Regionally, Colombia is in an active tectonic margin due to the interaction of the South American, Caribbean, and Nazca plates (Colleta et al., 1990; Dengo and Covey, 1993; Cooper et al., 1995a). This interplay caused the uplift of the three ranges (Western, Eastern, and Central) and also the separation of a Late Cretaceous–Paleogene regional basin into multiple basins (Gómez et al., 2003; Gómez et al., 2005a; Horton et al., 2010a; Parra et al., 2010; Saylor et al., 2012). Those are two of the main sedimentary basins in Colombia as they extend for hundreds of kilometers from south to north (Figure 1). The regional basin started as an extensional basin during the Jurassic; it was under a continental rift during the earliest Cretaceous, became a retroarc basin by the Late Cretaceous, and was separated by the uplift of the Eastern Cordillera during the Paleogene (Pindell and Erikson, 1994; Sarmiento, 2001; Horton et al., 2010b; Nie et al., 2010; Cediel et al., 2011; Bayona et al., 2013, 2020; Zapata et al., 2019). As a result, there are 140 My of sediment that preserves a complex structural evolution.

Colleta et al. (1990), Cooper et al. (1995a), Dengo and Covey (1993), and Villamill (1999) promoted regional geological models demonstrating how the tectonic plates interacted throughout time and how they produced the exhumation of the Colombian Cordilleras. Cortés (2006), Horton et al. (2010a), Mora et al. (2006), Mora et al. (2008), Mora et al. (2010a), Parra et al. (2010), and Sarmiento (2001) calibrated the timing of uplift of the Eastern Cordillera and presented some analysis of the sedimentology and basin-fill history during the Cretaceous and part of the Cenozoic. More recently, these works have been summarized and complemented in a series of paleogeographic maps and geological models presented by Bayona (2018), Bayona et al. (2020), Cardona et al. (2018), Cediel et al. (2011), Martínez and Roncancio (2011), Montes et al. (2019), Reyes-Harker et al. (2015), and Sarmiento (2011).

Regarding the Late Cretaceous to Paleogene geological model, Gómez et al. (2005b) and Parra et al. (2010) stated that the foredeep of the regional foreland basin extended from the Magdalena Valley in the west to the Llanos Foothills in the east, with most of the Llanos Basin being part of the forebulge. However, Bayona et al. (2003), Caballero (2010), and Terraza-Melo et al. (2002) indicated a gap in the sedimentary record, at least between the late Cretaceous to early-middle Paleocene. The data demonstrate the exhumation and uplifting of the western flank of the Eastern Cordillera from the latest Cretaceous to Paleogene and raise the question of how it was produced? Is that related to the forebulge located at the western flank of the Eastern Cordillera instead of the western Llanos Basin?

This study presents chronostratigraphic correlations using previous data from 14 locations across the Middle-Upper Magdalena Valley, Eastern Cordillera, and western Llanos Basin, including four new stratigraphic sections from the Llanos Foothills and Basin (Figure 1 and Table 1). The age of the strata in the new sections was established following the U-Pb ages of detrital zircons from Carvajal-Torres et al. (2019) and paleontological reports from the area. The correlations presented in Figure 1 were built following the methodology from Qayyum et al. (2017) and Wheeler (1959), along with results and analysis exposed by Carvajal-Torres et al. (2021). The interpretation of sedimentary environments was simplified based on the principles presented by Boggs Jr (2006), Nichols (2009), and Perry and Taylor (2007). We correlated the depositional environments about their position from the shoreline, regardless of the type of sedimentary shore (Figure 2B).

The delimitation of the forebulge was supported using paleogeographic palinspatically restored maps according to Bayona (2018), Cardona et al. (2018), and Mann et al. (2006). Its definition is relevant to constraining the geological model of the regional foreland basin. Likewise, understanding the forebulge unconformity is key to understanding how the tectonism controlled the accommodation and sedimentation throughout time. Those concepts can be applied to other analogous basins. Our results recognize the changes in the accommodation and sedimentation due to the first-order changes. Following Catuneanu (2019a, 2020) and Catuneanu and Zecchin (2017), we identified different stratigraphic surfaces. Two of them formed during the latest Campanian to Maastrichtian–Paleocene were classified as subaerial unconformities. They are located in the western flank of the Eastern Cordillera and the Western Llanos Basin, respectively.

These unconformities were evaluated as potential forebulge unconformities. The implications of its location should be reflected in the preserved sedimentary thickness. In the first scenario, where the forebulge is located along the western flank of the Eastern Cordillera, most of the Eastern Cordillera and westernmost Llanos Basin would comprise the backbulge. Therefore, the sedimentary thickness of the late Campanian to Paleogene succession should be lower than the sedimentary thickness preserved in the foredeep, the Upper and Middle Magdalena Valley Basin. Otherwise, the thickest sedimentary record should be conserved across the Upper and Middle Magdalena Valley and Eastern Cordillera basins.

Once the forebulge unconformity was located, its connection to the depocenter was established. Its association to a maximum regressive surface comprises the first-order compound surface formed when the retroarc foreland basin started. Likewise, a regional subaerial unconformity truncating the Paleocene–Eocene strata was determined as the upper first-order boundary of the sequence as it formed during the final stage of the regional foreland basin.

The study area is located in the central part of Colombia (Figure 1). The present-day configuration is a hinterland basin (Upper-Magdalena Valley) in the west, and a foreland basin in the east (Llanos Basin), separated by the Eastern Cordillera. Two fault systems delimit the latter, the Salina Fault system in the west and the Guaicaramo Fault system in the east (Figure 1). According to Colleta et al. (1990), Cooper et al. (1995b), and Horton et al. (2010b), the study area was part of a regional basin from the Jurassic until the Paleogene, when the exhumation of the Eastern Cordillera separated it.

We reviewed previous studies from 14 locations across the Upper and Middle Magdalena Valley and Eastern Cordillera. We incorporated new data from the eastern flank of the Eastern Cordillera and the western Llanos Basin (Figure 1). The Sagu and the Wells 1–3 were selected because they were in the marginal part of the late Campanian to Paleogene retroarc foreland basin. Therefore, their correlation with the other sections across the study area will support understanding the basin architecture throughout time. Likewise, the periods of sedimentation and subaerial exposure preserved in their rock successions are crucial for understanding the role of regional tectonism over sedimentation and accommodation rates.

Following Carvajal Torres et al. (2021), Catuneanu (2017, 2019a), Catuneanu et al. (2011), and MacEachern et al. (2012), we identified some stratigraphic surfaces in the strata from the Sagu and the wells. The first surfaces identified were those related to the following unconformable contact cases: 1) FA2 (lower delta front) overlaying FA1 (prodelta), 2) FA2 (lower delta front) overlaying FA5 (continental proximal fluvial), 3) FA5 (continental proximal fluvial) overlaying FA 2 (lower delta front), 4) FA5 (continental proximal fluvial) overlaying FA4 (delta plain), 5) FA5 (continental proximal fluvial) overlaying FA 6 (marginal fluvial).

Accordingly, either fluvial or marine facies overlies the unconformable contacts. In the case of fluvial facies on top of the unconformable contact, this was defined as the subaerial unconformity (SU) by Van Wagoner (1988) regardless of the underlain facies, for instance, the Cretaceous–Paleocene (Barco-Upper/Middle Guadalupe) and Paleocene–Eocene (Mirador-Los Cuervos) contacts in Figures 3, 4. Meanwhile, in subaqueous conditions, we defined the transgressive surface of erosion (TSE) from Catuneanu (2017) when the marine facies overlies shallower facies; see Carbonera C8 and Mirador contact in Figures 3, 4. Otherwise, it is determined as the regressive surface of marine erosion (RSME) as the one preserved in the Upper-Middle Guadalupe in the Sagu (Figure 3).

We also defined stratigraphic surfaces following the contacts between prograding and retrograding successions. If the contact separates a prograding succession below and a retrograding succession above, it is defined as the maximum regressive surface (MRS) by Posamentier and Vail (1988). On the contrary, if the retrograding sequence underlies the prograding interval, we determined it as the maximum flooding surface (MFS) by Mitchum (1977). We defined a late Campanian MRS and MFS within the Middle Guadalupe unit in the Sagu. They are absent in the wells.

Approximately 27 studies and reports were analyzed in the locations presented in Figure 1. Table 1 summarizes the information compiled from all these studies, showing the age, name, thickness, general lithological description, and reported environmental setting of all the rock units per section. We adjusted the author’s environmental interpretation to the conceptual depositional model presented in Figure 2B.

Subsequently, chronostratigraphic correlations along the adjacent basins were assembled (Figures 5–10). These correlations are intended to cover the main depocenter of the late Campanian to Paleogene retroarc foreland basin. Two correlations depict the chronostratigraphic relationship from the Upper-Middle Magdalena in the west to the Llanos Basin in the west (Figures 5, 7). They expose how the missing gap in the western flank of the Eastern Cordillera correlates with the active depocenters at that time. The other three correlations provide insight into the chronostratigraphic correlations through the main depocenters and both flanks of the Eastern Cordillera (Figures 7–10). We attempted to reconstruct the missing record in the Cobardes Anticline following the vitrinite reflectance, thermochronological data, and nearby stratigraphic reports from Ayala-calvo et al. (2009), Caballero (2010), Caballero et al. (2013a), Porta (1958), and Silva et al. (2013) (Figure 8).

Regarding the stratigraphic surfaces determined in the Sagu and wells, we identified a strong correlation between the subaerial unconformities across the basins. The Cretaceous–Paleocene and Paleocene–Eocene contacts are unconformable and overlaid by fluvial facies. In the Upper and Middle Magdalena Valley, 1) top of the La Tabla and the lower part of the Seca unit; 2) top of Palermo/Bache, Hoyon, and upper Seca units; and 3) top of Tesalia, Chicoral, and Gualanday units (Figure 7A) are evidenced. The latter is restricted to this depocenter, as in Figures 5A, 7A.

A Maastrichtian—Paleocene unconformity is reported along the Eastern Cordillera. Across the axial part and the eastern flank, it is preserved on top of the Bogota and Upper Socha units from the sections in Bogota, Medina, and Umbita (Figures 9A, 10A). In Uribe, this SU truncates the late Campanian Palmichal Gp (Figure 9A). Similarly, the late Campanian Upper Lidita and Labor units from the Carmen de Apicala Syncline and Fusa locations are truncated by another SU that indicates subaerial exposure until the latest Maastrichtian–early Paleocene (Figure 9A).

In contrast, the presence of the subaqueous unconformities (RSME and TSE) is limited to the ECB and western LLB. The TSE is evident on top of the Mirador and Regadera units in the axial part and eastern flank of the Eastern Cordillera (Figures 9A, 10A). This TSE is absent from the western flank of the Eastern Cordillera to the Upper and Middle Magdalena Valley, where only fluvial facies have controlled the deposition since the Paleocene (Figures 5A, 6A). The RSME identified in the Sagu is more difficult to correlate. According to Catuneanu (2017) and MacEachern et al. (2012), the RSME separates inner shelf, prodelta, or offshore sediments below from the lower shoreface deposits above. In Bogota, Umbita, and Medina, a discordant contact is placed in the contact between Middle and Upper Guadalupe units (e.g., Pérez & Salazar, 1978; Guerrero and Sarmiento, 1996; Bayona, 2018). This contact is defined as a non-depositional + erosion period in Bogota and Umbita by Bayona et al. (2013). We correlated this surface as an RSME across the Eastern Cordillera (Figures 9A, 10A).

Likewise, we tracked contacts between prograding and retrograding intervals. In Bogota and Umbita, an interval of upper shoreface was identified within the late Campanian lower shoreface Labor formation, suggesting the presence of another late Campanian MRS. In Medina, Guerrero and Sarmiento (1996) reported a change within the prodelta facies of the Lodolitas de Aguascalientes. In this unit’s uppermost part, the succession of marine black mudstones is interrupted by an heterolithic interval with bivalves and U-shaped ichnofossils. That is accompanied by an abrupt reduction of the marine microfossils and a sudden increase in pollen. Based on this evidence, we delimited the late Campanian MRS in this rock interval (Figure 10A).

The characteristics are similar across the Upper and Middle Magdalena Valley. From Aico to Guaduero, the deposits were deposited under inner shelf conditions. Therefore, there is no similar trend to the one from the ECB and LLB. Nevertheless, late Campanian units preserve a peak of terrestrial microfossils. This peak can be found within the upper part of Monserrate, Upper Lidita, and Umir units from Tesalia, Aico, Piedras, and San Juan de Rio Seco. For instance, Garzon et al. (2012), Guerrero et al. (2000), and Yepes (2001a) reported an increase in pollen and spores for the late Campanian deposits from Aico and Piedras. We interpret this peak as correlatable with the MRS defined in Sagu (Figures 5A, 6A, 7A). This MRS is absent in the sections from Fusa, Carmen de Apicala, and the wells, as a SU truncates the middle to late Campanian strata (Figures 5A, 8A).

Several studies have identified stratigraphic surfaces such as unconformities (e.g., Gómez et al., 2005b; Ramon and Rosero, 2006) and maximum flooding surface (MFS) (e.g., Caballero and Naranjo, 2015, 2017; Caballero et al., 2020). However, their relationship and meaning within the regional tectonic framework are not considered. Instead, the changes in the eustatic sea level are contemplated by some authors (e.g., Garzon et al., 2012) as the primary factor driving transgressions and regressions. Detailed studies focusing on the timing of the structures (e.g., Mora et al., 2010b, 2013b; Silva et al., 2013) demonstrated that active tectonism took place from late Campanian to Paleogene. Initially, the deformation started by folding, resulting in the exhumation of some areas prone to subaerial erosion. Since the Miocene, the contractional stress caused the formation of numerous reverse faults and the inversion of previous normal faults (e.g., Parra et al., 2012; Mora et al., 2013a; Jimenez et al., 2013; Tesón et al., 2013).

The changes in the tectonic setting caused the first-order changes in the study area and defined the lifespan of the regional foreland basin. We analyzed the relationship of our interpreted stratigraphic surfaces with the major and minor tectonic events from the late Campanian to the earliest Oligocene, which is the estimated lifespan of the retroarc foreland basin (e.g., Gómez et al., 2005b; Parra et al., 2009b; Mora et al., 2010a; Horton et al., 2010b; Villagómez, 2010; Moreno et al., 2011; Villagómez and Spikings, 2013). Thus, we determined the hierarchy of each one of those surfaces. Moreover, the basin architecture was estimated based on the sedimentary thickness of the units, which is exposed in Figures 5B, 6B, 7B, 8B, 9B, 10B.

According to Catuneanu (2019b), underfilled accommodation occurs below sea level in a sedimentary basin connected to the global ocean, whereas overfilled accommodation is controlled by fluvial deposition only. The retroarc foreland basin was connected to the global sea level (e.g., Mann et al., 2006). Therefore, we delimited these stages for the basin (Figures 5–9). To analyze how the sedimentation and accommodation were impacted by the tectonic activity and the orogenic advance, we combined the chronostratigraphic correlations with an update to the palinspatically restoration from Bayona (2018) with the new data from Sagu and wells and reviewed numerous paleogeographic restorations from previous authors (e.g., Sarmiento, 2001; Cediel et al., 2011; Reyes-Harker et al., 2015; Caballero et al., 2020). Figures 11, 12 highlighted the basin architecture and the tectonic framework during the first-order changes and the facies distribution during the different stages of the basin evolution.

According to Allen and Allen (2005), Catuneanu (2019b), and Nemčok et al. (2013), once the retroarc foreland initiates, the foredeep area subsides and hosts most of the sedimentation. In some scenarios, deep marine retrograding facies deposit across the foredeep, whereas shallow marine deposition occurs in the backbulge. However, the forebulge area uplifts and is subject to either subaerial or subaqueous conditions forming the forebulge unconformity. There are cases in which a condensed section is preserved on top of the forebulge, yet it requires dynamic subsidence, such as in the Western Interior Basin (Catuneanu, 2004, 2019a; Miall et al., 2008). This unconformity underlies the first-order foreland sequence, and its north-south diachroneity results from the diachronic exhumation of the forebulge (Catuneanu 2004).

In the central area of Colombia, once the uplifting of the Central Cordillera started, two areas were exposed. First, the western flank of the Eastern Cordillera was diachronically exposed from south to north. In the south, multiple authors (e.g., Caicedo et al., 2000, 2002; Ramon and Rosero, 2006) recognized an unconformity in the Carmen de Apicala synclines (Figures 8A, 11, 12), indicating exhumation and erosion from the Campanian to the Paleogene (Figure 12). Bayona et al. (2003) identified a similar unconformity in the Fusa section. The time gap associated implies that the exhumation and erosion occurred from the latest Campanian until the latest Maastrichtian–the earliest Paleocene. Then, the sedimentation was resumed with marginal fluvial conditions (Figures 11, 12). The Cobardes Anticline started to uplift north to the study area during the Middle Maastrichtian (Figure 11C). The sedimentation has not resumed since, and the Jurassic rocks are exposed in the structure’s core (Julivert, 1970). The facies overlaying this unconformable contact along the western flank of the Eastern Cordillera indicates it represents a SU with a younger northward pattern from south to north.

Also, most of the Llanos Basin area was uplifting, and it was under subaerial exposure from the late Campanian to the middle Paleocene (Figures 4, 11, 12). Once the compressional regime triggered the formation and uplift of the Central Cordillera in the late Campanian, the previous extensional basin experienced a reduction of accommodation. Subsequently, a northward regression of the shoreline took place. The compression also resulted in a lithospheric flexure that exhumed most of the Llanos Basin. Similar to the unconformity across the western flank of the Eastern Cordillera, it is underlined by the Barco formation’s fluvial facies and is classified as a SU (Figures 3, 8). However, the thickness of the preserved Paleocene sedimentary record is almost half of the one preserved in Fusa.

We evaluated scenarios to define which of the subaerial unconformities is the forebulge unconformity. In the first scenario, the presented estimated lateral extension supports the interpretation of a foredeep across the Upper and Middle Magdalena Valley, a forebulge through the Eastern Cordillera’s western flank, and a backbulge along the Eastern Cordillera and westernmost Llanos Basin. Figure 11A indicates that the presumed foredeep extended for ±60–100 km, whereas the backbulge and forebulge extended for ±80–100 km and ±150–250 km, respectively. Those distances may fit the regular extension for a foreland basin defined by DeCelles and Giles (1997). Nevertheless, the thickness of the late Campanian to Paleogene sedimentary record does not support this idea. Suppose the western flank of the Eastern Cordillera acted as the backbulge during the late Campanian to Paleogene. In that case, its sedimentary record should be just a tiny portion of the preserved thickness across the foredeep (DeCelles and Giles, 1997). Instead, the thickness measured in San Juan del Rio Seco in the Upper and Middle Magdalena Valley and Bogota in the Eastern Cordillera is similar (Figure 6B). The data suggest that the Eastern Cordillera and Upper/Magdalena Valley hosted equivalent sediment from the late Campanian to Paleogene. Therefore, the forebulge was not located along the western flank of the ECB.

In contrast, Gómez et al. (2005b), Gómez et al. (2003), and Parra et al. (2010) suggested that the forebulge of this late Campanian to Paleogene retroarc foreland basin was located across the Llanos Basin. The Middle/Upper Magdalena Valley, Eastern Cordillera, and Llanos Basin comprised the foredeep in this scenario. It extended for ±290–450 km by the latest Campanian, approximately. The thickness of the sedimentary record indicates a pattern of becoming thinner eastward (Figures 5B, 6B). Accordingly, the exhumation and subsequent erosion in Uribe, Sagu, and the wells’ area is associated with the forebulge deformation. The SU that truncates the late Campanian to Middle Maastrichtian strata is defined as the forebulge unconformity.

The diachroneity of the forebulge unconformity is related to the migration of the orogen. Villagómez (2010) stated that the uplifting of the orogen started in the south during the late Campanian and expanded northward. That is reflected in the sedimentary record. Figure 10A demonstrates how the interpreted forebulge unconformity truncates relatively younger strata northward. The unconformity truncates relatively older rock westward from west to east (Figures 5A,B). We interpret this as associated with the orogen’s eastward advance, suggesting the forebulge was not static.

During the late Campanian, the uplifting and exhumation of the forebulge resulted in the erosion of the Middle Guadalupe along the western Llanos Basin. Then, the forebulge expanded westward by the middle-late Maastrichtian, exhuming the rocks from the Sagu area (Figures 11, 12). This westward expansion reached the area of Medina during the early Paleocene (Figure 11E). Nonetheless, the data suggest that the forebulge migrated eastward by the middle-late Paleocene, promoting the deposition of Barco and Cuervos across the western part of the Llanos Basin (Figure 12). That interpretation aligns with the foredeep migration proposed by Bayona et al. (2007).

Catuneanu (2019b) stated that the forebulge unconformity is a first-order stratigraphic surface. It bounds the retroarc foreland sequence at the bottom. However, it is restricted to the forebulge area. Therefore, we analyzed the stratigraphic record across the late Campanian depocenters to determine the stratigraphic response of the depocenters during the start point of the retroarc foreland basin.

In the Sagu, the late Campanian deposits in the study area expose a progradation pattern and a subsequent retrogradation pattern. The upper part of the Middle Guadalupe unit preserves a contact between prograding facies below and retrograding facies above (Figure 3). That contact is absent in the wells because that area was prone to subaerial exposure and erosion. We classified that contact as an MRS. Across the Upper and Middle Magdalena Valley, the Upper Lidita, Buscavidas, and lowermost Umir units preserve a change in sedimentary structures and the distribution of microfossils. Garzon et al. (2012), Guerrero et al. (2000), and Yepes (2001a, 2001b) reported a decrease in marine palynomorphs while the continental microfossils (pores and spores) increased. This marks a coetaneous contact between the underlain prograding pattern and the above retrograding succession, which correlates with the MRS in the Sagu section (Figures 5, 6).

A similar change is recorded within the late Campanian strata across the Easter Cordillera. In the Medina area, the Lodolitas de Aguascalientes registers an increase in pollen content that coincides with an increase in sandy portion and shallow marine fossils and ichnofossils (Guerrero and Sarmiento, 1996). In Bogota and Umbita, at the top of the Middle Guadalupe unit, there is a contact of prograding below and retrograding facies above (e.g., Pérez and Salazar, 1978). We defined another MRS within those intervals. This MRS is absent in the western flank of the Eastern Cordillera due to that area was exhumed and prone to erosion during the late Campanian–Maastrichtian times. However, we define that as another SU.

The south-to-north extension of the late Campanian MRS throughout the depocenters is illustrated in Figures 6–8. The low-resolution isochrones impede the constraining of the exact age of this MRS. Nonetheless, the data indicate this surface is relatively younger in Guaduero than in Tesalia (Figure 6) and relatively younger in Medina than in Sagu (Figure 7). The surface is connected with the first SU formed in the western Eastern Cordillera and the Llanos Basin from west to east (Figures 5, 6). Hence, this suggests they were formed simultaneously, which is hierarchically similar. They are the first-order lower boundary of the retroarc foreland basin and along the overlying deposits belong to the first-order foreland basin sequence.

The upper boundary was formed during tectonic quiescence in the early-middle Eocene. According to Parra et al. (2009b), this tectonic quiescence is associated with orogenic unloading and the subsequent end of the retroarc foreland basin. At this time, the western flank of the Easter Cordillera separated the two depocenters, the Middle/Upper Magdalena Valle in the west, and the Eastern Cordiller–Llanos Basin in the east (Figures 2, 13). For that reason, we defined the SU truncating the Paleocene–Eocene strata across the study area as the upper first-order boundary of the retroarc foreland basin. The overlain units were deposited in a new tectonic setting where the Upper and Middle Magdalena Valley and Llanos Basin acted as independent depocenters.

Our analysis suggests that the diachroneity of the first-order upper boundary is related to the tectonic stages of orogenic loading and unloading proposed by Parra et al. (2009a). In the Upper and Middle Magdalena Valley, this SU tends to truncate older units northward (Figure 7A). In the Eastern Cordillera, the time gap and age of the rocks truncated by this SU decreases toward Bogota (Figure 9A). At the same time, this SU truncates older units from west to east (Figures 5A, 6A). We associate the west-east diachroneity with the ongoing eastward advance of the tectonic deformation. In contrast, the north-south diachroneity may involve interacting with the Caribbean tectonic plate to the north, which flipped its subduction since the Paleocene times (Montes et al., 2019).

Regarding the continuous sedimentation identified near the Bogota area, the resolution of the data limits the entire understanding of this fact. We associate it with the eastward migration of the depocenters. Bayona (2018) proposed that dynamic subsidence may have occurred in that area. Notwithstanding, Catuneanu (2019a) clarified that dynamic subsidence results in positive accommodation along the forebulge, differing from the new data incorporated from Sagu and wells 1–3. Analyzing the outcrops across the western flank of the Eastern Cordillera and their correlation with the presented data is suggested to constrain the basin architecture and evolution.

Additionally, thicknesses should be decompacted to estimate and establish their temporal variation. Biostratigraphy, thermochronology, and detrital zircon are essential information to improve the understanding of the evolution of the forebulge. The stratigraphic surfaces presented are a powerful element to control the facies correlation. We recommend carrying out a detailed calibration of the palinspastic restorations to obtain a more precise measure of the retroarc foreland basin extent and architecture.

Understanding the lifespan of the retroarc foreland basin is fundamental not only to comprehending its evolution but also to a better understanding of the timing of significant events and their implication on the sedimentary record. Our results indicate that the western flank of the Eastern Cordillera and the western Llanos Basin uplifted and were under subaerial erosion once the retroarc foreland basin started during the late Campanian. Following the sedimentary thickness across the Upper and Middle Magdalena Valley and the Eastern Cordillera, the estimated basin architecture suggest that the uplifting of the western Llanos Basin was related to the forebulge deformation that resulted in the erosion of the Middle Guadalupe units in the area of the wells (Figure 13). The uplifting of the western part of the Eastern Cordillera is understood to be associated with the ongoing compressional deformation. An SU is an unconformity truncating the late Campanian deposits from the Carmen de Apicala and Fusa area, following Catuneanu (2017) and Van Wagoner (1988).

At the same time, the late Campanian Middle Guadalupe from the Sagu preserve a contact separating prograding facies below from retrograding facies above (Figures 3, 13). A similar contact was recognized within the Upper Lidita, Lutitas y Arenas, Buscavidas, and Upper Guadalupe units from the Middle/Upper Magdalena Valley and Eastern Cordillera. That contact is defined as a maximum regressive surface (MRS) according to Catuneanu (2017) and William et al., 1996. The combination of this MRS, the SU from the western flank of the Eastern Cordillera, and the forebulge unconformity represents the lower first-order boundary of the retroarc foreland stratigraphic sequence (Figure 13).

The correlation of the forebulge unconformity to the Sagu, Uribe, and Medina sections indicates that this surface truncates relatively younger units northward and eastward (Figures 5A, 10A). Likewise, the missing record from the Upper Guadalupe to Barco in Sagu and Guaduas to Lower Socha in Medina evidence that the forebulge expanded westward from the latest Campanian to the earliest Paleocene. By the Middle Paleocene, sedimentation resumed in those areas because the forebulge migrated eastward, resulting in the deposition of the Barco and Los Cuervos rock units. The north-to-south diachroneity of the forebulge unconformity is associated with the orogen’s northward migration. Therefore, the Uribe section did not host sediments since the late Campanian. The orogen moved northward, and no accommodation was created in that area.

In the Paleocene, most of the deposition across the depocenter was fluvial. Although the sedimentary thickness reveals that the accommodation space created was higher than in late Campanian to Maastrichtian times, sediment supply and sedimentation outpaced the accommodation rate. That enhanced the northward migration of the shoreline. Nonetheless, the Sagu and Medina strata unveil that brackish ocean water flooded the area by the late Paleocene (Figure 12B). Our interpretation is that the last tectonic pulse that triggered the accommodation rates to outpace the sedimentation took place. That also promoted the area’s flooding to the north of the Nuevo Mundo Syncline (Figure 12B).

During the latest Paleocene and early Eocene, tectonic quiescence ceased accommodation across the central part of Colombia. The SU on top of Los Cuervos Fm in the Sagu and the wells is correlatable across the Eastern Cordillera and the Upper and Middle Magdalena Valley. At this time, the western flank of the Eastern Cordillera started to separate the depocenter (Figures 2, 13). Therefore, this SU represents the end of the retroarc foreland basin, and we determine it as the upper first-order boundary (Figure 13). The Eocene Mirador unit from the Sagu and the wells along with the Regadera, Picacho, Tesalia, Gualanday, Chicoral, and San Juan del Rio Seco was deposited during a transition period.

Our analysis also identified patterns related to the shoreline trajectories during the retroarc foreland basin. We recognized reported changes in the population of marine palynomorphs (e.g., Piedras and Aico), variations in organic matter content and sediment structures (e.g., San Juan del Rio Seco and Girardot), and the presence of coetaneous thick exploitable coal seems (e.g., Bogota). Some of those patterns are contemporaneous and can be related to events where the accommodation outpaced the sedimentation rate, resulting in marine transgressions. The first-order sequence boundaries we present in this study intend to constrain that framework and facilitate the correlation of similar events in the central part of Colombia.