Facies analysis and depositional model of the Midcontinent Rift System in Kansas, USA

The Midcontinent Rift System of North America is one of the oldest continental rifts but rifting ceased before continental breakup. The southern segment of the Midcontinent Rift System lies in Kansas, USA, where the stratigraphic succession and rift evolution are largely unknown. This study analysed the rift basin infill in this part of the Midcontinent Rift System to propose a depositional model. The Precambrian rift succession was described in discontinuous cores drilled in the Texaco Noel Poersch#1 well in Washington County. Sixteen lithofacies were identified and grouped into four different facies associations (fluvial, aeolian, lacustrine and alluvial fan). Overall, the studied succession comprises continental deposits accumulated dominantly in alluvial and aeolian settings, with the intermittent development of lacustrine systems. The proposed depositional models for the available core intervals indicate cyclic patterns of overfilled and underfilled phases within the rift basin. These changes in the accommodation‐to‐supply ratio were controlled by tectonism and probably modulated by climate during evolution in the syn‐rift phase. This study advances our understanding of variations across the Midcontinent Rift System.


| INTRODUCTION
The Midcontinent rift system (MRS) is a 2000 km long, U-shaped rift developed over a mantle plume (Nicholson et al., 1997) centred at the Lake Superior region in Central North America (Behrendt et al., 1990;Nicholson & Shirey, 1990;Paces & Bell, 1989) at ca 1.15 Ga.The rift was active for 25 Myr (Fairchild et al., 2017) and had a complex evolution, marked by extensive igneous activities and the subsequent formation of sedimentary basins, ultimately failing to open into a new ocean.Because it is buried beneath kilometres of Phanerozoic sedimentary rocks, there are uncertainties in the identification of the rift and its geometry (Hinze & Chandler, 2020).Studies using geophysical data such as gravimetric anomalies and other seismic investigations have demarcated the MRS (Elling et al., 2020;Hinze & Chandler, 2020;Stein et al., 2014Stein et al., , 2015)).It forms a failed triple junction that extends from Kansas to Lake Superior and across the southern peninsula of Michigan (Dickas et al., 1992;Ojakangas et al., 2001;Van Schmus & Hinze, 1985;Woodruff et al., 2020), while the third arm is suggested to extend into Ontario, Nipigon Embayment in Canada (Hinze & Chandler, 2020).
The intensity of the rifting is thought to have varied as it spread away from the centre.However, this remains conjectural due to the lack of outcrops in most places except for Lake Superior, where the Mesoproterozoic Keweenawan Supergroup is exposed and well-studied (Ojakangas & Dickas, 2002).Therefore, subsurface data are paramount, both for rift identification and to establish the connection between rift evolution and the sedimentary package.For the entire extension of the MRS, the subsurface stratigraphic succession has been described in detail in only two boreholes in the Upper Peninsula of Michigan (Ojakangas & Dickas, 2002), and roughly described in one borehole in Ohio (Dickas et al., 1992) and another in Kansas (Berendsen et al., 1988).More stratigraphic information is needed to advance our understanding of rift development and evolution, especially regarding thermal and tectonic variations in time and space.Hence, any analysis of the stratigraphic succession across the MRS is valuable.
Geophysical and borehole data have identified the southernmost segment of the MRS in north-central Kansas, USA, with a different rift development history proposed for this region compared to the central segment in Lake Superior (McSwiggen et al., 1987).That proposal is speculative, however, since there are no studies of rift evolution through time in Kansas that allows the various evolutionary stages to be integrated into a rift evolutionary model, in contrast to the Lake Superior region where more data are available (Woodruff et al., 2020).
An earlier study by Berendsen et al. (1988) identified two distinct successions in the Texaco Noel Poersch#1 (NP#1) well which is the only well to reach the MRS in Kansas.Berendsen et al. (1988) identified a lower sedimentarydominated succession (11,300-7429 ft/3444.2-2264.5 m deep) and upper volcanic-dominated succession (7429-2846 ft/ 2264.5-867.5 m deep) through wireline log data and well cuttings, reflecting a radical change during rift evolution.This study aims to provide information on rift sedimentation and depositional environments in the Kansas rift segment through facies analysis of the cores from NP#1 (Figure 1), particularly in the lower sedimentary succession.

| Geological Setting
The presence of the MRS segment in Kansas was identified through geophysical data that reported a positive gravity anomaly (Figure 1B), showing strong spatial correlation with the Mesoproterozoic Keweenawan rocks from Lake Superior (Thiel, 1956;Woollard, 1943).
However, the breakthrough in the identification and modelling of the rift was provided by Serpa et al. (1984), analysing the seismic data from the Consortium for Continental Reflection Profiling (COCORP) lines across north-eastern Kansas.Their study revealed a 40 km wide asymmetrical rift basin plunging to the west bounded by easterly dipping normal faults, which caused 29 km of crustal spreading (Serpa et al., 1984) (Figure 2).Based on the border faults and rift geometry, the MRS in Kansas is interpreted as a half-graben rift basin (Berendsen, 1997).McSwiggen et al. (1987) proposed that this southern terminus of the MRS might not have developed to the same extent as elsewhere in the MRS, possibly representing an arrested stage of rift development, similar to Bosworth's (1987) model for the East African Gregory Rift.This speculation was mostly based on the rift geometry and the relatively thin volcanic layers in the rift basins compared to northern MRS segments.The resulting marginal, half-graben rift basin worked as a depocentre for the rift volcanics and subsequent sedimentation.The scarcity and poor quality of seismic data hinder any attempt to detailed interpretations.
Rift geometry is dictated by varying rifting mechanisms and plays a vital role in rift sedimentation (Withjack et al., 2002).The faults and fault-related topography control the sedimentary systems within the rift basin by controlling accommodation (Barr, 1987;Leeder & Gawthorpe, 1987), that is, the space available for potential sediment accumulation (Jervey, 1988), and sediment transport and deposition (Withjack et al., 2002).Thus, the basin fill can be classified according to the different stages of tectonostratigraphic evolution, including the pre-rift, syn-rift and post-rift stages (Rapozo et al., 2021;Strugale & Cartwright, 2022; see discussion below).Despite the overall understanding of rift basin evolution, there are still some debatable issues regarding rift propagation (e.g.mantle plume vs. tectonic forces) and the controls on rift sedimentation (Holz et al., 2017).Sedimentological data from rift successions in different segments may offer insights about rift evolution.Hence, despite the discontinuous nature of the data in this study, it provides an opportunity to build onto the existing knowledge of the MRS.Bosence (1998) proposed a model in which two unconformities, the syn-rift and the post-rift unconformities, delimit three rift basin successions (pre-rift, syn-rift and post-rift).According to this model, the onset of rifting triggers a complex fault-related topography that guides sediment accumulation and controls the erosional and sedimentological processes.Igneous rocks can dominate this phase (pre-rift), and the sediments deposited during this phase are controlled mainly by the rift basin architecture (Berendsen, 1997).At this stage, the extended crust undergoes regional subsidence; the footwall crests are uplifted and eroded, becoming the source of the sediments, and resulting in the formation of the syn-rift unconformity that separates the pre-rift and syn-rift strata.As the rifting advances, a complex drainage basin and the changing relationship with the structures control the thickness of the syn-rift strata (Bosence, 1998).The synrift strata mark the main rift phase and are characterised by an onlap surface generated by tilting and rotation of structural blocks.Bosence's (1998) model proposes that syn-rift sedimentation fills the graben-generated accommodation in rift basins and usually contains deltaic and fluvial deposits.
The post-rift unconformity separates the syn-rift units from the overlying post-rift succession.It develops near the rift shoulders during a phase dominated by thermal subsidence, controlled by lithospheric cooling and an increase in the density of the lithosphere and asthenosphere, augmented by sediments and water loading (Bott, 1992).The post-rift strata are characterised by onlap and offlap, with surfaces dipping into the enlarging accommodation in the basin centre (Bosence, 1998), as the basin slowly evolves into either a passive oceanic basin margin or, in the case of MRS, a failed rift (Stein et al., 2018).
Due to its progressive and diachronous nature, rift basin evolution can differ in both time and space (Holz et al., 2017).Therefore, Bosence's (1998) model cannot be applied to the entire basin over the same time frame.Several authors have considered the rifting process as a function of two fundamental stages: a mechanical subsidence phase that creates accommodation and a phase of tectonic quiescence when the available accommodation is filled (Martins-Neto & Catuneanu, 2010).Holz et al. (2017) proposed an integrative sequence stratigraphic model for rift basins with three tectonic phases.A local fault characterises the rift initiation phase, with an incipient half-graben filled by fluvial and deltaic deposits, followed by a rift development phase that connects the faults and enlarges the depositional area, with sediment accumulation in deep lakes showing a retrogradational stacking pattern.Finally, the rift fill-up phase is marked by declining accommodation and basin infill by aeolian, fluvial and deltaic deposits, forming sedimentary successions with a progradational trend.
Since the MRS succession crops out only in the Lake Superior region, this area has been studied in detail  Van Schmus & Hinze, 1985;Woodruff et al., 2020).In general, the rift succession of the northern rift segments includes syn-rift and post-rift successions overlying the Precambrian crystalline basement.Ojakangas et al. (2001) proposed four sequences of sedimentary rocks reflecting the tectono-sedimentary framework of MRS in the Lake Superior region.Sequence 1, composed of quartzose sandstone, marks the onset of rifting and associated volcanism (Ojakangas & Morey, 1982).Sequence 2, otherwise known as the syn-rift succession, comprises lava flows (loosely termed 'Portage Lake Lava Formation'), interbedded with a 30 m thick succession of sedimentary rocks (known as the 'interflow sequence') deposited during volcanic quiescence (Merk & Jirsa, 1982).These are interpreted to have been deposited in a fluvial and lacustrine environment (Ojakangas & Dickas, 2002).A thick, Mesoproterozoic, post-rift unit of about 10 km in the Lake Superior region (Ojakangas & Dickas, 2002) overlies the syn-rift unit, which includes the Oronto Group (Sequence 3) and the Bayfield Group (Sequence 4).The Oronto Group, comprising the Copper Harbour Conglomerates, Nonesuch Formation and Freda Sandstone, primarily consists of red-bed, alluvial fan deposits, indicative of an oxidising environment.The Nonesuch Formation, characterised by grey-black siltstone, shale and sandstone suggests a contrasting reducing environment, potentially associated with a lake or marine setting in proximity to the continental red bed deposits above and below it, as supported by geochemical evidence (Ojakangas et al., 2001).The texturally and mineralogically more mature Bayfield Group is composed of the Orienta Sandstone, Devils Island Sandstone and the Chequamegon Sandstone.It represents an overall fluvial depositional environment punctuated by a lacustrine setting during the deposition of the Devils Island Sandstone (Adamson, 1997;Ojakangas & Dicaks, 2002).
In contrast to the data available for the rift succession in Lake Superior, the sedimentary fill of the rift in Kansas, loosely called the Rice Formation (Berendsen et al., 1988;Berendsen, 1997), is yet to be studied in detail.The Rice Formation is composed of reddish-brown, finegrained feldspathic sandstone, commonly conglomeratic (Berendsen, 1997).Scott (1966) suggested a marine or lacustrine depositional environment for this formation, as it contains limestone and dolomite interbedded with sandstone and shale.Although it is placed in the Keweenawan Supergroup by many, it lacks a proper radiometric age (Sawin et al., 2013) and detailed analysis to indicate its depositional characteristics related to rift evolution.

| Data source
The data for this study were obtained from the cores retrieved in Texaco Noel Poersch#1 (NP#1; Figure 1), drilled in the eastern flank of the geophysical anomaly that defines the rift in Kansas (Figures 1B and 2A) and on a basement high associated with the post-Mississippian Nemaha Uplift (Berendsen et al., 1988).Despite rock data being discontinuous and very sparse, these cores provide valuable insights into the rift fill, since they are the only ones available for the rift segment in Kansas (and some of the few for the entire MRS), allowing the characterisation of depositional environments formed during rift evolution.
Located in Washington County, NP#1 was drilled in 1984 as a wildcat well in an anticline recognised from seismic data.The COCORP seismic profile (Serpa et al., 1984) allows this well to be positioned on the flank of a basement high (Figure 2B).NP#1 is the deepest well ever drilled in Kansas, with a total depth of 11,300 ft (3444 m).The top of the Precambrian is at a depth of 2846 ft (1477 m).In the preliminary report produced by the Kansas Geological Survey, well cuttings were analysed, with the rock succession determined to comprise syn-rift deposits (Berendsen et al., 1988) based on correlation with the COCORP seismic profile and the textural and compositional immaturity of the deposits.The report concluded that the rift segment in Kansas is younger than other segments of the MRS to the north based on the much younger age (1021 ± 100 Ma) obtained from the volcanics in NP #1 (Berendsen et al., 1988).
The Precambrian stratigraphy in Kansas is divided solely according to the Eonothem/Eon (Archean and Proterozoic) and Erathem/Era (Palaeoproterozoic, Mesoproterozoic and Neoproterozoic) names (Sawin et al., 2013).The 'Rice Formation' was improperly defined and abandoned.Hence there is no stratigraphic column for the studied succession; the only subdivision comes from the previous study of NP#1 (Berendsen et al., 1988).Berendsen et al. (1988) identified two successions: the upper one (down to depths of 7429 ft/2264.4m) is dominantly composed of volcanic rocks, with sparse sedimentary interbeds, whereas sedimentary rocks dominate the lower succession (at depths of 7429-11,300 ft/2264.4-3444.2m), and thus was the focus of this study (Figure 3).The studied section, comprising a total of 66.16 ft (20 m) of cores, range from 5395 ft to 11,296 ft (1644.4 to 3443 m) meaning nearly 99% of the sedimentary sequence is missing.Despite this limitation, the analysis of the existing cores provides a glimpse of the depositional environments developed during rift evolution.
The wireline log data of NP#1 suggest two periods of igneous activity within the lower sedimentary-dominated succession (Figure 3) at depths of 10,840-10,622 ft (3304-3237.6 m) and 8810-8676 ft (2685.3-2644.4m).These volcanic intervals were used to divide the lower, sedimentary-dominated succession into three distinct intervals: sections A, B and C. On the other hand, the upper volcanic-dominated succession has some thin interbedded sedimentary units that reflect periods of quiescence (Berendsen et al., 1988).Three cores retrieved from such depths were studied for this research and were categorised as sections D and E (Figure 3).A total of 11 core intervals (CI) were studied for this research (Figure 3).

| METHODS
A detailed sedimentological description (scale = 1:30.5) of these cores for facies identification included lithology, texture, composition, sedimentary structures and diagenetic features.They are displayed as a composite sedimentological log (Supplemental Material).The concept of facies is that of Reading (1996), and the facies classification scheme used here was adapted from that proposed by Miall (1985) for fluvial systems.The initial letter, capitalised, denotes grain-size classification (e.g.G for gravel, S for sand and F for fine-grained sediments), while subsequent lowercase letters signify texture and/or sedimentary structures (e.g.t for trough crossstratification).The identified facies were systematically arranged in a lithofacies table, accompanied by illustrative plates for each lithofacies.These were then categorised into facies associations, representing a cluster of genetically linked facies indicative of sub-environments within a depositional system (Collinson, 1996;James & Dalrymple, 2010).Optical petrography of the different facies was carried out to augment the macroscopic description (Azmi, 2020).
Wireline data were used to distinguish igneous from sedimentary rocks, based on their gamma-ray and density signatures.Igneous rocks are characterised by low gamma ray (<50 API units) and high density (>2.8 g/cm 3 ), while sedimentary rocks display high gamma ray (>100 API units) and low density (ca2.6 g/cm 3 ).

| Facies analysis
Sixteen lithofacies were identified in the 66.2 ft of core collected from NP#1, summarised in Table 1.The most common facies are illustrated in Figure 4. See Supplemental Materials for details.
The studied section contains mostly sandstones, with minor conglomerates and mudstones.Conglomerates are mostly clast-supported, lithic conglomerates with subrounded pebbles in a medium sand matrix (Figure 4A), but subordinate intraformational conglomerates with mud intraclasts also occur.Texturally, the sandstones are very fine to coarse-grained (mostly fine to mediumgrained), poorly to well sorted, with sub-angular to subrounded clasts (Figure 4B through F).Normal and inverse grading was observed, with occasional millimetre-scale, horizontal and inclined calcite veins (Figure 4E).The mudstones are commonly massive, with fenestral and discontinuous clay laminations (Figure 4G).
The most abundant facies in the analysed cores is the sandstone facies (S), compared to minor mudstones (F) and conglomerates (G).The most abundant facies is Sle, followed by Sp, Sm and Sc (Figure 5).However, facies distribution is not homogeneous among the different sections.The bottom two sections, A and B, are devoid of any conglomerate and mudstone facies.The top two sections F I G U R E 3 Wireline log of Noel Poersch#1 in Kansas, showing the 11 core intervals used in this study (modified from Berendsen et al., 1988).

Lower Sedimentary Dominated Succession
Upper Volcanic Dominated Succession  Suspension settling from dominantly standing water, with plastic deformation of semi-consolidated sediments due to re-sedimentation and/or fluidisation from the volcanic-dominated succession (sections D and E) are dominated by conglomerate and mudstone facies, with subordinate sandstone facies (Figure 5).The lithofacies identified within the studied succession reflect different processes responsible for the deposition of that unit (Table 1).The conglomerate facies (Gcm and

| Facies association
The lithofacies identified in the cores were grouped into four facies associations: aeolian, fluvial, lacustrine and alluvial fan facies associations.The description and interpretation of each facies association are displayed in Figure 6, and their vertical distribution throughout NP#1 in Figure 7.These facies associations are not limited by stratigraphic surfaces, and they may be interbedded with each other (e.g.aeolian and fluvial facies associations) or isolated in a core interval (e.g.lacustrine facies association in core interval 6).

| Aeolian facies association
The aeolian facies association comprises 4.6 m thick successions composed of moderately to well-sorted sandstones, mostly with low-angle cross-stratification and inverse grading (Sle) interbedded with horizontally-stratified and planar cross-stratified sandstones with inverse grading (She and Spe, respectively) (Figure 6).The reddish-brown colour, presence of pinstripe lamination, inversely-graded wind ripple lamination, grain-flow and grain-fall strata,  (Fryberger & Schenk, 1988;Hunter, 1977;Scherer & Lavina, 2005).The alternations between Sle and She mark fluctuations in wind velocities between the transitional and upper-flow regimes.The interbedding between these facies is typical of sand sheet deposits adjacent to dune fields (Boggs, 2012;Scherer & Lavina, 2005).

| Fluvial facies association
The fluvial facies association comprises 0.3 to 3 m thick successions, composed of coarse to fine sandstones, poorly to moderately sorted, with planar cross-stratification (Sp), horizontal lamination (Sh) and/or massive (Sm), occasionally ripple cross lamination (Sr) and low-angle crossstratification (Sl) (Figure 6).This facies association occurs only in the lower sedimentary-dominated succession (Figure 5).
The abundance of tractive structures, poor sorting and overall fining-upward successions suggests a channel deposit (Miall, 2014(Miall, , 1977;;Scherer et al., 2007).presence of planar cross-stratification indicates sedimentation by currents under a lower flow regime, while low-angle crossstratification and horizontal lamination indicate higher energy (transitional to upper flow regimes) in the depositional environment (Miall, 2006).High-velocity, intermittent discharge leads to the preferential deposition of transitional to upper-flow regime beds, with overbank fines rare to absent (Miall, 2006).The absence of scour surfaces at the base of the successions, coupled with variations in flow regimes represented by facies Sh and Sr, suggests the prevalence of poorly channelised flows.Vertical successions featuring upper flow structures (Sh) and critical to supercritical flow conditions (Sl) at the base, grading upward to lower flow structures (Sp) indicative of waning flows (Miall, 2006), further support this interpretation.The presence of horizontal lamination, ripple cross lamination and indications of waning flow collectively suggest ephemeral river deposits (Bhattacharyya & Morad, 1993;Davies & Gibling, 2010;Picard & High, 1973), probably formed by flash-flood discharge during flooding events.This fluvial morphology is commonly associated with pre-vegetation times owing to the absence of rooted vascular plants to control bank stability (Long, 2006).

| Lacustrine facies association
This facies association comprises a 3.4 m thick sand body with crinkly lamination (Sc), interbedded with massive mudrocks (Fm), and subordinate rippled, fine sandstones with asymmetrical and symmetrical ripple cross laminations (Sr and Sw), and 0.6 to 1.3 m thick successions of laminated and massive siltstones (Fl and Fm), alternating with fenestral laminites (Ff) (Figure 6).The first succession occurs only in section B, and the second at the top of F I G U R E 6 Description and interpretation of the facies associations identified in the studied cores (not to scale).

F I G U R E 7
Depositional model for different sections of NP#1, based on the facies associations identified from the description of the available core intervals (CI).Symbols for lithology are the same as in Figure 6.The detailed core description is available in Supplemental Materials.the lower sedimentary-dominated succession and in the upper volcanic-dominated successions, in sections C and D (Figure 5).The dominance of silt grain size, with minor mud fractions, indicates the low energy of the depositional environment (Rogers & Astin, 1991).Crinkly laminations formed by microbial activity (Schieber et al., 2007) with interbedded Sw indicate fluctuations in the water level, with periodic displacement of the microbial mats by subaqueous, wave-reworked environments.The lack of evaporites and/or mud cracks and other structures indicative of subaerial exposure suggests a shallow, open lake with perennially wet margins.The fenestral laminites are interpreted as resulting from microbial activity in a lacustrine environment (Suarez-Gonzalez et al., 2019), where microbes bound and trapped the finer material onto microbial mats.Later partial desiccation formed the fenestrae, filled by cement during diagenesis (Demicco & Hardie, 1994).
The dominance of conglomerate facies fining up to massive sandstone suggests debris flow dominated fan deposits built primarily of sheet flow and hyperconcentrated flow deposits (Miall, 2006).The poor sorting and lack of stratification point to an en masse depositional process (Went, 2005).The fine-grained intervals were deposited by suspension settling in a standing water body (Fl) where synsedimentary deformations occurred (Fd).Altogether, interbedding of these high-energy, proximal, gravity-flow deposits with fine-grained, subaqueous deposits suggests a proximal fan delta environment (Blair & McPherson, 2008;McPherson et al., 1987;Benvenuti, 2003;Tamrakar et al., 2009) where the conglomeratic facies represents subaerial debris flow deposits, as has been shown in both modern environments (Blair & McPherson, 1998) and ancient successions (Benvenuti, 2003).Even though the facies association resembles those of modern-day, arid fans, climate reconstruction in pre-vegetation successions is difficult because the lack of land plants may make continental environments appear (Went, 2005).The palaeoclimate context will be discussed below.

| DISCUSSION
Facies analysis of the cores from the Noel Poersch#1 well shows that these sedimentary rocks accumulated in four different depositional environments, based on the facies associations described and illustrated in Figures 6 and  7.The discontinuous distribution of cores precludes the full understanding of rift evolution, but nonetheless the stratigraphic succession of the facies associations provides the basis for a rough evolutionary depositional model.Wireline log data proved useful to differentiate between sedimentary and igneous rocks (although they lacked resolution to recognise facies associations).Based on the modern analogue provided by the East African Rift System, where extension is facilitated and accommodated by magmatic activity (Corti, 2009), intervals dominated by igneous rocks in the studied well were interpreted as indicative of rift reactivation through extensional efforts.As such, rift evolution was characterised by distinct phases of extension punctuated by periods of relative quiescence, during which the sedimentary succession accumulated.
Due to paucity of data, the construction of the depositional model depicted in Figure 7, and subsequent discussions about the controls acting on the sedimentary succession, relied heavily on models for well-studied rift basins (Scherer et al., 2007;Holz et al., 2017;Corti et al., 2018), or comparisons with the better-known MRS segment along Lake Superior (DeGraff & Carter, 2023).
The seismic data for the MRS segment in Kansas (Serpa et al., 1984) outlined the existence of half-grabens like modern rift basins, with NP#1 located on the flank of a structural high (Figure 2B).Hence, it is reasonable to assume that the MRS basin in Kansas included both axial and transverse systems.Transverse systems develop along elevated hangingwall blocks, favouring the development of the alluvial fans common in rift basins (Gawthorpe & Leeder, 2000).
Interpretations of climate signals are admittedly speculative due to the difficulty posed by the limited data and the ancient age of the studied succession (accumulated before the evolution of metazoans and land plants), but some inferences can be made.Palaeogeographical reconstructions position the MRS in the centre of supercontinent Rodinia (Stein et al., 2018), which lay at (or very near) the palaeoequator (Swanson-Hysell, 2021).These conditions suggest that the climate was probably hot and arid, which would explain the abundance of aeolian deposits and the dominance of unconfined, ephemeral fluvial flows.
Following the rift stratigraphic framework proposed by Martins-Neto and Catuneanu (2010), the changes in depositional environments (represented by the facies associations) were interpreted in terms of changes in the accommodation-to-sediment supply (A/S) ratios through time.However, it is still possible that the vertical variations (interpreted here as temporal variations) might actually be lateral variations in the environment.Nonetheless, the stratigraphic analysis of the sedimentary succession of the MRS in Kansas provided some hints on rift evolution.
Figure 7 includes a depositional model for sections A, B, C, D and E (from base to top).Section A contains two core intervals.Core interval 1 (11,296-11,290 ft/3443-3441.2m) displays alternating aeolian and fluvial facies associations, followed by the fluvial facies association dominating core interval 2 (11,068-11,061 ft/3373.5-3371.4m), covered by a 66.5 m thick basalt (identified from the gamma-ray log; Figure 3).It is possible that section A was deposited by axial fluvial systems (Leeder & Gawthorpe, 1987) within the half-graben basin formed by rifting (Figure 7).The Rb-Sr ages obtained in the overlying basalt (1021 ± 100 Ma) indicate deposition in the Proterozoic (Berendsen et al., 1988).Proterozoic rivers were braided systems, as a consequence of the lack of vegetation and poor soil development to provide bank stability (Bose et al., 2012;Eriksson et al., 1998;Sønderholm & Tirsgaard, 1998).This agrees with the absence of mudrocks (both in the cores and un-cored intervals, as identified from wireline log data) and dominance of coarse to medium sized sand grain in section A, typical of bedload-dominated, braided river deposits (Miall, 1977(Miall, , 1978)).The fluvial system is interpreted as having developed along the axial zone on the downfaulted block, as suggested for asymmetrical half-graben basins, where channels tend to occupy the axis of maximum subsidence (Bridge & Leeder, 1979).Laterally adjacent to it, wind reworking allowed the development of dune fields.Avulsion of the axial river is a common phenomenon in continental half-grabens due to episodic tectonic tilting (Leeder & Gawthorpe, 1987).Rearrangement of tectonic blocks in the rift basin may result in lateral shifts of the fluvial system, resulting in interbedded fluvial and aeolian deposits (Figure 7).Alternatively, rather than concomitant with the fluvial system, the dune fields could have developed intermittently in response to wind reworking of ephemeral fluvial sediments during dry periods.This stage of evolution in NP#1 ended with volcanism that resulted in the extrusion of flood basalts.
The vertical facies succession suggests that section B contains core intervals that record at least two different depositional systems, from base to top, sections B-1 and B-2, respectively.Section B-1 (10,515-10,504 ft/3205-3201.6m) contains mostly lacustrine deposits, with minor interbeds of aeolian and fluvial deposits.The depositional model shows back-stepping of the fluvial systems and expansion of a lake (Figure 7).Based on the position of NP#1 in the rift structure (on a basement high) (Figure 2B), this change was probably due to an increase in the A/S ratios.
Following a gap of about 164.3 m above section B-1, section B-2 (9965-9950 ft/3037.3-3032.7 m) is characterised by the return of fluvial and aeolian systems in the basin (Figure 7).Following a gap of 223 m between core intervals 4 and 5, the aeolian systems dominated the top of section B-2 (9170-9160 ft/ 2795-2792 m).The end of section B was marked by another intense episode of igneous activity, resulting in the accumulation of 40.8 m of basalt.
Section C comprises lacustrine deposits (in core interval 6), followed by a gap of 150 m and fluvial-aeolian deposits (in core intervals 7 and 8).The depositional model focusses on the dominance of lacustrine systems in the rift basin (Figure 7), but the vertical succession in section C is like the one observed in section B, in which lacustrine systems are followed by fluvial-aeolian systems.
The overlying units (sections D and E) are part of the upper volcanic-dominated succession.Following a gap of 124.2 m, the presence of basalt at the base of core interval 9 points to renewed igneous activity within the basin in section D. Overlying the basalt, alluvial fan deposits indicate the progradation of proximal systems (Figure 7).Following a gap of 532.2 m, section E includes alluvial fan deposits at the base (core interval 10).A period of intense igneous activity was responsible for the extrusion of basalts, and lacustrine deposits (core interval 11) cap the studied succession.
The temporal succession of depositional models proposed above indicates a cyclic pattern of facies associations within the Midcontinent Rift basin in Kansas.Interpretations based on discontinuous and sparse core data preclude the identification of the rift sequences.Nonetheless, in the few segments where rock data are available, the basin infill might still be analysed in terms of the balance between accommodation (A) and sediment supply (S), or A/S ratios related to different systems tracts (Martins-Neto & Catuneanu, 2010).The cores in NP#1 record alternating underfilled phases, characterised by high A/S ratios (e.g.core intervals 3, 6, 10 and 11), and overfilled phases in which sediment supply exceeds accommodation, that is, low A/S ratios (e.g.core intervals 1-2, 4, 5, 7, 8 and 9).These phases might be related to the episodic nature of tectonism in the rift, where increased subsidence leads to greater accommodation, whereas the reactivation of rift shoulders leads to increased sediment supply.
The stratigraphic succession can be controlled by three allogenic processes: relative seal level, climate and tectonics.The studied section was deposited in a continental context; hence sea level was probably not an important control.The available data did not provide enough evidence of significant climate change, and the low palaeolatitutidal position of the MRS would favour a more equitable climate.Thus, tectonic movements controlled the A/S ratio during deposition of the studied rift succession.However, as in many rifts across the world, tectonic control was probably modulated by climate.While subsidence and uplift either create or destroy accommodation, the pathways for bringing sediments into the basin and the climate that influences weathering, sediment transport and facies are equally important in the evolutionary history of the rift (Nichols & Uttamo, 2005).The strong tectonic control of the sedimentary succession in NP#1 implies it was probably accumulated during the syn-rift stage.

| CONCLUSIONS
This study proposes an evolutionary depositional model for the rift section of the MRS in Kansas, based on facies analysis of cores from the Noel Poersch #1 well.This is the first-ever detailed description with photodocumentation of the rift deposits, supporting the facies analysis, interpretation of sedimentary processes and preliminary discussion of rift evolution in Kansas.
Overall, the studied succession comprises continental deposits accumulated dominantly in alluvial and aeolian settings, with the intermittent development of lacustrine and peri-lacustrine systems.Despite the scarcity of rock data, analysis of these cores provided an insight into the rift fill and allowed the characterisation of depositional environments during rift evolution.The definition of a stratigraphic framework, however, was limited by the existing rock data.Nonetheless, it was possible to identify changes in the A/S ratio, probably related to tectonic pulses that created accommodation, followed by infilling of that accommodation.This suggests that the studied succession is expected to have accumulated during the syn-rift stage, contrary to the post-rift interpretation provided by Ojakangas and Dickas (2002).The younger ages of the Kansas succession when compared to the syn-rift stage in the Lake Superior region suggests non-coeval rift development.Although further insights into the role of climate and tectonics were hindered by data availability, this work advances our understanding of the variations across the MRS.
as a source of the sedimentary fill of the rift basin (DeGraff & Carter, 2023; Ojakangas & Dickas, 2002; F I G U R E 1 (A) Location of Midcontinent Rift System showing the rift segment in Kansas (red rectangle-inset).Map extracted from https:// geona rrati ve.usgs.gov/ mrs_ miner al_ depos its/ (B) The inset displays the Bouger gravity anomaly map with the studied well, Noel Poersch #1 (NP#1), on the eastern flank of the gravity anomaly in Kansas.

F
I G U R E 2 (A) Distribution of Precambrian rocks in north-eastern Kansas (afterBickford et al., 1981), with the approximate location of NP#1 (white star).(B) Geological model for the MRS basin structure in Kansas (afterSerpa et al., 1984) along the A-A' transect depicted in Figure1B.The approximate position of NP#1 is indicated with a black line.
Gci) are bedload deposits transported en masse by strong currents.All but one of the sandstone facies indicate deposition in a unidirectional flow that varied from lower to upper flow regimes.The symmetrical ripple, crosslaminated sandstone (Sw) indicates deposition under oscillatory flow conditions.The mudstone facies settled F I G U R E 4 Main lithofacies identified within NP#1 cores (core surface is wet with water for better clarity).(A) Clast-supported massive conglomerate (Gcm).(B) Sandstone with crinkly lamination (Sc, red arrows).(C) angle cross-laminated sandstone with inverse grading and millimetrescale sub-vertical epidote vein (Sle).(D) Horizontally laminated sandstone with inverse grading (She).(E) Massive sandstone with millimetre-scale calcite veins (Sm).(F) Planar cross-laminated sandstone (Sp).(G) Mudstone with fenestrae (Ff).Scale bar = 1 inch (2.54 cm).standing water, either with or without associated microbial activity.

F I G U R E 5
Facies distribution within the entire core (pie chart at the top) and in the different sections of NP#1.Fd Mudstone with soft-sediment deformation structures Sl Sandstone with low-angle cross strati cation Sl(e) Sandstone with low-angle cross strati cation and inverse grading Sp Cross-strati ed sandstone Sm Massive sandstone Gci Intraformational conglomerate Sh Sandstone with horizontal lamination Sh(e) Sandstone with horizontal lamination and inverse grading Key Gcm Clast-supported conglomerate and the well-sorted nature of the sand bodies indicate an aeolian deposit Summary of lithofacies observed in the cores collected from Texaco Noel Poersch#1.
FmReddish-brown, massive mudstone, locally with millimetre-scale faint lamination marked by very fine sand laminae; haematite staining Suspension settling from standing water; low depositional energy Fd Reddish-brown mudstone with convolute lamination, irregular silt streaks, discontinuous millimetre-scale clay laminae and faint climbing ripples