Geomorphological Characterization of the Colombian Orinoquia

Abstract

The Colombian Orinoquia was shaped within a tectonic and sedimentary framework linked to the uplift of the Andean cordilleras during the Oligocene–Miocene. This orogenic event generated two tectonic fronts and facilitated extensive fluvial sedimentation across a broad alluvial geosyncline. The present geomorphological configuration reflects the cumulative interaction of tectonic and erosional processes with Quaternary climatic dynamics, which together produced complex landscape assemblages characterized by plains with distinctive drainage patterns. To delineate and characterize geomorphological units, we employed multidimensional imagery and Machine Learning techniques within the Google Earth Engine platform. The classification model integrated dual polarizations of synthetic aperture radar (L-band) with key topographic variables including elevation, slope, aspect, convexity, and roughness. The analysis identified three major physiographic units: (i) the Foothills and the Floodplain, both dominated by fluvial environments; (ii) the High plains and Serranía de La Macarena (Macarena Mountain Range), where denudational processes predominate; and (iii) localized aeolian environments embedded within the Floodplain. These contrasting dynamics have generated a broad spectrum of landforms, ranging from terraces and alluvial fans in the Foothills to hills and other erosional features in La Macarena. The Floodplain, developed over a sedimentary depression, illustrates the combined action of fluvial and aeolian processes, whereas the High plains is characterized by rolling plains and peneplains formed through the uplift and erosion of Tertiary sediments. Such geomorphic heterogeneity underscores the interplay between tectonic activity, climatic forcing, and surface processes in shaping the Orinoquia landscape. The geomorphological classification using Random Forest demonstrated high effectiveness in discriminating units at a regional scale, with accuracy levels supported by confusion matrices and associated Kappa indices. Nevertheless, some degree of classificatory overlap was observed in fluvial environments, likely reflecting their transitional nature and complex sedimentary dynamics. Overall, this methodological approach enhances the objectivity of geomorphological analysis and establishes a replicable framework for assessing landform distribution in tropical sedimentary basins.

1. Introduction

Thematic mapping via remote sensing represents the most efficient method for quantifying and monitoring landscape structure [1] and geomorphological attributes [2,3,4]. Synthetic Aperture Radar (SAR) imagery is particularly valuable, providing data on the physical and dielectric properties of terrain irrespective of cloud cover or time of day, with some penetration capability through dense vegetation [5,6,7,8,9]. Technological advances, including improved spatial resolution and revisit frequency, now enable reliable, continuous data acquisition [10,11,12]. The utility of SAR is enhanced by its frequency-dependent penetration depth [13,14]; L-band systems, for instance, are sensitive to soil texture and moisture [5,15]. When combined with Digital Elevation Models (DEMs) from SAR interferometry—which allow for the calculation of key terrain parameters like slope and roughness [4,14,16,17,18,19]—these tools form a powerful, multidimensional basis for landscape analysis.

To translate these data into objective landform classifications, Machine Learning (ML) algorithms have become essential. These non-parametric methods are highly suited for classifying remote sensing data, which often deviate from normal statistical distributions, and for handling multidimensional imagery [20,21,22]. The Random Forest (RF) algorithm is notably robust, constructing multiple decision trees from randomly sampled data to produce models with high variance but low bias, thereby reducing overfitting [23,24,25]. Cloud-based platforms like Google Earth Engine (GEE) have revolutionized the processing of such large datasets by providing parallel computing power and hosting petabytes of planetary-scale geospatial data [26,27,28,29]. This eliminates traditional bottlenecks of data download and local computational capacity, facilitating the flexible application of analytical functions [30,31].

Despite these technological advancements, regional-scale geomorphological mapping in complex, low-relief terrains like the Colombian Orinoquia often remains reliant on subjective visual interpretation of optical imagery. While a methodological framework for analytical geomorphological mapping exists in Colombia [25], and studies confirm the superiority of integrated DEM and SAR data for feature identification in the region [32,33,34,35], a comprehensive, reproducible map produced via automated classification is lacking. This gap limits the objective analysis of landscape control on ecological patterns.

To address this, our study has two primary objectives: (i) to produce the first analytical geomorphological map of the Colombian Orinoquia using an integrated SAR, DEM, and ML approach within the GEE platform, and (ii) to quantitatively evaluate the distribution and influence of the mapped geomorphological environments within the region’s major physiographic units. This work establishes a reproducible, objective baseline for studying biodiversity patterns and ecosystem dynamics across environmental gradients in the region.

2. Materials and Methods

2.1. Regional Framework

The tectonic conditions that enabled the formation of the Orinoquia are associated with the uplift of the Eastern cordillera during the mid-Tertiary (Oligocene–Miocene), expressed in two principal fronts: a long front in the region between the Duda and Upía rivers, and a shorter yet equally intense front in the El Cocuy region, coupled with continuous sedimentation from river systems descending from the newly formed mountain range [36,37]. The region developed over an extensive alluvial sedimentary geosyncline, influenced by tectonic events, Quaternary climatic dynamics, ancient erosional processes in the Guiana Shield massifs, and more recent erosion of the Andes [38,39,40,41,42]. Although often described as a territory with relatively homogeneous natural features, gentle slopes, and a landscape largely dominated by grasses [43], its physiography is closely related to both chronological criteria and its proximity to the Eastern cordillera and the Meta River. The latter constitutes a key hydrological boundary that determines drainage conditions and shapes relief variation: the poorly drained Orinoquia to the north of the river, characterized by alluvial and aeolian fans and plains, and the well-drained Orinoquia to the south, dominated by flat to undulating uplands, hills, and terraces [44,45,46,47]. The sedimentary deposits, of both Tertiary and Quaternary age, overlie an ancient platform belonging to the Guiana Shield, composed of Precambrian rocks with ages ranging from one to 1.8 billion years. Outcrops of these formations are particularly prominent in Serranía de La Macarena (La Macarena Mountain Range, hereafter La Macarena) and the High plains, where they appear as hill ranges and inselbergs near the Guaviare and Orinoco rivers, locally referred to as the residual High plains [37,43,48].

2.2. Study Area

The Colombian Orinoquia is a vast natural region within the Orinoco River basin. It extends from the eastern flank of the Eastern Cordillera of the Andes to the banks of the Orinoco River, forming a plain with a pronounced environmental gradient from west to east. Its general boundaries are defined by major fluvial systems: to the north by the Arauca and Meta rivers; to the east by the Orinoco River, which marks the border with Venezuela; to the south by the Guaviare and Guayabero rivers; and to the west by the foothills of the Eastern Cordillera [39,40].

From a physiographic standpoint, the region comprises a sequence of units that condition its ecological and hydrological dynamics. To the west, the Foothills (Piedemonte) represent the transition zone between the Andean mountains and the plain, characterized by dissected relief and moderate to steep slopes. To the south, the system of dissected plateaus of La Macarena, a biogeographic enclave of Precambrian origin, stands as a geomorphological and biodiversity singularity. Eastward, the Floodplain (Llanura aluvial), with active fluvial systems and extensive flood zones, and the High Plains (Altillanura), an extensive surface of ancient terraces with acidic soils and well-established drainage, are distinguished [41,42].

The region’s climate is typical of a tropical savanna, characterized by pronounced seasonality with a single, well-defined dry season. Annual precipitation exhibits a strong gradient, decreasing from approximately 4387 mm in the western foothills to around 1661 mm in the floodplain. Rainfall follows a unimodal, bi-seasonal regime distinguished by an abrupt transition between phases, which can drive a rapid shift from hydrological deficit to surplus within weeks. The spatial distribution of precipitation and evapotranspiration is governed primarily by orographic effects—notably from the Eastern Cordillera and the La Macarena Massif—and by synoptic wind patterns, including the Easterly Trade Winds and convection associated with the Intertropical Convergence Zone. This marked seasonality fundamentally controls regional hydrological regimes, governing river discharge cycles, floodplain inundation dynamics, and resultant vegetation patterns. The natural vegetation reflects this environmental heterogeneity, forming a mosaic that includes natural grasslands (upland, lowland, and marshy types), gallery forests along watercourses, foothill forests, and—in La Macarena—tropical humid forests with high levels of endemism. Vegetation composition and structure are closely linked to topography, water-table depth, and flood duration [46,47,48,49].

2.3. Delineation of Physiographic Units

The delineation of the physiographic units for this study was conducted using an integrated cartographic approach, combining altitudinal, hydrographic, and geomorphological criteria. As a general base, the vector cartographic layer of Colombia’s hydrographic zonation [49] and the NASA SRTM Version 3 Digital Elevation Model (DEM) at 30 m resolution, available in Google Earth Engine [50], were employed. The general boundaries of the study region follow the definition by Jaramillo & Rangel-Ch. [39,40].

The specific criteria for each unit were as follows. Foothills: Its eastern and western delimitation was based on altitudinal thresholds defined by hydrographic basin [32]. South of the Meta River, within the Guamal River basin, the upper and lower limits were set at 575 m and 350 m, respectively. North of the Meta River, the thresholds were defined as follows: 675 m (upper) and 200 m (lower) for the Meta and Pauto river basins; 600 m and 200 m for the Casanare River basin; 425 m and 200 m for the Cravo Sur River; and 375 m and 175 m for the Arauca River. Its northern boundary was the Arauca River, and its southern boundary was the Ariari River basin. La Macarena: It was delimited to the west by the altitudinal thresholds of 575 m in the Guayabero River basin and 775 m in the Guaviare River basin; to the north and east by the boundaries of the Ariari River basin; and to the south by the Guayabero River and the limits of its respective basin. Floodplain: It was defined to the west by the eastern limit of the foothills, to the north by the Arauca River, to the east by the international border with Venezuela, and to the south by the Meta River. High Plains: Its delimitation was established to the west by the eastern limits of the foothills and La Macarena, to the north by the Meta River, to the east by the Orinoco River, and to the south by the Guaviare River (Figure 1).

2.4. Methodological Framework

Integrated systems were employed, wherein the outputs or results of one procedure served as the inputs or parameters of the subsequent process. The workflow (Figure 2) was applied through differentiated geomorphological modeling for each physiographic unit, rather than a single, overarching model for the entire study area. This approach was adopted to account for the distinct environmental and morphometric conditions characterizing each subregion, thereby increasing model accuracy. A single, generalized model would have been compromised by the high spatial heterogeneity inherent to the diverse landscape, leading to a potential loss of precision in capturing local-scale processes and forms.

The geomorphological model incorporated two L-band SAR polarizations and five parameters considered fundamental terrain descriptors, as proposed by Franklin [26]: (i) HH/HV polarizations from an annual Alos-PalSAR mosaic for 2021 (Global PALSAR-2/PALSAR collection, available in GEE), which underwent Speckle correction [51] using a focal mean filter with a 50 m window to improve image classification accuracy [22,52]; (ii) the NASA SRTM v.3 DEM, from which slope, aspect, and convexity cartographic layers were derived; (iii) terrain slope expressed as a percentage, (iv) aspect, or slope orientation, expressed in degrees relative to North; (v) convexity, calculated from the relative elevation of a seven-pixel moving window compared to its surroundings, categorized as concave (window elevation lower than its surroundings), convex (window elevation higher), or planar (window elevation similar); and (vi) terrain roughness, or surface irregularity, quantified through a local variance analysis of slope using seven-pixel moving windows.

This parameter selection is grounded in comprehensive geomorphometric characterization that considers both primary geometric attributes and the spatial expression of terrain texture and moisture. Elevation constitutes the primary attribute, from which slope (degree of inclination) and aspect (direction of maximum inclination) are directly derived. Convexity describes terrain curvature, distinguishing convex (ridgelines) from concave (valleys) morphologies. Finally, roughness is quantified through spatial texture analysis, providing a robust measure of surface heterogeneity that surpasses the limitations of traditional statistical metrics [27]. The integration of these five descriptors, alongside the SAR polarizations, offers a complete quantitative foundation for geomorphological analysis.

The performance of RF classification depends on the quality of the training samples, which must unequivocally represent the thematic categories within the multidimensional image [53]. The training areas for the geomorphological model were defined using digital elevation profiles derived from the DEM, combined with the visual interpretation of L-band SAR imagery. Polygonal entities of the identified landforms were delineated according to the geomorphological units described by the Colombian Geological Survey [25]. The training areas covered more than 0.25% of the study area, ensuring both representativeness and validity of the sample [54]. This allocation considered the spectral breadth of the multidimensional images as well as the landscape representativeness of the classes, since balanced samples across thematic categories yield higher accuracy by reducing both commission and omission errors in underrepresented classes [55,56].

Although it is generally assumed that a higher number of decision trees improves the performance of the RF model, processing time increases linearly with this parameter, thereby justifying its calibration to an optimal number of trees [52]. Accordingly, the classifiers were configured using sampling variables derived from multidimensional images and vector layers of the training areas. The dataset was subsequently partitioned into 70% for training and 30% for accuracy assessment, applying an iterative function in increments of 10 trees until reaching a total of 150. Finally, predictions were generated based on the sampling variable and the sequential tree parameters, and classification accuracy was plotted against the number of trees [32].

Subsequently, RF classification was performed, in which training variables for the classifiers were defined using the input layers and training areas, following an approach analogous to the sampling variables employed during the parameterization procedures. The classifiers were then configured with the training variables and the number of trees that yielded the highest accuracy during parameter tuning.

The classification outputs were initially exported as 30 m resolution rasters, consistent with the input data. A three-step generalization process was applied to achieve a 5-hectare minimum mapping unit (MMU) before final conversion to vector format. The procedure comprised the following: (i) thematic aggregation using a 2 × 2 moving-window majority filter to reduce pixel isolation and mitigate inherent speckle noise from the input SAR imagery; (ii) identification of contiguous pixel groups for each class, considering 8-neighbor adjacency, and the removal of all groups below the 5-ha MMU threshold; (iii) iterative elimination and reassignment of remaining sub-5-ha patches via a focal majority filter, which reclassifies them to the value of the surrounding dominant class that meets the area requirement.

Model accuracy was assessed on the original 30 m classification output, before any generalization, using the same training data and Random Forest parameters [24]. The 70%/30% split was maintained, and standard accuracy metrics (confusion matrix, overall/producer/user accuracy, Kappa) were calculated. Evaluating accuracy after the 5-ha generalization would be inappropriate, as the post-processing smooths and aggregates pixels, artificially inflating agreement metrics and not reflecting the classifier’s true performance [32].

3. Results

As a result of the physiographic delimitation, the calculated area of the Orinoquia region is 233,546.19 km², which is smaller than the 260,000 km² previously reported by FAO [44] and Riveros [57]. Four geomorphological environments were identified: (i) the denudational, defined by the combined action of moderate to intense processes of weathering, erosion, and gravity- and rainfall-driven transport; (ii) the fluvial, encompassing erosional processes generated by river currents as well as the deposition of sediments in adjacent areas; (iii) the aeolian, determined by wind activity and the consequent transport, fragmentation, and deposition of particles of varying sizes; and (iv) the structural, resulting from processes related to Earth’s internal dynamics, mainly associated with rock folding and faulting [25].

3.1. Foothills Geomorphology

The estimated extent of the Foothills physiographic unit is 19,123.71 km², representing 8.19% of the total Orinoquia region. This delineation exceeds a previously reported area for this unit by 6537.75 km² [44,57]. The geomorphological map of the Foothills (Figure 3) reveals a clear predominance of fluvial landforms, which occupy 80.83% of the area. Within this fluvial environment, sub-recent accumulation terraces are particularly significant. These terraces are associated with valley widening by rivers descending from the Eastern Cordillera and are located along the eastern Foothills margin, coinciding with the unit’s lowest elevations and gentlest slopes. They largely form progressive, transitional units towards the Floodplain and show no apparent signs of degradation [39,40,58,59].

In contrast, ancient accumulation terraces of Pleistocene age are exposed in areas of higher elevation and steeper slope, shaped by neotectonic activity and surface runoff. These older terraces are overlain by sediments deposited following episodes of eastern subsidence, which subsequently gave rise to lower-gradient Holocene terraces. Within these deposits, anastomosing channels indicative of intense fluvial dynamics are clearly identifiable [37,60]. The next most representative features are active alluvial fans, formed by torrential and fluvial accumulation in a radial pattern where streams debouch onto flat areas. Originating from the higher cordillera, these fans are frequently situated between structural units or along the higher western margins of the Foothills, juxtaposed with the lower-lying accumulation terraces. The torrential sediment load from the cordillera is deposited discontinuously across these fans, creating conditions of both actual and potential instability and helping to define the western boundary of this physiographic unit.

Ancient alluvial fans, discontinuously distributed along the cordillera-adjacent margin, were deposited during the Pleistocene. Subsequent deformation by recent tectonic activity—including folding, range-parallel faulting, and episodes of subsidence and uplift—has shaped these features. These processes have produced multiple, often difficult-to-differentiate terrace levels, which are traversed by braided channels [61]. Floodplains are also prominent, associated with the main braided channels descending from the cordillera. Within these wide, dynamic channels, gravel is deposited, with clast size fining eastward in response to decreasing slope and transport capacity, while finer sediments accumulate along the banks [60]. Accumulation terraces are commonly located between sub-recent alluvial fans and sub-recent terraces. The denudational environment covers 11.70% of the Foothills, characterized by erosive hillslopes along the steep cordillera margin, frequently associated with structural units. Representing the smallest areal extent, the structural environment accounts for 7.47% of the Foothills. It is located along the cordillera margin and is dominated by structural hills and ridges.

3.2. Geomorphology of La Macarena

Based on this study, La Macarena occupies an estimated area of 20,747.63 km², constituting 8.88% of the Orinoquia region. Denudational environments are dominant within this physiographic unit, covering 46.88% of its surface (Figure 4). The most extensive features within this environment are dissected hills, shaped by intense denudational processes. These are primarily located in the northern sector, along the margin adjoining the High Plains and adjacent to the floodplains of the Ariari, Guejar, and Guaduas rivers. Undissected hills, associated with erosional and soil-creep processes, are subdominant and are found mainly adjacent to the floodplains of Caño Talanquera and the Cunimia, Guejar, and Guayabero rivers. Erosive hillslopes are also noteworthy, predominantly situated in lower-elevation areas across both the massif and the adjacent cordillera.

The fluvial environment accounts for 43.83% of La Macarena. Floodplains predominate within this category and are primarily associated with the major river channels descending from the cordillera. In terms of spatial extent, these are followed by ancient alluvial fans, located mainly in higher-altitude areas along the cordilleran margin. These fans were formed by sediment contributions from the Ariari, Uruime, Yamanito, Guape, Guejar, and Duda rivers. Additionally, sub-recent accumulation terraces are significant, particularly those associated with the floodplains of Caño Berlín and the Guejar, Duda, Guayabero, Papamene, and Leiva rivers. Representing the smallest proportion of the massif, the structural environment encompasses 9.29% of the area. It is characterized by structural hills distributed across much of the massif and by structural slopes defined by preferential planes parallel to the terrain’s dip, which are often located in the lower-altitude sectors.

3.3. Floodplain Geomorphology

The Floodplain was estimated to cover 52,187.13 km², representing 22.35% of the Orinoquia region. This extent aligns with the value reported by Jaramillo & Rangel-Ch. [39,40] but is smaller than the 65,250 km² previously recorded by other authors [44,57]. This unit is characterized by extensive terraces of poorly consolidated materials that converge into a vast, frequently inundated plain featuring numerous natural levees and mobile sandbanks. These features expand across longitudinal belts of residual Tertiary sedimentary layers, which contain distinct alluvial paleoforms. The intermediate depressions, locally known as bajos and esteros, undergo seasonal flooding during the rainy periods [35,37,62,63]. These conditions promote continuous river course migration, with lateral sedimentation leaving few prominent traces due to the dominance of vertical overbank deposition [39,40,60,63].

The fluvial environment predominates within the Floodplain, occupying 51.52% of its total area (Figure 5). Floodplains constitute the most extensive landform in this environment and are closely associated with the river channels descending from the cordillera. In the northwestern sector, these inundated zones cover more extensive areas than in the south, forming a complex system affected by overbank flooding and alluviation. This is particularly evident along the Arauca, Lipa, Ele, Cravo Norte, Cusay, Cuiloto, Casanare, and Ariporo rivers, as well as their major tributaries (e.g., Caño Colorado, Caño Palmer, Caño Caranal, Caño Cuarteles, Caño Matepalma). Here, fill–spill plains are common, exhibiting poor drainage, permanent saturation, and active siltation [35,37,60]. Sub-recent accumulation terraces are the next most representative feature. These are found in the unit’s higher-altitude zones, particularly along its western margin bordering the Foothills. Recent accumulation terraces are frequently located in intermediate altitudinal sectors between the Foothills and the Meta River, where their formation is also influenced by aeolian processes [63].

The aeolian environment extends across the eastern margin of the Floodplain, covering 17,346.60 km² (33.10% of the unit)—an area 3422.15 km² smaller than previously reported [44,57]. Its origin is linked to the intense deposition of alluvial material during the Pleistocene and Holocene, which created extensive unconsolidated sediment covers. These deposits were subsequently reworked, transported, and redistributed by strong winds, a process reinforced by the marked alternation of dry and wet seasons following past desertification phases [37,43,60]. Within this environment, loess mantles are predominant. They formed through aeolian suspension transport and subsequent compaction of silts over older sediments under conditions favorable for deflation, and are primarily located in poorly drained flat areas along the southeastern margin [58,60]. Sand sheets follow in extent, concentrated on the northeastern margin, alongside dune fields. These dunes, formed of sand mounds that frequently overlie preexisting landforms, occur in the northeastern-most sector between the Meta and Arauca rivers. They are typically oriented by prevailing trade winds into elongated or longitudinal units reaching up to 15 km in length [64].

Occupying the smallest proportion (15.38%) of the Floodplain is the denudational environment, represented exclusively by plains associated with floodplains in the northern sector. Here, complex fluvial systems frequently display channel bifurcations, as the low-energy, weakly hierarchical drainage network lacks the capacity to effectively dissect the terrain [35,37,63].

3.4. Geomorphology of the High Plains

The High Plains constitute the most extensive physiographic unit within the Orinoquia, covering an estimated area of 141,487.71 km², which represents 60.58% of the region. This unit is composed of sedimentary deposits originally accumulated in marine and coastal environments, later redeposited following the uplift of the Eastern Cordillera, and subjected to prolonged weathering that has significantly altered their mineralogical composition [43]. The delineated extent is smaller than a previously reported area of 188,668.61 km² (62.50% of the region) [65], as the present classification assigns the terraces extending from the cordillera to the Metica and Upía to Ariari rivers to the Foothills unit.

The denudational environment predominates, covering 66.03% of the High Plains (Figure 6). The most extensive landforms are residual plains, located along the eastern margin within the influence of the Orinoco River. These gently inclined surfaces, associated with ancient erosion surfaces and residual soils, are composed of unconsolidated Quaternary fluvial sediments. They are frequently contiguous with the broad floodplains of major Orinoco tributaries, fill–spill plains on former overflow surfaces, and terrace sequences [37,58,59].

Peneplains are the next most representative features, occupying the central margin under the influence of the Meta River. Bounded by residual plains to the east, hill systems influenced by the Manacacías River to the west, and dissected hills of the Vichada River basin to the south, these undulating surfaces are marked by low hills and intermontane valleys that form a dense reticular drainage pattern, indicative of the dissection of an ancient high plain [59]. Dissected and highly dissected hills collectively account for approximately one-third of the unit’s area. These rounded elevations, with short to moderate slopes, are formed by fluvial erosion of unconsolidated Tertiary sediments and typically exhibit sub-dendritic drainage patterns [59]. Highly dissected hills, concentrated along the western margin near the Manacacías River, are characterized by steep, escarped slopes, V-shaped valleys, and advanced degradation features such as petroferric crusts, gullies, and subsurface tunnels resulting from intense runoff [58]. In contrast, dissected hills are widely distributed across the central High Plains, featuring inclined slopes with moderate dissection, U-shaped valleys, and often crusted margins with flat floors [37].

The fluvial environment covers 33.78% of the High Plains. Floodplains associated with major river channels are the most distinctive features, followed in spatial extent by accumulation terraces, which are most extensive in areas influenced by the Guaviare River in the south. The aeolian environment is minimally represented, consisting exclusively of loess mantles. These occur in low-gradient zones adjacent to floodplains near the confluence of the Manacacías River. These compacted silt sheets are homologous remnants of more extensive loessic formations north of the Meta River, which characterize the southwestern margin of the Floodplain.

4. Discussion

The parameterization of the Random Forest (RF) classifiers determined that the optimal number of decision trees for the highest model accuracy ranged from 50 in the Foothills to 120 in the High Plains. The overall model accuracy, calculated as the proportion of correctly classified pixels within the training areas, varied between 0.875 (Foothills) and 0.920 (Floodplain). The Kappa coefficient, which measures the agreement between the classifier and the reference data beyond chance [66,67], consistently exceeded 0.86 across all model runs, confirming a high probability of correct classification relative to a random assignment of pixels. Confusion matrices revealed maximum omission and commission errors of 28% and 34%, respectively, predominantly associated with units of the fluvial environment. Despite this, Synthetic Aperture Radar (SAR) imagery proved highly effective for discriminating geomorphological units, including those in densely vegetated areas like the southern High Plains [5,6,9]. Its efficacy is attributed to the sensor’s sensitivity to surface dielectric properties, which correlate with soil moisture and aid in distinguishing floodplains from other units [14,18,68]. Furthermore, the differential backscatter response and penetration capability of SAR signals allow for the discrimination of units based on surface texture, which is intrinsically linked to material grain size—a property influenced by erosional, transport, and depositional processes [13,15,18,51].

The integration of L-band SAR imagery with Digital Elevation Models (DEMs) and their derivatives constitutes a highly effective data source that complements traditional, visually based geomorphological mapping in the Orinoquia, which has historically relied on optical imagery and aerial photographs [4,18,19,59,63,65]. This digital methodology reduces the subjectivity inherent in visual interpretation by implementing coherent, reproducible procedures where the derivation of quantitative metrics from DEMs is essential [26]. The use of multidimensional data enables innovative technological approaches that leverage diverse data types to address complex classification challenges [31]. Increasing the number of input variables within these models enhances overall accuracy and reduces classification variance [22,68,69].

Random sampling of training data ensured objectivity for both confusion matrix estimates and the Kappa statistic [70,71]. However, during sample definition, certain rare classes with limited spatial distribution were identified. While sampling was proportional to class representativeness, future studies could employ stratified sampling designs based on existing thematic cartography to guarantee the accurate inclusion of all classes regardless of their spatial extent [21,55]. Beyond its classification performance, the RF technique provides a valuable ranking of variable importance for class discrimination. This feature is particularly useful for refining the variable set or incorporating new data in subsequent studies [23,72].

4.1. Foothills

The geomorphology of the Foothills is fundamentally controlled by thrust faults along the tectonic boundary separating the cordillera from the alluvial plains. This structural configuration produces an abrupt slope break, which drives the rapid dissection, transport, and torrential deposition of detrital material [37,58]. The extensive fluvial landforms observed—including accumulation terraces, alluvial fans, and floodplains—collectively reveal a complex system where erosion, sedimentation, and denudation processes interact continuously. These terraces were formed primarily by erosional and alluvial sedimentation across ancient floodplains during phases when depositional dynamics dominated over lateral or vertical dissection [37]. Their morphology has been further shaped by ongoing tectonic activity, rock weathering, denudation, basement formation, and intense regional shear deformation [39,40]. Consequently, the high sediment supply promotes frequent bed aggradation in river channels, leading to the development of anastomosing channel patterns [37,58]. Similarly, active alluvial fans, particularly in the northern Foothills, exhibit intense erosive processes [39,40]. This geomorphic framework underscores the combined influence of neotectonics and historical climatic variability on landscape evolution in the region.

The torrential sediment load derived from the mountain range presents both current and potential instability hazards, highlighting the critical importance of geomorphological information for environmental management. Such knowledge contributes directly to the following: (i) mitigating natural risks associated with erosion, sedimentation, and fluvial dynamics; (ii) managing watersheds and conserving ecosystems through restoration, preservation, and recreational strategies; and (iii) planning sustainable productive activities. In this context, geomorphological analysis provides essential parameters for understanding landscape dynamics and sensitivity, thereby clarifying the processes that sustain and influence ecological systems [73,74].

4.2. La Macarena

La Macarena constitutes a Precambrian rock massif with a significant influence on orogenic processes. Its present geomorphological configuration results from the dynamic interplay between tectonics, denudation, and fluvial sedimentation. With an elevation of 1598 m and extending over 130 km, this massif is associated with the Guiana Shield and was subsequently overlain by Andean sediments [43]. Its origin is linked to crustal rifting, the exhumation of basement rocks, and structural deformation within a positive flower structure. Geological ages range from the Precambrian to the Tertiary, reflecting orogenic processes such as uplifts and folding associated with the Andean orogeny [39,40].

Three primary morphogenetic environments characterize its geomorphic diversity. The structural environment, although occupying the smallest area, is foundational. It reflects the region’s tectonic history through structural hillslopes, revealing a past of uplift and deformation that has fundamentally shaped the massif’s architecture [39,40]. The denudational environment predominates, covering nearly half of the surface. This intense geomorphic activity is responsible for highly dissected hills and erosional slopes, producing a landscape of pronounced altitudinal contrasts and a complex drainage network. The fluvial environment, encompassing most of the remaining area, features floodplains and ancient alluvial fans. These landforms record the continuous sediment flux from the adjacent cordillera, a process that has facilitated the development of accumulation terraces and significantly transformed the landscape. This geomorphic triad—structural, denudational, and fluvial—highlights a landscape where active tectonics and surface processes interact continuously. The structural framework, driven by near-surface deformation and basement uplift [39,40], provides the template upon which intense denudation and prolific fluvial sedimentation operate, resulting in the region’s distinctive and diverse topography.

4.3. Floodplain

The Floodplain occupies a lateral structural depression or sedimentary basin east of the Foothills, an area characterized by active subsidence. This basin was initially infilled with Cretaceous to Tertiary epicontinental sediments, including interbedded volcanic materials [37,43,58]. The contemporary landscape is overwhelmingly shaped by the fluvial environment, which covers more than half of the surface. This dominance highlights the prevalence of overbank flooding and lateral accretion processes, which have generated a mosaic of landforms including active floodplains and accumulation terraces. The capacity of these fluvial systems to remodel the terrain is now spatially limited, reflecting the intricate dynamics of a mature drainage network operating within a subsiding basin. Aeolian processes have also played a significant role in configuring the surface. Extensive loess sheets and dune fields attest to past and present wind activity, intensified by cyclical alternations between dry and humid climatic periods. The morphology of these aeolian deposits is subsequently modified by surface runoff, which promotes rill formation and sediment deposition in interdune areas [37,58,63,64,75].

4.4. High Plains

The High Plains originated through a complex tectonic history involving the compression and subsequent rupture of the proto-oceanic crust, followed by the uplift of Tertiary sedimentary sequences [39,40]. This structural framework established the fundamental architecture of the region. A major eastward-tilting regional fault, deeply incised by transverse drainage channels, defines the western boundary of the High Plains and shapes its morphology into a large, dissected alluvial fan. The landscape is interrupted by isolated sierras, a pattern explained by their genesis over uplifted blocks of Precambrian basement rock [39,40]. The resulting geomorphic configuration is a mosaic of low hills, rounded knolls, and slopes of varying gradients, interspersed with depressions and sinuous valleys superimposed on broader domal structures [43].

The sedimentary record indicates extensive aggradation from the Andean cordillera to the Orinoco Basin throughout the Tertiary and Pleistocene. This phase culminated in the Pleistocene with the development of the prominent fault along which the modern Meta River flows, which now forms the definitive physiographic boundary separating the High Plains from the eastern Floodplain [37,58,60,65]. The presence of this fault underscores the profound influence of deep-seated geological processes on contemporary landscape partitioning and regional hydrology. The High Plains are predominantly a denudational landscape, where prolonged erosion has sculpted terrain ranging from gently rolling colluvial plains to highly dissected relief. This topography is largely the result of a long history of Andean-sourced sedimentation subsequently modified by pervasive weathering and erosive processes. Fluvial and aeolian environments, while less extensive, contribute significantly to the region’s geomorphic diversity. Floodplains and accumulation terraces record the persistent role of fluvial systems, while extensive loess mantles attest to the importance of wind dynamics and broader climatic patterns in sediment redistribution and surface shaping over time.

4.5. Regional Considerations

The preservation of regional geomorphological and edaphic settings is paramount for sustaining the ecological processes that govern the structure and dynamics of natural vegetation. These landforms and soil systems regulate hydrological and nutrient cycles while providing the foundational substrates that determine species distribution, community composition, and ecosystem persistence. Notably, geomorphic heterogeneity increases microenvironmental variability, creating ecological niches that support high biodiversity. Concurrently, edaphic properties shape the functional traits of plant assemblages and their adaptive capacity to environmental stressors [76]. Consequently, the integrity of these geomorpho-pedological units is essential for ecological stability, as their degradation could initiate cascading effects on vegetation cover, soil fertility, and hydrological regimes. From a conservation standpoint, protecting these physical–biotic interactions ensures the persistence of natural vegetation mosaics and maintains their critical functions in carbon sequestration, habitat connectivity, and long-term ecosystem service provision. Therefore, integrating geomorphological and edaphic conservation strategies into regional planning is fundamental for safeguarding the ecological functionality and robustness of these landscapes under current and future environmental pressures [73,74,77].

5. Conclusions

The spatial extent of the Orinoquia region in Colombia was delineated using precise geomorphological criteria, most notably the abrupt slope gradient along the Foothills, which provides a clear physiographic boundary separating the Andean domain from the Orinoquian plains. This topographic break profoundly influences river morphology and sediment transport from the cordillera, reflecting a complex interaction between regional tectonics and coupled erosion–sedimentation dynamics, further modulated by underlying crustal flexure in the region’s southern sector. This study establishes a precise delimitation of the region’s four constituent physiographic units—Foothills, High Plains, Floodplain, and La Macarena massif—and details their characteristic geomorphological environments. The structural environment, though limited in spatial extent, is significantly represented in the Foothills and La Macarena. The fluvial environment predominates in the Foothills and Floodplain, the denudational environment is most prominent in the High Plains and La Macarena, and the aeolian environment forms a significant component of the Floodplain.

Methodologically, this work demonstrates the substantial potential of an integrated digital approach for regional-scale geomorphological mapping. Random Forest modeling successfully classified stacked, multidimensional imagery into discrete and representative units corresponding to predefined field-validated categories. Model validation via confusion matrices and associated metrics confirmed acceptable accuracy levels, with the highest misclassification rates occurring among fluvial units. The strong agreement between overall accuracy values and Kappa indices confirms that correct pixel allocation is systematic and non-random, underscoring the model’s robustness.

Finally, the conservation of these geomorphological and edaphic systems is not merely the protection of abiotic structures but a fundamental prerequisite for sustaining the region’s ecological integrity. Maintaining the integrity of soil–landform interactions preserves the functional heterogeneity that supports species persistence, regulates biogeochemical cycles, and ensures long-term ecosystem persistence. The degradation of these foundational templates would directly compromise vegetation structure, ecological stability, biodiversity, and ecosystem service provision. Therefore, prioritizing their conservation represents a strategic imperative for both ecological science and sustainable environmental management in the Orinoquia.