1

TITLE: Tectonic and climatic influences on fluvial channel architecture in the Upper Cretaceous-Paleogene Raton basin, Colorado-New Mexico, USA

RUNNING TITLE:Fluvial channel architecture in the Raton basin

AUTHORS:Theresa M. Schwartz (corresponding author; tmschwartz@usgs.gov)Geosciences & Environmental Change Science Center, U.S. Geological Survey, Denver, CO 80225303-236-7881

Marieke Dechesne (mdechesne@usgs.gov)Geosciences & Environmental Change Science Center, U.S. Geological Survey, Denver, CO 80225303-236-1289

Kristine Zellman (kzellman@usgs.gov)Geosciences & Environmental Change Science Center, U.S. Geological Survey, Denver, CO 80225303-236-5838

KEYWORDS (5): fluvial fan; perennial river; ephemeral river; flashy discharge; Laramide basin

123456789

10111213141516171819202122

2

ABSTRACTThe Raton basin of Colorado-New Mexico, USA, is the southeasternmost basin of the Laramide

intraforeland basin network of North America. It hosts a thick succession (15,000 ft, or 4.5 km) of Upper Cretaceous to Paleogene marine and continental strata that were deposited in response to the final regression of the Western Interior Seaway and the onset of Laramide intraforeland deformation. The Upper Cretaceous-Paleogene Raton and Poison Canyon Formations have previously been described as meandering and braided river deposits that represent distal and proximal members of rivers that drained the basin-bounding Sangre de Cristo-Culebra Laramide uplift. We present new observations of fluvial channel architecture that suggest both formations contain the deposits of sinuous fluvial channels. Channel fills of the Raton Formation preserve perennial fluvial channels that likely had extrabasinal sources, whereas overlying channel fills of the Poison Canyon Formation preserve ephemeral, locally sourced fluvial channels, perhaps constituting a fluvial fan that emanated from the Sangre de Cristo-Culebra uplift. We suggest that the transition in fluvial style reflects both tectonic and climatic influences on the basin, including emergence of the Sangre de Cristo-Culebra uplift and changes in precipitation patterns and runoff due to climatic changes that occurred during the Paleogene.

INTRODUCTIONThe past few decades have witnessed renewed, focused interest in determining the geologic and

topographic history of the Laramide intraforeland province of the North American Cordillera. Studies on both local and regional scales have highlighted its paleoelevation history (e.g., Davis et al., 2009b; Chamberlain et al., 2012; Fan et al., 2014; Copeland et al., 2017), sediment routing within (e.g., Carroll et al., 2008; Davis et al., 2009a; Schwartz and Schwartz, 2013; Smith et al., 2017) and beyond (Blum and Pecha, 2014; Sharman et al., 2017) the Western Interior region, and the influence of paleoclimatic events on landscape evolution (Sewall and Sloan, 2006; Fan et al., 2014; Schwartz et al., 2019). As part of this, multiple recent studies have utilized Laramide basin fills to document local sedimentary responses to global climatic events (e.g., Foreman et al., 2012; Foreman, 2014; Plink-Bjorklund et al., 2014; Wang and Plink-Bjorklund, 2019; Zellman, 2019). Most of these studies have relied on data collected from relatively few, well-studied basins that lie in the interior of the Laramide province, such as the Bighorn, Wind River, Green River-Washakie, and Uinta-Piceance basins. In contrast, relatively few studies have focused on similar issues in basins on the eastern periphery of the Laramide province.

The Raton basin of Colorado-New Mexico, USA is the southeasternmost “perimeter” basin of the Laramide intraforeland province (Fig. 1; nomenclature after Dickinson et al., 1988; Lawton, 2008). Similar to many other Laramide basins, the Raton basin contains a kilometers-thick succession of Upper Cretaceous-Paleogene strata (Pillmore et al., 1984) that were deposited synchronous with the transition from Sevier- to Laramide-style deformation in the Western Interior, the concomitant building of topography, and significant changes in global climate. The Raton and Poison Canyon Formations of the Raton basin provide a nearly uninterrupted record of fluvial deposition that spans the Cretaceous-Paleogene (K/Pg) boundary and the early Paleogene. In spite of dense vegetation, roadside outcrops afford a detailed view into fluvial depositional processes through time. Within the context of paleotectonic, paleoelevation, and paleoclimate studies across the Western Interior region, these deposits record environmental conditions for a Laramide basin located distal to the Cordilleran orogen, on the eastern periphery of the Laramide intraforeland province.

We provide detailed descriptions of Upper Cretaceous (Maastrichtian) to Paleocene, and possibly Eocene, fluvial deposits of the Raton and Poison Canyon Formations. We document marked temporal

2324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

3

trends in fluvial channel architecture, overbank deposit character, channel dimensions, and sediment caliber that demonstrate a transition from deposition by perennially active sinuous channels that were laterally confined by coal-rich overbank deposits to deposition by ephemeral sinuous channels that were laterally mobile and relatively unconfined by fine-grained overbank deposits. These new interpretations and existing sediment provenance data highlight changes in paleodispersal patterns and fluvial processes that likely occurred in response to regional tectonic and global climatic events in the early Paleogene.

STUDY AREA & PREVIOUS WORKRaton basin

The Raton basin is the southernmost of three large “perimeter” Laramide basins located east of the Colorado Plateau, flanking the Laramide deformation front of Colorado and New Mexico (Fig. 1; Dickinson et al., 1988; Lawton, 2008). The basin is a north-trending, asymmetric syncline that is bound on the west by a series of high-angle, east-verging reverse faults that uplift Mesoproterozoic crystalline basement of the Sangre de Cristo Mountains and Culebra Range over Pennsylvanian-Permian and younger strata (Lindsey, 1998). The synformal axis of the basin (the La Veta Syncline) closely parallels the range front (Fig. 1). The sedimentary rocks on the west limb of the syncline dip eastward toward the basin axis, whereas sedimentary rocks on the east limb are largely sub-horizontal (Tyler et al., 1995; Topper et al., 2011; this study). The northern sector of the basin fill is intruded by Oligocene-Miocene stocks and radial dike swarms of the Spanish Peaks (Knopf, 1936; Muller, 1986; Penn and Lindsey, 2009).

Cretaceous-Paleogene stratigraphyWe use the subdivisions of Cather (2004) and separate the Laramide depositional history of the

Raton basin into three phases: (1) an early transitional phase represented by marine to coastal plain deposits of the Upper Cretaceous (Campanian-Maastrichtian) Pierre, Trinidad, and Vermejo Formations; (2) a middle phase represented by terrestrial deposits of the uppermost Cretaceous (Maastrichtian) to Paleocene(?) Raton and Poison Canyon Formations; and (3) a late phase represented by terrestrial deposits of the Eocene(?) Cuchara, Huerfano, and Farisita Formations (Fig. 1).

The Cretaceous Pierre Shale (up to 580 m or 1,900 ft thick), Trinidad Sandstone (30-40 m or 130 ft thick), and Vermejo Formations (up to 120 m or 395 ft thick) record the final regression of the Western Interior Seaway from the region (Pillmore et al., 1984; Flores, 1987). The three formations form a conformable, northeast-prograding succession of offshore and prodelta shale and siltstone, beach-delta sandstone, and heterolithic coastal plain deposits (Pillmore et al., 1984). The succession youngs toward the northeast, with the top of the Trinidad Sandstone inferred to be ca. 71.5 Ma near Raton, NM, and ca. 68.7 Ma near Walsenburg, CO (Berry, 2016).

The Vermejo Formation is unconformably (in the west) to conformably (in the east) overlain by the uppermost Cretaceous—Paleocene Raton Formation (up to 650 m or 2,100 ft thick; Pillmore et al., 1984; Flores, 1987; Berry, 2016). Palynostratigraphy, megafloral biostratigraphy, and ammonite biostratigraphy suggest that the magnitude of the Vermejo-Raton unconformity is spatially variable, representing up to 4 Myr in the southeastern part of the basin near Raton, NM (ca. 71.5 Ma to 67.5 Ma) and less than 1 Myr in the northern basin near Walsenburg, CO (Berry, 2016). The unconformity represents a period of erosion that occurred between the early and middle phases of Raton basin sedimentation (after Cather, 2004).

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109

4

The base of the Raton Formation is marked by a chert pebble conglomerate with a quartz sandstone matrix that extends across much of the basin (Flores, 1984; Pillmore et al., 1984). Above the basal conglomerate, the Raton Formation is an assemblage of sandstone, siltstone, carbonaceous shale, and coal that has traditionally been subdivided into a lower coal-rich zone, a cliff-forming “barren series” that contains few coal layers, and an upper coal-rich zone (Pillmore et al., 1984; Flores and Pillmore, 1987; Topper et al., 2011). Notably, the upper part of the lower coal zone preserves the K/Pg boundary layer, recorded in the Raton basin by palynology records (Tschudy et al., 1984) and a variably preserved, iridium-enriched, kaolinitic claystone layer (Pillmore et al., 1984). The presence of the K/Pg boundary layer provides a robust age datum of 65.4 ± 0.1 Ma (Obradovich, 1993) where preserved.

The Paleocene(?) Poison Canyon Formation (up to 750 m or 2,500 ft thick) interfingers with the Raton Formation adjacent to the Sangre de Cristo-Culebra uplift, but ultimately overlies the Raton Formation across much of the basin (Pillmore et al., 1984; Flores, 1987). In the west, the Poison Canyon Formation is differentiated from the Raton Formation based on its coarser sediment caliber (granular to conglomeratic), increased feldspar abundance, lack of coal, sandy siltstone overbank facies, and laterally amalgamated to stacked channel assemblages (Flores, 1987). In central to eastern parts of the basin, the Poison Canyon Formation can be strikingly similar to the underlying Raton and Vermejo Formations but contains less organic material (Flores, 1987).

The Poison Canyon Formation is overlain by the Eocene(?) Cuchara, Huerfano, and Farisita Formations (cumulatively > 2,000 m or 6,500 ft thick; Johnson, 1969; Tyler et al., 1995; Cather, 2004). The Cuchara, Huerfano, and Farisita Formations are unconformity-bound successions of varicolored (red, pink, white, tan) conglomerate, sandstone and mudstone that represent continued deposition in fluvial channel and overbank environments (Johnson, 1969; Rasmussen & Foreman, 2017).

METHODSOutcrop limitations and regional dip corrections

We acknowledge that the lack of lateral outcrop continuity and age control in the Raton and Poison Canyon Formations limits our ability to correlate individual fluvial channel fills across the basin. However, outcrops located east of the basin axis (Fig. 1) are nearly horizontal, dipping approximately 0.5° toward the north-northeast on average. Thus, along the east-west oriented outcrop transects described below, we use surface elevation as a proxy for relative stratigraphic height. We correct for regional dip between the two outcrop transects by inferring that a regional dip of 0.5° equates to an approximately 25 foot-per-kilometer change in elevation in the direction of dip (see Supplementary Information).

Depositional environments and paleodispersalWe provide detailed descriptions of channelized intervals of the Raton and Poison Canyon

Formations that are well-exposed in roadside outcrops along highways CO-12 and NM-555 in the central part of the Raton basin, generally east of the axis of the La Veta Syncline. Individual outcrops are named according to nearby localities (Fig. 1; Table 1). We purposefully highlight channels in the interior of the basin, distal to the range front, in order to minimize the presence of basin margin (e.g., alluvial fan) facies in the successions to focus solely on environments that are demonstrably fluvial. Both highways provide access to fresh outcrop surfaces that are not obscured by dense vegetation, providing insight into sub-decimeter-scale sedimentary features within fluvial channel fills. In addition, the road cuts transect the channel fills roughly perpendicular to their paleoflow directions (Fig. 1), enabling the measurement of paleochannel widths and depths. Representative stratigraphic sections through each channel complex

110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153

5

were measured at 5-cm resolution to document trends in grain size and sedimentary structures (Fig. 2; Supplementary Information). Fluvial channel deposits were photographed using a digital camera and were photostitched to generate detailed photo panoramas onto which channel fill architectural trends were interpreted (e.g., Figs. 3-8).

Apparent channel widths and depths were measured using a laser range finder when possible and were otherwise estimated using measurement tools available in Google Earth (Table 1). Due to the limitations of outcrop exposure and the effects of compaction during burial, the reported measurements are considered minimum dimensions. Paleocurrent directions were measured from ripple and dune cross-stratification, parting lineations, flute casts, and the orientations of transported logs. Similarly, channel orientations were estimated using the strike/dip of exposed channel margins and/or the surfaces of accretion sets within channel fills that roughly paralleled channel margins, wherein the paleoflow directions of host channels were estimated to be approximately parallel to the strikes of measured surfaces. Paleocurrent and channel orientation measurements were plotted using the freeware Stereonet (Allmendinger et al., 2013; Cardozo and Allmendinger, 2013; see Supplementary Information).

RESULTS: FLUVIAL CHANNEL DIMENSIONS AND ARCHITECTUREWe subdivide our descriptions of channel architecture into three sub-groups: 1) isolated channel

fills of the lower coal zone of the Raton Formation; 2) amalgamated channel-fills of the “barren series” of the Raton Formation; and 3) laterally amalgamated channel-fills of the Poison Canyon Formation. Importantly, we note that the three sub-groups share various characteristics and are not strictly unique from one another. Changes in fluvial architecture are transitional between the sub-groups, and subdivisions are made only for ease of description based on lithostratigraphic trends.

Isolated channel fills of the lower coal zone of the Raton FormationThe lower coal zone of the Raton Formation contains laterally isolated, white to light tan,

sandstone-dominated fluvial channel assemblages that are encased in overbank successions of coal, dark gray carbonaceous shale, and gray siltstone. Overbank siltstone is commonly interbedded with tabular, sheet-like beds of very fine- to fine-grained, current ripple-bearing sandstone up to 50 cm thick that pinch out over distances > 100 m. Sandstone and siltstone intervals may be heavily rooted and may contain fossilized tree casts in growth position, especially when associated with coal horizons. We highlight three channels including Zebra, Don’t Shoot, and Valdez. Individual channel elements range from ~30-50 m wide and ~4-8 m high (Table 1), with laterally nested channel assemblages > 90 m wide (e.g., Fig. 4a-b). Channel fills are sandstone-dominated, but may also be heterolithic at the tops of channel fills or near channel margins where sandstone beds pinch out and/or transition into overbank mudstone. Individual channel fill successions fine upward from lower medium- or upper fine-grained sandstone to upper very fine-grained sandstone and/or siltstone (Fig. 2). Disseminated organic debris is abundant and tends to be concentrated in heavily laminated intervals of sandstone channel fills; coalified logs are common as lags at basal and intra-channel erosional surfaces.

The internal architectures of individual channels vary. Zebra displays some cut-and-fill structures that parallel the channel margin, but no internal erosion surfaces with < 1 m relief. Coaly rip-ups occur at the channel base, and dark, organic-rich zones of laminated sandstone extend inward from the channel margin toward the channel axis (Fig. 3a-b, Fig. 9a). Zebra is dominated by parallel to gently undulating laminations (Fig. 9b) with some possible tangential dune cross-stratification that suggests paleoflow toward the north-northwest (Fig. 3a). Undulating surfaces may also reflect soft-sediment deformation.

154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197

6

Approximately 45 m (150 ft) up-section, Don’t shoot is heavily amalgamated with a relatively homogenous grain size, rendering internal structures difficult to discern (Fig. 3c-d). Similar to Zebra, Don’t Shoot contains abundant parallel to undulating lamination with zones of apparent soft-sediment deformation manifested as trains of meter-high antiforms and synforms with oversteepened limbs (Fig. 3d). The basal surface of Don’t Shoot is lined by siltstone rip-up clasts and an array of sub-parallel tree trunks that indicate paleoflow toward the northwest (Fig. 3c), which is corroborated by rare flute casts. Another 45 m (150 ft) up-section, Valdez includes multiple channel elements that are dominated by inclined, sandstone-dominated macroforms that parallel exposed channel margins (Fig. 4a-b). Macroform sets are bound by subtle erosion surfaces that generally parallel macroform surface (Fig. 4a-b); they also stack on top of one another to form a nested, composite channel-fill succession (Fig. 4c). Macroform sets are sandy and amalgamated in their lower (i.e., toeset) to middle (i.e., foreset) parts, but become heterolithic toward their tops (Fig. 4c-d). The sandstone portions of macroforms contain small dune (< 20 cm amplitude) and ripple cross-lamination (Fig. 9c). The heterolithic tops of macroforms also contain ripple cross-lamination that is commonly overprinted by rootlets. Siltstone interbeds at the tops of macroforms preserve large plant fragments (e.g., palm fronds, leaves, and ferns; Fig. 9d). Macroform orientations and paleocurrent measurements from cross-laminae indicate mean flow toward the north-northwest (Fig. 4d).

Amalgamated channel fills of the “barren zone” of the Raton FormationThe “barren zone” of the Raton Formation is characterized by laterally and vertically

amalgamated, light gray to tan, sandstone-dominated fluvial channel assemblages that are separated vertically by relatively thin overbank successions similar to those in the lower Raton Formation, but with minor coal. We highlight two channel complexes from this interval including Lightning and Coal Beans. Due to dense vegetation and outcrop orientations, the dimensions of individual channel elements are difficult to measure. However, individual channel elements are estimated to be ~5-8 m thick and at least 80-90 m wide based on the dimensions of exposed macroforms (see description below). Channel fill assemblages are sandstone-dominated, more so than underlying examples, but do contain some heterolithic intervals that are especially apparent near the base of the “barren zone”, such as within the Lightning channel fill. Higher in the section, at Coal Beans, heterolithic intervals were only observed at the tops of channel fills. In both areas, individual channel fill successions fine upward from medium-grained to upper very fine-grained sandstone with siltstone.

Similar to the lower Raton Formation, the internal architectures of channel fills in the “barren zone” vary. The Lightning channel fill succession begins with meter-scale inclined heterolithic strata that are poorly exposed in outcrop (no photographs available). These are incised and overlain by a sandstone-dominated channel element that contains pervasive climbing ripple cross-lamination (Fig. 9e) with some tangential dune cross-stratification (< 20 cm amplitude; Supplementary Information). Within climbing ripple successions, climb surfaces and ripple partings are crudely defined by concentrations of oxidized minerals (dark laminations in Fig. 9e); these features may be confused for dune cross-strata, resulting in uncertain paleoflow indicators. Paleocurrent measurements from a collection of dune and ripple surfaces provide an inconclusive paleoflow direction for Lightning, but channel margin orientations suggest northeast-directed paleodispersal. Approximately 128 m (420 ft) up-section, Coal Beans is dominated by large (~7.5 m high, ≥30 m wavelength) accretionary macroforms that drape an erosion surface across a lower sandstone interval (Fig. 5). Channel margins are not clearly exposed. Both the lower and upper sandstone intervals exhibit soft-sediment deformation with folds ≥ 1 m high. Soft-sediment folds are

198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241

7

apparent at the top of the lower sandstone interval and at the toes of macroforms in the upper sandstone interval, with apparent crumpling of macroforms toward the northeast (Fig. 5). Both the lower and upper sandstone intervals contain trains of small mudstone intraclasts and disseminated plant debris that are aligned along low-angle to convolute laminations. Ripple lamination (Fig. 9f) was only observed near the tops of macroforms where they interfinger with siltstone. No paleocurrent measurements were collected at this location, but the northeastward vergence of deformed macroforms suggest either 1) northeastward paleodispersal if the macroforms accreted downstream or 2) northwest or southeast paleodispersal if the macroforms accreted laterally. Given consistent northward paleocurrent measurements from all other observed channels, we prefer a northwest or northeast paleoflow direction for the Coal Beans channel.

Laterally amalgamated channel fills of the Poison Canyon FormationThe contact between the Raton and Poison Canyon Formations is gradational, but in general,

sandstone-dominated intervals of the Poison Canyon Formation are coarser than those of the Raton Formation. Fluvial channel successions consist of laterally to vertically amalgamated channel elements containing orangish-tan sandstone that interfingers with reddish-brown, sandy siltstone. Notably, silt-rich sandstone makes up a significant percentage of some channel fills, resulting in a heterolithic character (e.g., Fig. 7). Coals are absent in the Poison Canyon Formation, but large plant fragments and mats of plant debris are abundant in both overbank deposits and the finer-grained intervals of channel fills (Fig. 2b). We highlight four channel fill successions from the Poison Canyon Formation including Flank Steak, Wavy Gravy, Big Muddy, and CHS1885. Individual channel elements range from ~25-70 m wide and ~3-5 m thick. Channel elements fine upward from pebble-bearing coarse-grained sandstone to lower fine-grained sandstone and siltstone (Fig. 2b-c). Sandstones contain abundant mica. Pebbles and rare cobbles (< 5 cm) are present as thin lags or trains within sandstone beds; clasts are typically angular and include granite, meta-granite, chert, whitish quartzite, red quartzofeldspathic sandstone, and megacrystalline quartz and feldspar.

Again, the internal architectures of individual Poison Canyon channel fills vary. The Flank Steak channel fill, which occurs at the mapped contact of the Raton and Poison Canyon Formations, is heavily amalgamated with few discernable internal features (Fig. 6a-b). Crude amalgamation surfaces are lined by coalified to petrified logs, and disseminated organic material is relatively abundant. Approximately 61 m (200 ft) up-section, Wavy Gravy is similarly amalgamated but preserves multimeter-high cut-and-fill features with internal sedimentary structures. Bedding dominantly consists of undulating lenses of structureless sandstone that are bound by laminated intervals of organic-rich sandstone (Fig. 6b-c). Lenses contain low-angle inclined to wavy bedding, with minor (possible) tangential dune cross-stratification that suggests northeast-directed paleoflow (Fig. 6c). Some undulation may be related to soft-sediment deformation. Another 9 m (30 ft) up-section, Big Muddy includes many channel elements that are dominated by inclined macroforms composed of alternating, pebbly, very coarse- to coarse-grained sandstone and silty, very fine-grained sandstone (Figs. 2b, 7a-b). Shale drapes are also preserved between some sandstone beds (Fig. 2b), but are commonly eroded to form trains of intraclasts (Fig. 9g). The macroforms tend to parallel exposed channel margins, but are separated from one another by distinct erosion surfaces that are sub-parallel to macroform surfaces (Fig. 7d). Macroforms are laterally amalgamated to form a >170 m- (550 ft-) wide composite channel-fill succession (Fig. 7a-b). Gravelly beds are concentrated near the axes of channel elements. Many of the inclined macroforms are lenticular and/or have pinch-swell geometries, and some display soft-sediment folding that verges toward the channel element axis (Fig. 7c). Internally, sandstone beds contain rare dune cross-stratification (< 1 m

242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285

8

amplitude; Fig. 9g), low-angle lamination, and/or convolute lamination. Intervening finer-grained intervals commonly contain ripple cross-lamination and organic material. Macroform orientations and paleocurrent measurements from ripples and dunes indicate north-northwestward paleoflow (Fig. 7d). Near the top of the Poison Canyon Formation, channel elements of CHS1885 are heavily amalgamated to heterolithic (Fig. 8). Sandstone beds vary from pebbly with a granular or very coarse-grained sandstone matrix to medium- or fine-grained sandstone (Fig. 2c). Rip-up clasts (up to 10 cm) of sandy siltstone are common. Amalgamated, sandy channel fills appear mostly structureless with undulating to convex amalgamation surfaces, inclined macroform surfaces, and soft-sediment deformation features including gentle folds (Fig. 8a-d). Some inclined units contain crude bedding-parallel lamination. A wedge of graded, bedding-parallel laminae (each up to 1 cm thick) also occurs near the top of one channel element (Fig. 8e-f). Meter-scale tangential dune cross-strata lined by pebbles and granules (Fig. 8e) are apparent in one area, indicating east-northeast paleoflow (Fig. 8a). Similar to Big Muddy, finer-grained interbeds and overbank material are organic-rich, micaceous, and heavily rippled.

INTERPRETATIONSFluvial Style and Processes

Isolated channel fills of the lower coal zone of the Raton Formation. --- We interpret the observed channel fills from the lower Raton Formation as two channel types within a classical sinuous channel network: 1) perennial, sinuous trunk channels (e.g., Valdez) and 2) transient cut-off or crevasse splay channels (e.g., Zebra and Don’t Shoot). Rationale for each interpretation is provided below. These interpretations are consistent with those of Flores and Pillmore (1987) and Flores (1987), who interpreted the lower Raton Formation in the middle of the Raton basin to record meandering streams that were laterally confined by voluminous levee-overbank, splay, and coal swamp deposits.

The sigmoid, channel margin-parallel macroforms that dominate the Valdez channel fill (Fig. 4a-b) are interpreted as laterally accreting point bar macroforms (Walker and Cant, 1984; Flores and Pillmore, 1987). The presence of point bar deposits supports the interpretation of sinuous fluvial channels that were sustained by paired cut bank erosion and point bar migration (Walker and Cant, 1984). Lateral nesting of erosion surface-bound point bar sets (Fig. 4a-b) reveals lateral shifts in the position of the channel following high-energy events (floods) that modified previous channel geometries, whereas vertical stacking and nesting of point bar sets (Fig. 4a-c) reveals channel aggradation (Walker and Cant, 1984). The high volume of fine-grained overbank deposits in the Raton Formation (estimated to be > 50%; Flores, 1987) restricted sinuous channels from migrating great distances, and instead prompted channel and floodplain aggradation paired with periodic channel avulsion. Ripple and dune cross-stratification (Fig. 9c) throughout the point bar deposits reflect the migration of ripples and small, sinuous-crested dunes across point bar surfaces under subcritical flow conditions. Inclined to horizontal heterolithic deposits at the tops of channel fills (Figs. 2a, 4c-d) represent gradual channel abandonment, wherein mudstone beds represent suspension deposition during low flow energy or ponded conditions and rippled sandstone beds represent periodic, higher-energy flows into the semi-abandoned channel. The presence of rooted horizons within upper heterolithic intervals supports an overall quiescent setting during abandonment. Given the lack of rooted horizons or paleosols within the lower parts of point bar successions, we interpret the sinuous channels to have flowed perennially.

In contrast, cut-off and crevasse splay-related channels lack sigmoid point bar macroforms (Fig. 3a-d), indicating that the channels were not sustained by progressive meander migration. Rather, these channels were more likely rapidly excavated and filled by flood events that breached levees and flowed

286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329

9

away from trunk channels (Walker and Cant, 1984). Examples discussed here occur immediately above a coal (Zebra, Don’t Shoot; Fig. 3a-d) and/or at the top of coarsening-upward crevasse splay successions (Don’t Shoot; Fig. 3c-d; after Flores, 1987), recording incision into levee-overbank environments. Concentrated, channel-parallel tree casts lining the erosional bases of channels (e.g., Don’t Shoot) also support catastrophic flooding and rapid erosion into overbank environments. Apparent inclined heterolithic strata at the base of Zebra (Fig. 4a-b) are interpreted as rotated blocks of overbank deposits that slumped into the channels closely following excavation (after Flores and Pillmore, 1987), rather than heterolithic bar foresets deposited under fluctuating flow conditions. Within the sandstone-dominated channel-fills, transient, rapid deposition is supported by a lack of well-developed point bar macroforms and dune cross-stratification, the presence of abundant soft-sediment deformation, and an abrupt vertical transition from sandy channel fill back to overbank mudstone (Fig. 4a-d). In addition, some of the undulating laminated intervals observed at Zebra and Don’t Shoot (e.g., Fig. 9b) may represent aggradational supercritical flow structures such as those produced by antidunes (McBride et al., 1975; Allen, 1984; Cheel, 1990), which may have formed due to rapid deposition during the initial, high-energy phases of flood events.

Amalgamated channel fills of the “barren zone” of the Raton Formation. --- Following the rationale of Flores and Pillmore (1987), we interpret the “barren zone” of the Raton Formation to record a period of deposition by laterally mobile, avulsion-prone sinuous channels that lacked the robust overbank confinement that was present for lower Raton Formation channels. The lack of aggradational overbank deposits in the “barren zone” allowed sinuous channels to rapidly incise, fill, and avulse to new positions, resulting in laterally and vertically amalgamated channel fill assemblages with minor preservation of overbank deposits (Flores and Pillmore, 1987). Incision into overbank areas is supported by the presence of inclined heterolithic intervals at the base of Lightning that we interpret as foundered overbank deposits (similar to those pictured for the lower Raton Formation in Fig. 3b). Periodic channel filling and avulsion are recorded by the lateral incision and nesting of channel fills, such as at Coal Beans where large, accretionary macroforms drape an erosional surface through an underlying fluvial sandstone deposit (Fig. 5). Based on available field data, it is unclear whether the accretionary macroforms at Coal Beans represent laterally or downstream-accreting point bars; however, both are possible in a sinuous channel system (e.g., Walker and Cant, 1984; Miall, 1988; Almeida et al., 2016; Ghinassi et al., 2016). Climbing ripple successions at Lightning (Fig. 9e) reveal rapid flow deceleration paired with contemporaneous suspension load fallout and bedload transport (McKee et al., 1967; Picard and High, 1973); their dominance in the channel fill shows that flows carried abundant sediment in suspension, only depositing climbing ripple successions as flows waned to subcritical conditions. Rapid sedimentation is also indicated by soft sediment deformation throughout channel fills, on both the bed- and macroform-scale (e.g., Fig. 5). Overall, channel systems in the “barren zone” were morphologically similar to those in the lower Raton Formation, but with higher lateral mobility.

Laterally amalgamated channel fills of the Poison Canyon Formation. --- We interpret the observed channel fills from the Poison Canyon Formation as ephemeral, laterally mobile and avulsion-prone sinuous channels that were poorly confined by overbank material. This interpretation revises previous work that recognized features associated with sinuous channels (Flores and Pillmore, 1987) but generalized the coarse-grained channel fills of the Poison Canyon Formation as sand-rich alluvial plain and braided stream deposits (Flores, 1987). Although the Poison Canyon channels were laterally mobile, they lack the trademark architectural features of braided river systems such as cross-cutting, compound bar forms containing multiple scales of dune cross-strata (Walker and Cant, 1984).

330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373

10

Similar to the lower Raton Formation, the inclined, channel margin-parallel macroforms that are observed in most Poison Canyon channel fills are interpreted as laterally accreting point bar deposits (Flores and Pillmore, 1987), the presence of which suggests channel migration via cut bank erosion and point bar migration (Walker and Cant, 1984). However, multiple features within the channel fills indicate that channel migration was highly episodic, and perhaps sometimes catastrophic, rather than sustained through time. First, point bar sets are laterally nested and bound by erosion surfaces that obviously truncate earlier macroforms (e.g., Big Muddy; Fig.7a-b, d). Lateral nesting of erosion surface-bound point bar sets reveals lateral shifts in the channel position following high-energy events (floods). Unlike the erosion surfaces that only subtly modify macroform surfaces in the lower Raton Formation, macroform sets are significantly excavated by similar erosion surfaces in the Poison Canyon Formation. Second, some point bar sets are incised and overlain by composite lenses of heavily amalgamated, gravelly sandstone (e.g., Wavy Gravy and CHS1885; Figs. 6c-d, 8a-d). The nesting of amalgamated sandstone lenses within heterolithic point bar deposits suggests re-excavation and filling of existing channel forms by highly erosive flood events.

In addition to episodic channel reorganization induced by high-energy events, multiple features suggest that floods were flashy and resulted in rapid sediment deposition. Unlike the lower Raton Formation, point bar sets in the Poison Canyon Formation are heterolithic, marked by couplets of gravelly coarse-grained sandstone and silty fine-grained sandstone. The finer-grained horizons commonly extend into and across channel thalweg areas (e.g., Big Muddy; Fig. 7a-b). The alternating distribution of coarse- and fine-grained point bar deposits indicates fluctuations in discharge and flow velocity through time (Plink-Bjorklund, 2015), wherein coarse intervals were deposited during high-energy stages of flood events and finer intervals were deposited by lower energy flows or during the waning stages of flood events. Fluctuating discharge is supported by the presence of soft-sediment deformed sedimentary structures within coarse intervals that indicates rapid deposition followed liquefaction, as well as well-defined ripple lamination within fine intervals that records rapid deposition under lower flow regime conditions. In addition to small-scale soft-sediment deformation, some point bar successions display multimeter-scale slumping of point bars toward the channel thalweg with slump topography infilled/healed by subsequent heterolithic deposits (e.g., Big Muddy; Fig. 7c); such slumping again supports rapid deposition of point bar macroforms and failure due to liquefaction.

Within amalgamated sandstone lenses, abundant scour-and-fill features that lack internal cross-stratification (Fig. 8a-d) reveal Froude supercritical flow conditions followed by rapid sediment aggradation and infilling of scour features (Plink-Bjorklund, 2015). Rip-up clasts at scour surfaces support erosive upper flow regime conditions, as well as indicate the erosion of in-channel fine-grained intervals (Plink-Bjorklund, 2015). Undulating lamination sets at Wavy Gravy support the formation and aggradation of antidune-like supercritical flow structures during high-energy phases of flow (McBride et al., 1975; Allen, 1984; Cheel, 1990). Similarly, the graded, centimeter-thick laminae at CHS1885 reveal rapid deposition and aggradation that is consistent with upper flow regime conditions and heavy sediment loads (e.g., “overthickened laminae” of Plink-Bjorklund, 2015). The combination of high flow velocity, high deposition rates, and suspension deposition largely suppresses the formation of dunes and cross-stratification (Allen and Leeder, 1980). However, rare dune cross-stratification in gravelly sandstones (e.g., CHS1885; Fig. 8e) reflects periodic flows that were strong but subcritical. Similar to the lower Raton Formation, the upper parts of Poison Canyon Formation channel fills may also be composed of heterolithic to muddy strata that represent channel abandonment (e.g., Big Muddy; Fig. 7e).

Vertical trends in grain size and channel element dimensions

374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416

417418

11

Considered in succession, channel elements from the Raton and Poison Canyon Formations display systematic changes in grain size distributions and channel element dimensions. Figure 10a summarizes the minimum, maximum, and average sediment size recorded for each observed channel element (Table 1). We note that these data do not reflect systematic, laboratory-based analysis of grain size distributions, and are based purely on field observations made while measuring stratigraphic sections through each channel fill complex (e.g., Fig. 2). The average grain size of channel fills increases gradually from fine-grained sand in the lower Raton Formation to lower coarse-grained sand in the Poison Canyon Formation. Likewise, minimum grain size increases from lower fine- to medium-grained sand, and maximum grain size increases from upper fine-grained sand to pebbles and cobbles. The grain size data reveal net upward-coarsening of channel fills during deposition of the Raton and Poison Canyon Formations and concurrently highlight a temporal widening of grain size distributions. These trends suggest a change in sediment source, a decrease in sediment sorting between source and sink, and/or a decrease in weathering capacity.

Figure 10b shows vertical trends in the dimensions of individual channel elements by summarizing pairs of channel element widths and depths as aspect ratios (where aspect ratio is calculated as element width divided by depth; Table 1). When all measured channel elements are considered together, there is a weak increase in aspect ratio through time (white circles; Fig. 10b), indicating that channel element width generally increases relative to channel element depth through time. To better highlight this trend, we separate channel elements whose margins and heights were fully visible in outcrop (Zebra, Don’t Shoot, Valdez, Wavy Gravy, Coal Beans Big Muddy 1, CHS1885 1) from those whose full dimensions were obscured (Boom, Lightning, Flank Steak, Big Muddy 2, CHS1885 2). Although the dimensions remain variable, there is a slightly stronger positive correlation between aspect ratio and stratigraphic height (gray circles; Fig. 10b). Finally, we separate channel elements that are considered to represent well-developed trunk fluvial channels (Valdez, Coal Beans, BM1, CHS1885-1) from channels that are considered to be subsidiary. These trunk channels (black circles; Fig. 10b) show a strong positive correlation between stratigraphic height and aspect ratio (R2=0.992), supporting that trunk fluvial channels became wider and shallower through time.

Paleodispersal patternsWith one exception, paleocurrent indicators from all observed channels suggest persistent

paleodispersal toward the northern hemisphere (Fig. 1). Paleocurrent measurements from the lower Raton Formation consistently indicate a north-northwest paleodispersal direction, whereas measurements from the Poison Canyon Formation show north to east-northeast paleodispersal directions. These results are generally consistent with environmental reconstructions depicted by Flores et al. (1985). We interpret the north- to northeast-directed paleoflow directions to reflect axial flow through the middle of the Raton basin, approximately parallel to the emerging range front. Channel-specific variations in average flow direction may be explained by channel sinuosity and/or differences between the absolute flow directions of trunk and subsidiary channels.

DISCUSSIONTemporal trends in fluvial depositional systems

Based largely on lithostratigraphic mapping, previous work defined the Raton and Poison Canyon Formations as more-or-less conformable and interfingering with one another, with different stratigraphic intervals representing the end-members of a classical fluvial depositional spectrum (e.g., Rust and Koster,

419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462

12

1984). The Poison Canyon Formation and overbank-poor intervals (“barren series”) of the Raton Formation were interpreted as range-proximal alluvial fan to braided fluvial deposits, whereas overbank-rich, coal-bearing intervals of the Raton Formation were interpreted as distal, meandering fluvial deposits (Flores et al., 1985; Flores and Pillmore, 1987). Accordingly, temporal variations in fluvial style were attributed to periodic basinward advance of proximal environments across distal environments.

Our new interpretations allow us to modify the previous model by highlight ing five systematic trends in the Raton-Poison Canyon stratigraphy: 1) net upward-coarsening and upward-widening of in-channel grain size distributions (Fig. 10a); 2) net upward-widening and -shallowing of channel elements (Fig. 10b); 3) an upward-decrease in the volume of overbank fines and lower preservation of organic material; 4) an increase in apparent channel mobility; and 5) a transition in fluvial architecture that suggests a shift from perennially to ephemerally occupied channels (Figs. 4, 7). Despite these temporal changes, we consistently interpret basin-center fluvial systems to be composed of northward-flowing sinuous channel networks. On the most fundamental level, we argue that basin-center fluvial deposits of the Raton and Poison Canyon Formations record changes in the lateral confinement of sinuous channels, and therefore changes in channel stacking patterns, rather than alternating deposition by meandering and braided rivers. Sinuous channels were strongly confined by thick overbank successions during deposition of the lower Raton Formation, producing vertically stacked channel fill assemblages (e.g., Valdez; Fig. 4a-c). Channel confinement decreased during deposition of the “barren series” of the Raton Formation, allowing for both vertical and lateral amalgamation of channel fills. Channel confinement continued to decrease during deposition of the Poison Canyon Formation, allowing for the deposition of laterally extensive, nested channel fill assemblages (e.g., Big Muddy; Fig. 7a-b). On a broader scale, we speculate that the gradual transition from isolated, sinuous channels in the lower Raton Formation to laterally amalgamated channels in the Poison Canyon Formation represents progradation of fluvial fans (e.g., DeCelles and Cavazza, 1999; Leier et al., 2005; Hartley et al., 2010; Weissmann et al., 2010; Latrubesse, 2015; Ventra and Clarke, 2018) across the Raton basin. In this context, the coal-bearing intervals of the lower Raton Formation are deposits of a basin-axial river system, whereas the “barren series” of the Raton Formation and the Poison Canyon Formation are deposits of sinuous channels located progressively higher on a fluvial fan surface. Progradation of a fluvial fan system over the basin is supported by the observed trends in grain size (Fig. 10a), channel geometry (Fig. 10b), overbank deposit character, degree of channel avulsion, and evidence for perennial vs ephemeral channel occupation (REFS). It is reasonable that uplift along the Sangre de Cristo-Culebra Laramide uplift and/or Late Cretaceous to Paleogene climatic events induced changes in sediment supply, sediment transport mechanisms, sediment routing, and basin subsidence, each of which may affect the migration of depositional environments within the Raton basin. Detrital zircon data from Bush et al. (2016) suggest regional drainage reorganization during deposition of the Raton Formation, wherein the lower Raton Formation received sediment from the Upper Cretaceous arc and the upper Raton and Poison Canyon Formations were fed by local sediment sources. In other words, fluvial systems of the lower Raton Formation were far-reaching with headwaters beyond the Raton basin, whereas overlying fluvial fans emanated from the emerging Sangre de Cristo-Culebra uplift.

Implications for tectonic and climatic controlsPrevious studies have invoked tectonism as the primary driver of sedimentation in the Raton

basin, with only minor emphasis on potential climatic influences. Although uplift along the Sangre de

463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506

13

Cristo-Culebra Laramide uplift certainly influenced subsidence rates and patterns (Flores, 1987; Lindsey, 1998; Higley, 2007; Bush et al., 2016), sediment source areas (Flores, 1987; Bush et al., 2016), and sediment routing (Cather, 2004; Bush et al., 2016), our new data underscore the concurrent influence of climate on sediment availability and transport.

Available sedimentary and paleobotanical evidence indicate a transition in the Raton basin from warm and perennially wet during deposition of the lower Raton Formation to drier with periodic rainfall during deposition of the upper Raton Formation and Poison Canyon Formation. First, this interval hosts a marked transition from volumetrically abundant, gray to black, clay- and silt-rich overbank deposits (e.g., FIG) to lower volume, reddish-orange and sandy overbank deposits (e.g., Figs. 3, 4); accompanying this is better preservation of detrital feldspar grains through time (Flores, 1987). These trends suggest increasingly poor weathering of source rocks, possibly due to climatic drying and/or cooling that would have inhibited chemical weathering of feldspathic and lithic sediment to silt and clay. Paleobotanical records from the K/Pg boundary record a subhumid climate during deposition of the lower Raton Formation followed by a rapid-onset, humid, high-precipitation climate after the K/Pg impact event (Wolfe and Upchurch, 1987); such conditions would foster high-intensity weathering of source rocks at this time, both degrading feldspathic sediment and providing abundant fine-grained sediment to the basin. The wet, post-impact climate persisted into the early Paleocene (Wolfe and Upchurch, 1987), presumably becoming drier afterward. Although unroofing of different basement rocks in the Sangre de Cristo-Culebra Laramide uplift is an alternative explanation for changes in the character of sediment in the basin (e.g., Flores, 1987; Bush et al., 2016), the increasingly poor accumulation and preservation of organic material supports climatic drying and/or cooling through the Raton-Poison Canyon transition. Finally, the transition in fluvial channel architecture described herein reveals a transition in fluvial style from perennial, axial rivers to ephemeral channels located on a fluvial fan, ostensibly related to climatic drying and/or cooling.

We propose two possibilities for the apparent transition in climate preserved by the Raton and Poison Canyon Formations. Basinal drying, cooling, and increasingly periodic rainfall may reflect the influence of Laramide uplifts on atmospheric patterns, wherein the growth of Laramide topography may have produced localized rain shadows across the Laramide intraforeland province. However, Paleogene paleoclimate models suggest summertime (monsoonal) winds blowing northwestward from the Gulf of Mexico toward the Raton basin (Sewall and Sloan, 2006, their fig. 2f). Since the Sangre de Cristo-Culebra uplift bounds the Raton basin on its western margin, it is unlikely that precipitation-bearing winds from the Gulf would have experienced rain-out prior to reaching the Raton basin, making a rain shadow effect across the Raton basin unlikely. Alternatively, we posit that drying, cooling, and increasingly periodic rainfall may instead reflect global Paleogene climatic events, such as gradual cooling during the Paleocene (REFS) or abrupt, transient warming associated with the Paleocene-Eocene thermal maximum (PETM; REFS). Pollen data (REFS??! Pillmore??) from the NM-555 outcrop belt support that the Raton Formation is Maastrichtian near the base of the “barren series” and is Paleocene throughout the upper Raton Formation (Fig. 1). Above the Raton-Poison Canyon contact, pollen samples yield species that are indeterminately Paleocene or Eocene. Although the pollen ages are not definitive, they prompt further investigations into the age of the Poison Canyon Formation, and potentially place the Paleocene-Eocene boundary up to 2,500 feet (~760 m) lower in the Raton basin stratigraphy. If all or part of the Poison Canyon Formation is Eocene, it is reasonable that Paleocene drying or the PETM is captured within the Raton-Poison Canyon interval. Accordingly, the changes in fluvial architecture described herein may have been a response to enhanced monsoonal precipitation, discharge, and sediment

507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550

14

flux during the PETM, such as has been observed for other Laramide basins in North America (Foreman et al., 2012; Foreman, 2013; Zellman et al., 2020).

CONCLUSIONSWe demonstrate that basin-center fluvial deposits of the Maastrichtian-Eocene(?) Raton and

Poison Canyon Formations in the Raton basin of Colorado-New Mexico, USA, record marked temporal changes in fluvial style. Although channel systems remained sinuous throughout that time period, channel fill architectures reveal a transition from regional, through-going, perennial river systems (exemplified by sandstone-dominated, aggradational point bar successions) to more locally sourced, ephemeral river systems (exemplified by heterolithic, laterally amalgamated point bar successions). We speculate that the transition in river type, fluvial style, and stratigraphic architecture occurred due to 1) partitioning of the Western Interior Region by Laramide intraforeland uplifts and 2) global and/or local climatic trends of drying, cooling, and increasingly periodic rainfall during the early Paleogene. Although available age data are not conclusive, we speculate that the contact of the Raton and Poison Canyon Formations may occur near the Paleocene-Eocene boundary, up to 2,500 ft (~760 m) lower in the Raton basin stratigraphy than previously placed. These new findings provide a stronger context for understanding the depositional history of the Raton basin, and ultimately lay a foundation for additional studies regarding basin-wide chronology as well as the effects of tectonic and climatic drivers on basin and landscape evolution.

ACKNOWLEDGEMENTSWe gratefully acknowledge Vermejo Park Ranch for land access and support with field work in

the New Mexico sector of the basin. We also thank _____________, ___________, and __________ for constructive reviews that improved this work. This draft manuscript is distributed solely for purposes of scientific peer review. Its content is deliberative and predecisional, so it must not be disclosed or released by reviewers. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

REFERENCESAllen, J.R.L., 1984, Sedimentary Structures, Their Character and Physical Basis: Elsevier, New York,

593 p.Allen, J.R.L. and Leeder, M.J., 1980, Criteria for the instability of upper-stage plane beds:

Sedimentology, v. 27, p. 209-217.Allmendinger, R.W., Cardozo, N.C., and Fisher, D., 2013, Structural Geology Algorithms: Vectors &

Tensors: Cambridge, England, Cambridge University Press, 289 p.Almeida, R.P., Freitas, B.T., Turra, B.B., Figueiredo, F.T., Marconato, A., and Janikian, L., 2016,

Reconstructing fluvial bar surfaces from compound cross-strata and the interpretation of bar accretion direction in large river deposits: Sedimentology, v. 63, p. 609-628.

Berry, K., 2016, Biostratigraphy of the Campanian-Maastrichtian, Mid-Maastrichtian, and Cretaceous-Tertiary boundaries in the Raton Basin: Implications for models of early Laramide tectonism and for the origin of the enigmatic disturbed zone: The Mountain Geologist, v. 53, no. 2, p. 75-91.

Blum, M. and Pecha, M., 2014, Mid-Cretaceous to Paleocene North American drainage reorganization from detrital zircons: Geology, v. 24, p. 607-610.

551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592

15

Bush, M.A., Horton, B.K., Murphy, M.A., and Stockli, D.F., 2016, Detrital record of initial basement exhumation along the Laramide deformation front, southern Rocky Mountains: Tectonics, v. 35, p. 2117-2130.

Cardozo, N.C. and Allmendinger, R.W., 2013, Spherical projections with OSXStereonet: Computers & Geosciences, v. 51, no. 0, p. 193-205, doi: 10.1016/j.cageo/2012.07.021.

Carroll, A.R., Doebbert, A.C., Booth, A.L., Chamberlain, C.P., Rhodes-Carson, M.K., Smith, M.E., Johnson, C.M., and Beard, B.L., 2008, Capture of high-altitude precipitation by a low-altitude Eocene lake, western U.S.: Geology, v. 36, p. 791-794.

Cather, S.M., 2004, Laramide orogeny in central and northern New Mexico and southern Colorado, in Mack, G.H. and Giles, K.A., eds., The geology of New Mexico: a geologic history: New Mexico Geological Society, Special Publication 11, Socorro, NM, p. 203-248.

Chamberlain, C.P., Mix, H.T., Mulch, A., Hren, M.T., Kent-Corson, M.L., Davis, S.J., Horton, T.W., and Graham, S.A., 2012, The Cenozoic climate and topographic evolution of the western North American Cordillera: American Journal of Science, v. 312, p. 213-262.

Cheel, R.J., 1990, Horizontal lamination and the sequence of bed phases and stratification under upper-flow-regime conditions: Sedimentology, v. 37, p. 517-529.

Copeland, P., Currie, C.A., Lawton, T.F., and Murphy, M.A., 2017, Location, location, location: The variable lifespan of the Laramide orogeny: Geology, v. 45, p. 223-226.

Davis, S.J., Mix, H.T., Wiegand, B.A., Carroll, A.R., and Chamberlain, C.P., 2009a, Synorogenic evolution of large-scale drainage patterns: Isotope paleohydrology of sequential Laramide basins: American Journal of Science, v. 309, p. 549-602.

Davis, S.J., Mulch, A., Carroll, A.R., Horton, T.W., and Chamberlain, C.P., 2009b, Paleogene landscape evolution of the central North American Cordillera: Developing topography and hydrology in the Laramide foreland: GSA Bulletin, v. 121, p. 100-116.

DeCelles, P.G. and Cavazza, W., 1999, A comparison of fluvial megafans in the Cordilleran (Upper Cretaceous) and modern Himalayan foreland basin systems: GSA Bulletin, v. 111, p. 1315-1334.

Dickinson, W.R., Klute, M.A., Hayes, M.J., Janecke, S.U., Lundin, E.R., McKittrick, M., and Olivares, M.D., 1988, Paleogeographic and paleotectonic setting of Laramide sedimentary basins in the central Rocky Mountain region: GSA Bulletin, v. 100, p. 1023-1039.

Fan, M., Hough, B.G., and Passey, B.H., 2014, Middle to late Cenozoic cooling and high topography in the central Rocky Mountains: Constraints from clumped isotope geochemistry: Earth and Planetary Science Letters, v. 408, p. 35-47.

Flores, R.M., 1984, Comparative analysis of coal accumulation in Cretaceous alluvial deposits, southern United States Rocky Mountain basins, in Scott, D.F. and Glass, D.J., eds., The Mesozoic of Middle North America: Canadian Society of Petroleum Geologists Memoir 9, p. 373-385.

Flores, R.M., 1987, Sedimentology of Upper Cretaceous and Tertiary siliciclastics and coals in the Raton Basin, New Mexico and Colorado, in Lucas, S.G. and Hunt, A.P., eds., Northeastern New Mexico: New Mexico Geological Society 38th Annual Fall Field Conference Guidebook, p. 255-264.

Flores, R.M. and Pillmore, C.L., 1987, Tectonic control on alluvial paleoarchitecture of the Cretaceous and Tertiary Raton Basin, Colorado and New Mexico, in Ethridge, F.G., Flores, R.M., and Harvey, M.D., eds., Recent Developments in Fluvial Sedimentology: SEPM Special Publication 39, p. 311-230.

593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635

16

Flores, R.M., Pillmore, C.L., and Merewether, E.A., 1985, Overview of depositional systems and energy potential of Raton Basin, Colorado and New Mexico, in Macke, D.L. and Maughan, E.K., eds., Rocky Mountain Section Field Trip Guide, p. 41-74.

Foreman, B.Z., 2014, Climate-driven generation of a fluvial sheet sand body at the Paleocene-Eocene boundary in north-west Wyoming: Basin Research, v. 26, p. 225-241.

Foreman, B.Z., Heller, P.L., and Clementz, M.T., 2012, Fluvial response to abrupt global warming at the Paleocene/Eocene boundary: Nature, v. 491, p. 92-95.

Ghinassi, M., Ielpi, A., Aldinucci, M. and Fustic, M., 2016, Downstream-migrating fluvial point bars in the rock record: Sedimentary Geology, v. 334, p. 66-96.

Hartley, A.J., Weissmann, G.S., Nichols, G.J., and Warwick, G.L., 2010, Large distributive fluvial systems: Characteristics, distribution, and controls on development: Journal of Sedimentary Research, v. 80, p. 167-183.

Higley, D.K., 2007, Chapter 2 Petroleum Systems and Assessment of Undiscovered Oil and Gas in the Raton Basin-Sierra Grande Uplift Province, Colorado and New Mexico, in Higley, D.L., ed., Petroleum Systems and Assessment of Undiscovered Oil and Gas in the Raton Basin-Sierra Grande Uplift Province, Colorado and New Mexico – USGS Province 41: U.S. Geological Survey Digital Data Series DDS-69-N, 124 p.

Johnson, R.B., 1969, Geologic map of the Trinidad Quadrangle, south-central Colorado: USGS Miscellaneous Geologic Investigations Map I-558, 1:250,000.

Knopf, A., 1936, Igneous geology of the Spanish Peaks region, Colorado: GSA Bulletin, v. 47, p. 1727-1784.

Latrubesse, E.M., 2015, Large rivers, megafans and other Quaternary avulsive fluvial systems: A potential “who’s who” in the geological record: Earth-Science Reviews, v. 146, p. 1-30.

Lawton, T.F., 2008, Laramide sedimentary basins, in Miall, A.D., ed., Sedimentary Basins of the World: The Sedimentary Basins of the United States and Canada, volume 5: Elsevier, Netherlands, p. 429-450.

Leier, A.L., DeCelles, P.G., and Pelletier, J.D., 2005, Mountains, monsoons, and megafans: Geology, v. 33, p. 289-292.

Lindsey, D.A., 1998, Laramide structure of the central Sangre de Cristo Mountains and adjacent Raton basin, southern Colorado: The Mountain Geologist, v. 35, p. 55-70.

McBride, E.F., Shepherd, R.G., and Crawley, R.A., 1975, Origin of parallel near-horizontal laminae by migration of bed forms in a small flume: Journal of Sedimentary Petrology, v. 45, p. 132*139.

Miall, A.D., 1988, Reservoir heterogeneities in fluvial sandstones: lessons from outcrop studies: AAPG Bulletin, v. 72, p. 682-697.

Muller, O.H., 1986, Changing stresses during emplacement of the radial dike swarm at Spanish Peaks, Colorado: Geology, v. 14, p. 157-159.

McKee, E.D., Crosby, E.J., and Berryhill, M.L., 1967, Flood Deposits, Bijou Creek, Colorado, June 1965: Journal of Sedimentary Petrology, v. 37, p. 829-851.

Obradovich, J.D., 1993, A Cretaceous time scale, in Caldwell, W.G.E. and Kauffman, E.G., eds., Evolution of the Western Interior Basin: Geological Association of Canada, Special Paper 39, p. 379-396.

Penn, B.S. and Lindsey, D.A., 2009, 40Ar/39Ar dates for the Spanish Peaks intrusions in south-central Colorado: Rocky Mountain Geology, v. 44, p. 17-32.

636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678

17

Picard, M.D. and High, L.R., 1973, Sedimentary structures of ephemeral streams: Developments in Sedimentology 17, Elsevier, Amsterdam, 223 p.

Pillmore, C.L., Tschudy, R.H., Orth, C.J., Gilmore, J.S., and Knight, J.D., 1984, Geologic framework of nonmarine Cretaceous-Tertiary boundary sites, Raton Basin, New Mexico and Colorado: Science, v. 223, no. 4641, p. 1180-1183.

Plink-Bjorklund, P., 2015, Morphodynamics of rivers strongly affected by monsoon precipitation: Review of depositional style and forcing factors: Sedimentary Geology, v. 323, p. 110-147, doi: 10.1016.j.sedgeo.2015.04.004

Rasmussen, D.M. and Foreman, B.Z., 2017, Provenance of lower Paleogene strata in the Huerfano basin: implications for uplift of the wet mountains, Colorado, USA: Journal of Sedimentary Research, v. 87, p. 579-593.

Rust, B.R. and Koster, E.H., 1984, Coarse Alluvial Deposits, in Walker, R.G., ed., Facies Models Second Edition: Geoscience Canada, Reprint Series 1, St. John’s, Newfoundland, Canada, p.53-69.

Schwartz, T.M. and Schwartz, R.K., 2013, Paleogene postcompressional intermontane basin evolution along the frontal Cordilleran fold-and-thrust belt of southwestern Montana: GSA Bulletin, v. 125, p. 961-984.

Schwartz, T.M., Methner, K., Mulch, A., Graham, S.A., and Chamberlain, C.P., 2019, Paleogene topographic and climatic evolution of the Northern Rocky Mountains from integrated sedimentary and isotopic data: GSA Bulletin, v. 131, p. 1203-1223.

Scott, G.R. and Pillmore, C.L., 1993, Geologic and structure-contour map of the Raton 30’ x 60’ Quadrangle, Colfax and Union Counties, New Mexico, and Las Animas County, Colorado: USGS Miscellaneous Investigations Series Map I-2266, 1:100,000.

Sewall, J.O. and Sloan, L.C., 2006, Come a little bit closer: A high-resolution climate study of the early Paleogene Laramide foreland: Geology, v. 34, p. 81-84.

Sharman, G.R., Covault, J.A., Stockli, D.F., Wroblewski, A.F.-J., and Bush, M.A., 2017, Early Cenozoic drainage reorganization of the United States Western Interior-Gulf of Mexico sediment routing system: Geology, v. 45, p. 187-190.

Smith, M.E., Cassel, E.J., Jicha, B.R., Singer, B.S., and Canada, A.S., 2017, Hinterland drainage closure and lake formation in response to middle Eocene Farallon slab removal, Nevada, USA: Earth and Planetary Science Letters, v. 479, 156-169.

Topper, R., Scott, K., and Watterson, N., 2011, Geologic Model of the Purgatoire River Watershed within the Raton Basin, Colorado: Colorado Geological Society, Coalbed Methane Stream Depletion Assessment Study, 53 p.

Tschudy, R.H., Pillmore, C.L., Orth, C.J., Gilmore, J.S., and Knight, J.D., 1984, Disruption of the terrestrial plant ecosystem at the Cretaceous-Tertiary boundary, Western Interior: Science, v. 225, p. 1030-1032.

Tyler, R., Kaiser, W.R., Scott, A.R., Hamilton, D.S., and Ambrose, W.A., 1995, Geologic and hydrologic assessment of natural gas from coal: Greater Green River, Piceance, Powder River, and Raton Bains, Western United States: Bureau of Economic Geology, Report of Investigations No. 228, 219 p.

Ventra, D. and Clarke, L.E., 2018, Geology and geomorphology of alluvial and fluvial fans: current progress and research perspectives: Geological Society of London Special Publications, v. 440, p. 1-21.

Walker, R.G. and Cant, D.J., 1984, Sandy Fluvial Systems, in Walker, R.G., ed., Facies Models Second Edition: Geoscience Canada, Reprint Series 1, St. John’s, Newfoundland, Canada, p. 71-89.

679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722

18

Wang, J. and Plink-Bjorklund, P., 2019, Variable-discharge-river macroforms in the Sunnyside Delta interval of the Eocene Green River Formation, Uinta Basin, USA: Sedimentology, doi: 10.1111/sed.12688.

Weissmann, G.S., Hartley, A.J., Nichols, G.J., Scuderi, L.A., Olson, M., Buehler, H., and Banteah, R., 2010, Fluvial form in modern continental sedimentary basins: Distributive fluvial systems: Geology, v. 38, p. 39-42.

Wolfe, J.A. and Upchurch, G.R., Jr., 1987, Leaf assemblages across the Cretaceous-Tertiary boundary in the Raton Basin, New Mexico and Colorado: Proceedings of the National Academy of Sciences, v. 84, p. 5096-5100.

Zellman, K.L., 2019, Investigating links between climate perturbations, river discharge variability and fluvial fans in the Paleogene San Juan Basin, New Mexico, USA [PhD thesis]: Colorado School of Mines, Golden, CO, 206 p.

TABLESTable 1. Location, grain size, and dimensional data for fluvial channels observed in this study.

FIGURESFigure 1. (A) Simplified map of the Raton basin illustrating modern extent of the Cretaceous-Paleogene basin fill (gray shading) and major structures within the basin. White circles denote outcrops observed for this study. Pollen data are adapted from REF. (B) Stratigraphic column summarizing the Upper Cretaceous-Paleogene fill of the Raton basin.

Figure 2. Stratigraphic sections measured through trunk fluvial channels of (A) the lower Raton Formation (TKr) at Lower Segundo and the Poison Canyon Formation (Tpc) at (B) Armstrong and (C) Stubblefield. Additional measured sections are provided in the Supplemental Information.

Figure 3. Photopanels and line drawing interpretations of splay-related channels in the lower Raton Formation: (A-B) Madrid West channel; (C-D) Valdez channel. Note that both channels are associated with thick successions of carbonaceous overbank deposits. Average paleoflow for the channels is toward the north-northwest. Yellow – sandstone; light gray – siltstone; dark gray – carbonaceous shale; black – coal.

Figure 4. Photopanels and line drawing interpretations of sinuous channel deposits in the lower Raton Formation: (A-B) Lower Segundo point bar complex; (C) nested erosion surfaces filled by heterolithic bedding in the upper part of Lower Segundo; (D) point bar deposits recording aggradation and translation within the sinuous channel system. Average paleoflow from Lower Segundo is toward the north-northwest. Yellow – sandstone; light gray – siltstone; dark gray – carbonaceous shale; black – coal.

Figure 5. (A) Photopanel and (B) line drawing interpretation of large-scale accretion sets in the “barren series” of the Raton Formation (Incline channel complex). Note that accretion sets show evidence of liquefaction and failure toward the east-northeast. Inferred paleoflow is toward the north. Yellow – sandstone; light gray – siltstone; black – coal; green - vegetation.

723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765

19

Figure 6. Photopanels and line drawing interpretations of sinuous channel deposits in the lower Poison Canyon Formation: (A-B) Contact channel complex; (C-D) Little Crow channel complex. Average paleoflow from Little Crow is toward the northeast. Yellow – sandstone; tan – sandy siltstone; green - vegetation.

Figure 7. Photopanels and line drawing interpretations of sinuous channel deposits in the Poison Canyon Formation: (A-B) Armstrong channel complex; (C) heterolithic channel margin deposits that show liquefaction and failure toward the channel thalweg; (D) lateral amalgamation of heterolithic accretion sets bound by erosion surfaces; (E) channel element filled by sandy siltstone beds that drape the channel form. Yellow – sandstone; tan – sandy siltstone; green - vegetation.

Figure 8. Photopanels and line drawing interpretations of sinuous channel deposits in the upper Poison Canyon Formation: (A-B) lower channel element at Stubblefield; (C-D) upper channel element at Stubblefield; (E) meter-scale, gravelly, tangential cross-strata; (F) cm-thick, aggradational planar laminations. Yellow – sandstone; tan – sandy siltstone; green - vegetation.

Figure 9. Characteristic sedimentary structures in the Raton (TKr) and Poison Canyon (Tpc) Formations: (A) organic-rich laminations extending from the margin of the Madrid West channel (TKr) toward its thalweg, person for scale; (B) undulating to low-angle laminations filling the Madrid West channel (TKr); (C) ripple to dune cross-lamination lined by dark, disseminated organic material in the Lower Segundo channel (TKr); (D) carbonaceous fern fossils from heterolithic overbank deposits associated with the Lower Segundo channel (TKr); (E) climbing ripple sets with oxidized mineral partings in the Upper Segundo channel (TKr); (F) aggradational and slightly climbing ripples from the tops of macroforms at the Incline channel complex (TKr); (G) mud rip-up clasts lining crude tangential cross-strata in the Armstrong channel complex (Tpc); (H) plant fossils from heterolithic overbank deposits associated with the Armstrong channel complex (Tpc).

Figure 10. (A) Grain size and (B) channel dimension trends through the Raton (TKr) and Poison Canyon (Tpc) Formations. Grain size distributions show that channels coarsen upward and have progressively wider grain size distributions (vf – very fine; f – fine; m – medium; c – coarse; vc – very coarse; gr – granular; peb – pebbles; cob – cobbles; -l – lower; -u – upper). Aspect ratios of trunk channels indicate that individual channel elements become shallower and wider through time.

766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798