we infer for the Yulong mountains are the result of the interaction of (1) vigorous river incision controlled erosional processes that balance rock uplift rates; (2) a closed network of normal faults that accommodate differential rock uplift rates; (3) weak middle to lower crust that flows in response to imposed surface loads. Localized uplift probably began in response to unloading along the range bounding faults, but high erosion rates are responsible for sustaining the anomalous, localized rock uplift that continues to the present. Unlike other described examples where localized exhumation along a major river is associated with a river capture event, in the Yulong mountains the closed network of normal faults permit the range to respond independently of its surroundings (and therefore have effectively no flexural strength) and accommodates localized uplift. A final key characteristic of this system is that it is superimposed upon a landscape that was already characterized by the significant incision and entrenchment of the major river networks on a rising landscape. Without this pre-existing entrenchment, a river could simply be diverted around a localized uplift, rather than being "forced" to adjust and match the higher uplift rates.
T33D-2094
The Paleo-Lena River – 200 My of Transcontinental Zircon Transport in Siberia
An immense wedge of Carboniferous to Jurassic siliciclastic deltaic and marine strata accumulated on the Verkhoyansk passive margin of the Siberian craton. U-Pb ages of detrital zircons from Pennsylvanian to Middle Jurassic sandstones are remarkably consistent and show a systematic change in the proportion of zircon age populations through the stratigraphic section that reflects the evolution of the ancient drainage system. Most zircons originated from the orogenic belts along the southern margin of Siberia and were transported thousands of kilometers to the Verkhoyansk margin by a major transcontinental river system, the paleo-Lena, that existed for about 200 m.y. Specific sources for zircons are the Angara-Vitim batholith of Transbaikalia (315 and 291-288 Ma age peaks), plutons of the Altay-Sayan region of the Central Asia foldbelt (494-482 Ma), Proterozoic granitoids of northern Transbaikalia and the East Sayan Range (1888-1832 Ma), and minor contributions from the Siberian Platform and Aldan Shield (2900–2300 Ma). The long sedimentary transport documented for the paleo-Lena River system was comparable to Permian and Jurassic transcontinental rivers that transported sand from the Appalachians to the Colorado Plateau region (Dickinson and Gehrels, 2003). This contrasts with sands of the modern Amazon where the population of far-traveled zircon decreases with increasing distance from the headwaters and eventually becomes overwhelmed by more proximally derived zircon (Rino et al., 2004; Mapes et al., 2005). The paleo-Lena met its demise when the Verkhoyansk margin was deformed in the Late Jurassic and Cretaceous, and sediment was diverted north to the Arctic Ocean. Thus, the life span of major transcontinental drainage systems can be comparable to that of the plate boundaries that surround them.
T33D-2095
Rates, Pattern and Timing of Tectonic Uplift in the Bolivian Andes from River Profile Analysis.
We examine the profiles of a number of tributaries in the upper Beni River watershed, Bolivia, with the goal of constraining the timing and rates of rock uplift. The tributaries drain across the Eastern Cordillera of the Bolivian Andes, and vary in drainage area from about 2 x 107 to 5 x 109 m2. Our study area is focused on a region of very high local relief, where the channels exhibit a number of interesting characteristics. We hypothesize that there was a period of high uplift rates across the region of high relief, and we analyze patterns in the upper Beni tributary profiles and profiles produced using a numerical model to constrain the timing and pattern of uplift rates in this region. In the Beni tributaries, the average upstream channel steepness index (channel slope normalized by drainage area) follows a distinct trend with tributary drainage area. While many of the mid-size tributaries exhibit distinct transient signals, such as large knickpoints and regions of anomalously high channel steepness index, many of the smaller streams at the front of the high relief region are relatively well adjusted and have smaller average channel steepness indices. The apparent graded profiles of these smaller tributaries may indicate that uplift rates were smaller at the front of the high relief region, or possibly these channels have been forced with high uplift rates for a long enough period that they are now fully adjusted. Furthermore, some of the largest streams, which drain off the Altiplano, also appear to be more adjusted than the mid-sized streams. One explanation for the apparent adjustment of the largest channels is that they gained drainage area during their evolution. An influx of water and sediment could have increased erosion rates and accelerated the pace of river profile adjustment. We use numerical model simulations to explore whether the observed channel profile patterns result from a distinct spatial or temporal uplift pattern. We find that a spatially variable pattern of uplift can explain the behavior of both the small and mid-sized drainages, but the profile patterns in the largest simulated channels do not match those observed in largest Beni tributaries. This difference could be due to drainage capture, which is not captured in the numerical simulations.
T33D-2096
Long-term Canyon Incision From Low-Temperature Thermochronology
River incision is one of the most dramatic landscape responses to the growth of orogenic plateaus. However, using incision as a proxy for surface uplift is fraught with problems. Few methods exist that record long-term incision directly, and it is often difficult to quantify the lag time that may exist between surface uplift and the onset of incision, particularly in arid regions where knickpoints may propagate slowly through a river system, or where major changes in climate have occurred. Approaches that can quantify long-term incision in multiple locations along a river offer the best potential for giving meaningful data that can be corrected for lag time. Low-temperature thermochronometers can record transient thermal adjustment induced by incision in areas where incision exceeds the initial depth of the closure temperature. We present apatite (U-Th)/He data from 31 samples collected in Cotahuasi-Ocoña Canyon, Peru, which cuts to depths of over 3 km through the western margin of the Altiplano plateau. Interpretation of the canyon incision history from cooling ages is complicated by a southwest to northeast increase in temperatures at the base of the crust due to active subduction, as well as by three-dimensional variations in sample cooling histories that depend on the style of landscape evolution. We address these complications with coupled finite-element thermal, landscape evolution, and thermochronometer age-prediction models to quantify the range of topographic evolution scenarios consistent with observed cooling ages. Geological evidence for early canyon depths of at least 200 m and comparison of 210 model simulations to observed cooling ages and regional heat flow determinations identify a best-fit history with 0.2 to 0.5 km of incision starting prior to ~14 Ma and 2.7 to 3.0 km of incision starting between 8 and 11 Ma. Young thermochronometer ages and a 40Ar/39Ar age on a valley- filling volcanic flow imply that incision ended between ca. 5 and 2.21 ± 0.02 Ma. Initial results from 4He/3He dating of a subset of the samples suggest that even higher precision in the cooling history and interpreted canyon incision can be derived with this method. Furthermore, the potential for the entire incision-induced cooling signal to be recorded in a single sample implies that the timing of incision can be deduced for multiple locations along the valley bottom, allowing for tests of knickpoint propagation through the system.
T33D-2097
Paleogeography, Paleo-drainage Systems, and Tectonic Reconstructions of Eocene Northern South America Constrained by U-Pb Detrital Zircon Geochronology
Thick, Eocene to Miocene clastic sedimentary basins are widespread across on- and offshore northern South America and have been identified using seismic reflection data in offshore basins of the Leeward Antilles, the Lesser Antilles arc and forearc, and the Barbados accretionary prism. Several 3 to12-km-thick Paleogene depocenters occur in shelf to deep basinal settings along the offshore margins of Venezuela, Trinidad and Tobago, and Barbados. Previous studies proposed that the proto-Orinoco River has been the single fluvial source for these distal, continentally-derived sandstone units along northern Venezuela as part of the early Eocene to Miocene, proto-Maracaibo fluvial-deltaic system that emanated from the northern Andes of western Venezuela and Colombia. Those distal sandstones were displaced eastward with the movement of the Caribbean plate by several hundred kilometers and are now found in basins and islands of the southeastern Caribbean region. We collected nine Eocene age sandstone samples from well cores and outcrops along the northern South America margin, including Lake Maracaibo, Trinidad and Tobago, and Barbados Island. In total, 945 single detrital zircon grains were analyzed using LA-ICP-MS. The objective is to reconstruct the paleogeography, paleo-drainage system, and tectonic history during Eocene time. New data show that the Eocene Misoa Formation of Lake Maracaibo was characterized by a mixture of Precambrian, Paleozoic, and Mesozoic ages matching age provinces from eastern Cordillera and the Guayana Shield, which is consistent with previous proto-Orinoco River model flowing from the western Amazonian region of Colombia and Brazil through the Maracaibo basin into the area of western Falcon basin. However, coeval Eocene samples from Barbados and Trinidad show a much different age population dominated by Precambrian matching the eastern part of the Guyana shield to the south, which suggests that the western onland system and eastern offshore units belong to different systems. We postulate that a series of smaller, north-flowing drainages provided a line of sediment source dispersal of Eocene sandstone from the north central and eastern edge of the Guyana shield onto the Eocene passive margin that extended from central Venezuela to Trinidad instead of being tectonically transported to their present locations suggested by earlier studies.
T33D-2098
Evolution of fluvial drainage networks evolving in response to an emerging Cenozoic Panama Cordillera
We use correlations of Early Miocene to Pliocene fluvial stratigraphic, paleocurrent, and provenance data from across Panama to reconstruct the evolution of fluvial drainage networks evolving in response to the emerging Cenozoic Panama Cordillera. Late Miocene Panama is generally thought to have been comprised of a volcanic arc archipelago that by the end of the Miocene had merged into a single land mass. Yet the timing of, the exact location, and manner in which the Isthmus of Panama emerged and eventually closed is poorly constrained. We use this paleodrainage network data set coupled with preliminary constraints on the extent and timing of magmatic events and structural deformation to better constrain the paleogeographic evolution of the Isthmus of Panama within a Caribbean tectonic framework.
T33D-2099
Record of drainage rearrangement and erosion in a transpressive orogen: relative role of horizontal and vertical rock advection in drainage evolution
Along transpressive orogens, both range-transverse and range-parallel motions influence drainage network evolution. Range-parallel motions promote stretching of drainage networks, river lengthening or shortening, and sudden shortenings by river capture. Range-transverse motions induce river course shortening or lengthening, and generates stronger rock uplift. River incision patterns are influenced by rock uplift and waves of incision resulting from drainage rearrangement. Thus, under steady conditions of wrenching, drainages evolve by continued deformation and discrete rearrangements. Therefore, a significant part of erosion can be achieved in a state of significant departure from dynamic equilibrium. The frequency, intensity, and duration of these events set the timescale over which their integrated effects can be regarded as the expression of a long-term dynamic equilibrium. We document the growth of a 103-104 km2 catchment drained by the Chixoy River in Guatemala. The catchment covers a large part of a 50 km wide orogen located astride the North American - Caribbean plates boundary (Sierra de las Minas – Sierra de Chuacus range). The range is wrenched by sinistral tectonics with a varying amount of transpression and transtension. On the northern flank of this range, the Polochic Fault (PF) accumulated 130 km of total strike-slip displacement, but the Chixoy River only displays a 25 km tectonic bend. Geological evidence indicates that the river probably experienced a diversion that reset earlier tectonic bending. Upstream, the catchment stands out as a large (110x30 km) zone of enhanced erosion (2500 km3 removed since the Middle Miocene). The catchment retains many paleovalleys that we use as markers to track drainage rearrangement, bedrock deformation and changes in erosion rates. Study of the paleovalleys includes: satellite image detection, field mapping of river deposits, analyses of grain-size, clast provenance, heavy mineral provenance, deposit architecture, geochemical analyses, Ar40 -Ar39 dating of volcanic tuffs, 10 Be-26 Al burial dating, and apatite He cooling ages of the bedrock. Coupled analyses of erosion and drainage rearrangement show that, in the studied case, catchment growth occurred over 107 years. Most of the catchment erosion and growth is attributable to uplift along the PF rather than strike-slip motion, although both motion types contribute to the rearrangement. Growth of the catchment is strongly catalyzed by a wealth of other factors, such as river avulsion, volcanism, karstic flow, phreatic flow, and aridity resulting from catchment deepening.
T33D-2100
Constraining Paleotopography of the Sierra Nevada, California, Through U-Pb Detrital Zircon Analysis of Eocene River Deposits
The provenance of fluvial deposits found in paleochannels along the western flank of the central and northern Sierra Nevada can provide insight into the topography of adjacent uplands. We present U-Pb detrital zircon ages from the deposits, which allows us to trace the sourcing of paleorivers and to constrain the spatial and temporal development of a Sierran range front. The distribution of zircon ages in most samples is bimodal, with a dominant age peak between 110 and 90 Ma and a smaller, but significant peak occurring in the mid to late Jurassic. Our age distribution patterns match closely with apparent magmatic flux events in the central and northern Sierra Nevada, indicating that detrital zircons are of Sierran arc affinity. With the exception of one sample, which was collected in close proximity to the Devonian Lake Bowman batholith, fewer than 5% of all sample grain populations have pre-Mesozoic ages. This suggests that Eocene river systems were sourced exclusively in the Jurassic - Cretaceous Sierra Nevada batholith and therefore had higher gradients and drained smaller areas than is commonly thought. Furthermore, our data indicate that the early to mid Cenozoic Sierra Nevada functioned as a major topographic barrier and was characterized by a western flank with a slope similar to that of the modern one.
T33D-2101
Evolution of the Colorado River: Culmination of a Major Cenozoic Transformation of SW North American Drainage Patterns
The modern Colorado River is a large, youthful, unequilibrated continental drainage system that was established abruptly between 5 and 6 million years ago in conjunction with Gulf of California rifting and establishment of the modern river course through the western Grand Canyon and lower Colorado river region. New detrital zircon U-Pb analyses provide insight into details relating to the cause, timing and consequences of river inception. These samples encompass (1) the modern Colorado River delta, (2) major tributaries including the Green, "Grand", San Juan, Little Colorado and Gila rivers (3) late Miocene to Pliocene sediments along the lower Colorado (4) late Miocene to Pleistocene deltaic and fluvial sediments of the Imperial and Palm Spring Groups in the western Salton Trough, and (5) late Miocene- early Pliocene Bidahochi Formation of eastern Arizona. Data from the western Salton Trough and modern delta yield strata yield remarkably homogeneous age distributions that indicate there was little evolution in Colorado River sediment composition since 5.3 Ma. Detrital zircon is dominated by a mix of local southwest US cratonal basement (1.7 and 1.4 Ga) plus reworked supracrustal sequences of the Colorado Plateau that provide Neoproterozoic, 1.1 Ga, and early Paleozoic zircons. The strong homogeneity of the detrital zircon record from late Miocene to the present combined with the fact that the most easterly tributaries most strongly resemble the delta is consistent with the 'lake spillover model' for inception and integration of the modern Colorado River drainage. Abrupt integration of the lower Colorado River after 5.6 Ma is clearly recorded by detrital zircon ages from the laucustrine Bouse Formation and Bullhead alluvium aggradational package. Fluvial-laucustrine deposits of the Bidahochi Formation may represent a lake that overtopped the Kaibab upwarp to initiate western Grand Canyon incision.