GPR in the Maya Lowlands

GPR has been used by archaeologists worldwide for over 40 years to guide physical excavations. However, in the Maya area, particularly in the Northern Maya Lowlands, GPR has had difficulty serving as an accurate remote sensing tool for generating data and insights on landscape use and transformation. In the broader Maya area, GPR surveys are typically conducted using either a low- or high-frequency single-antenna unit (ranging from 100 to 500 MHz), which is pulled or pushed on a utility cart (Barba et al. Reference Barba, Ligorred, Blancas and Ortiz2019:251; Munro-Stasiuk and Manahan Reference Munro-Stasiuk and Manahan2010:129). Variations in antenna frequency and survey line spacing result in different resolutions in the recovered data (Conyers Reference Conyers and Johnson2006:135). Antenna frequency and line spacing should be calibrated to the specific research goals and methodologies used in the comparative GPR surveys conducted at distinct built environments in the broader Maya area.

Demonstrating variation in research design and methodologies, GPR surveys have effectively verified suspected archaeological remains of subsurface buildings (Barba et al. Reference Barba, Ligorred, Blancas and Ortiz2019:247; Chinchilla Mazariegos Reference Chinchilla Mazariegos2018:465; Dunning et al. Reference Dunning, Mccane, Swinney, Purtill, Sparks, Mann, Mccool and Ivenso2015:183; Pérez-Pérez et al. Reference Pérez-Pérez, de Tapia, Barba-Pingarrón, Gama-Castro and Peralta-Higuera2012:117; Skaggs et al. Reference Sheldon, Powis, Rucker and Micheletti2016:1), detected deep soil pockets that potentially illuminate agricultural land use and water management (Carpenter et al. Reference Carpenter, Adams and Lenczewski2013:232; Munroe-Stasiuk and Manahan Reference Munro-Stasiuk and Manahan2010:123), and located ancient agricultural fields (Dixon Reference Dixon2006:172; Sheets et al. Reference Sheets, Dixon, Guerra and Blanford2011:5). This research is complemented by the search for archaeological anomalies (Sauk and Desmond Reference Sauk and Desmond1999:4–7; Valdés and Kaplan Reference Valdés and Kaplan2000:337). While the results of several investigative surveys were indeed successful, the scope of application tended to be limited, focusing more on exploring isolated features, salvage, and cultural resource management rather than visualizing hidden ancient landscapes (see also Barba, Blancas, Ortiz, and Carballo Reference Barba, Blancas, Ortiz and Carballo2009:17; Barba, Blancas, Ortiz, and Ligorred Reference Barba, Blancas, Ortiz and Ligorred2009:27; Valdés and Kaplan, Reference Valdés and Kaplan2000:329). Consequently, GPR use in the broader Maya area has been regarded as a promising small-scale and independent remote sensing technique.

Yet, larger-scale GPR surveys are not entirely without precedent within the broader Maya area. At the Maya site of Nim Li Punit in Belize, a nearly 4,000 m² GPR survey was divided into 13 distinct grids of varying sizes. The survey was conducted using a single SIR-3000 Subsurface Interface Radar Unit equipped with a 400 MHz shielded antenna (Daniels Reference Daniels2014:32). This extensive survey revealed many promising features in the urban core of Nim Li Punit. However, while exploring partial plazas both on and off acropolis groups, the researchers faced significant challenges in fully representing probable architectural features and in distinguishing anomalies of archaeological interest from those potentially caused by ground-surface conditions and tree roots (Daniels Reference Daniels2014:31). Nonetheless, this study demonstrates promise for large-scale GPR surveys in the Maya region, albeit in a relatively cleared tourist setting.

In a much larger survey, investigations at El Baúl, Guatemala, initially covered four sectors of the site, each roughly a hectare in size, surrounding the central acropolis using GPR and magnetometry (Chinchilla et al. Reference Chinchilla, Lipo, Neff, Gómez, Cosenza, Reyes, Bradford, Safi, Plaskey, Laporte, Arroyo and Mejía2007:843). The broad survey of this project uncovered rectangular subsurface features, some of which were confirmed as architectural walls through excavation (Chinchilla et al. Reference Chinchilla, Lipo, Neff, Gómez, Cosenza, Reyes, Bradford, Safi, Plaskey, Laporte, Arroyo and Mejía2007:848). Following the initial GPR survey, an expanded area of 124,000 m² was covered during the 2006, 2007, and 2008 field seasons, revealing an extended view of a causeway (portions of which had been previously excavated in 2001) and linear features that likely represent subterranean building walls (Safi et al. Reference Safi, Mazariegos, Lipo and Neff2012:418). Furthermore, the GPR surveys at El Baúl revealed several subsurface anomalies ranging from suspected buildings to platforms and primary deposits, illustrating the potential of GPR for remote surveys.

Significantly, El Baúl is not located within the karst Maya Lowlands but rather in the volcanic highlands along the Pacific coast of southern Guatemala. As with the success of GPR surveys at Kaminaljuyu (Valdés and Kaplan Reference Valdés and Kaplan2000), the success of the investigations at El Baúl can, in part, be attributed to local geological conditions. The basalt boulders used in construction stand out against the surrounding fill in the survey results.

Finally, at the site of Pacbitun, Belize, a broad GPR grid survey was conducted at Plaza A, covering an area of approximately 50 × 50 m (Skaggs et al. Reference Sheldon, Powis, Rucker and Micheletti2016:5). Aware of the obstacles faced in other surveys, Skaggs and his colleagues confidently labeled the GPR subsurface survey of Plaza A a success—revealing features and subsurface buildings that were later confirmed through excavation. In the published results of their survey, Skaggs and colleagues conclude:

While the use of GPR in the Maya Lowlands presents certain challenges, the case studies of grid surveys at the sites of Nim Li Punit, El Baúl, Pacbitun, and Yaxuná (discussed below) show promise. GPR’s future is not as an independent method for remotely sensing small-scale features, even though its value in this regard is well-established in archaeology. Rather, the prospects for GPR surveys lie in serving as an augmentative lens for remotely sensing landscapes.

Central Plaza GPR Survey at Yaxuná, Yucatán, Mexico

The Northern Maya Lowland site of Yaxuná is located in the tropical dry broadleaf forest ecoregion of Yucatán, Mexico. Geologically, northern Yucatán features a semi-arid karst plateau that largely lacks both thick soils and surface water (Munro-Stasiuk and Manahan Reference Munro-Stasiuk and Manahan2010:125), except for intermittent sinkhole features such as rejolladas (round depressions with thicker soils) and cenotes (collapsed depressions where the water table is accessible). At its peak, Yaxuná covered a dense area of roughly 1 km², approaching urban proportions (Stanton and Collins Reference Collins2022:115).

From 2013 to 2016, investigations at the Yaxuná Central Plaza were conducted by Proyecto de Interacción Política del Centro de Yucatán (PIPCY; Stanton and Collins Reference Collins2022:99). During this period, PIPCY researchers excavated 113 units, each measuring 2 × 2 m (Figure 1). A total of 33 excavated units reached bedrock, providing extensive data on the Central Plaza’s construction sequence and stratigraphy.

Figure 1.

Lidar map of the Yaxuná settlement. Map by author; lidar data courtesy of Travis W. Stanton.

The Central Plaza exemplifies an E Group, a category of pyramid plaza complexes that is significant because researchers associate their presence with early mound-building traditions across the Maya Lowlands (e.g., Freidel et al. Reference Freidel, Chase, Dowd and Murdock2017). Morphologically, most E Groups exhibit an east–west orientation: a pyramid typically bounds a plaza on the west, while a long raised platform delineates the east. Notably, E Groups broadly share numerous features, including similar architectural patterns, dimensions, and ritualistic activities (such as caching) throughout the Maya Lowlands and primarily date to the Middle and Late Formative Period (1000 BC–AD 250). As a result, investigations of E Groups benefit from extensive comparative data that can enhance project design and the interpretative process.

The Central Plaza at Yaxuná features a relatively uniform flat surface measuring approximately 100 m from east to west and 100 m from north to south. Given the Central Plaza’s antiquity, which represents the oldest known occupation yet detected at Yaxuná, with likely origins around the ninth century BC (Stanton et al. Reference Stanton, Góngora, Collins, Slater and Walker2023:434), this space underwent some of the most extensive architectural modifications at the site. Excavations in the eastern portion of the Central Plaza revealed significant features and 11 phases of floor construction. Because the western portion of the Central Plaza remained largely unexcavated, the research team developed a GPR survey plan to remotely image and map subsurface potential features that may illuminate ritual activities, civic developments, and changes in building traditions throughout the entire plaza.

The goals of the GPR survey in the Yaxuná Central Plaza were threefold. First, in the eastern portion of the Central Plaza, the GPR survey aimed to prospect for promising anomalies that could guide ongoing physical excavations already at the sixth stratigraphic level of 11, thereby reducing time investment and maximizing the research study’s objectives. The second aim of the GPR survey was to detect known or suspected continuations of features in the unexcavated western portion of the Central Plaza that had been identified through previous investigations. Demonstrating the ability to detect archaeologically inferred features and stratigraphy would provide a methodological baseline for assessing and interpreting the GPR survey data within the culturally modified landscape. The final goal, also in the western portion of the Central Plaza, was to inform the planning of investigations for the subsequent field season. To achieve these objectives at Yaxuná, GPR data collection focused on two areas within the Central Plaza.

GPR surveys were conducted during the rainy season from June 25 to June 29, 2015, at the Central Plaza, using a GSSI SIR-3000 GPR system with a 400 MHz central-frequency antenna (Figure 2). The system was mounted on a utility cart with a distance-encoding wheel, triggering data collection every 2 cm. To set the average dielectric value using the SIR-3000 interface, hyperbolic matching was used. This was accomplished by identifying hyperbolas along the anticipated continuation of a targeted subsurface feature previously detected in nearby excavations. The strategy of selecting areas of suspected subsurface disturbances at anticipated depths to set the dielectric average was intended to help calibrate velocity to accurately represent depth. The maximum survey depth was between 1.5 and 2 m below the surface.

Figure 2.

Map of E Group excavation units with GPR survey area showing anomalies. Note units 14N-30W, 12N-30W, and 10N-30W, where excavations tested in 2016 detected anomalies. Map by the author. (Color online)

Figure 2 Long description

A map depicting excavation units and a Ground Penetrating Radar (GPR) survey area on a textured terrain. The excavation units are marked with outlined squares, while the GPR survey area is highlighted in a shaded rectangle. Labeled sites include ′WaterPipe Survey′ and ′Expedition GPR Survey′ among others. A scale bar indicates distances from 0 to 100 meters. A north arrow is present for orientation. The legend identifies excavation units and the GPR survey area with distinct symbols.

At the Central Plaza, geophysical grids were established using surveyors’ tape measures, and individual transects were obtained at 50 cm intervals. Grid corners were recorded in latitude/longitude using a handheld GPS unit. Markers were placed at the start and end positions of the individual transects. These markers were then mapped with a GPS unit that had a nominal horizontal accuracy of 1.5–2.0 m. For both gridded and transect data, GPS positions were either incorporated into the file header of GPR files using GSSI’s RADAN 7.0 software or geo-rectified in ArcGIS to match GPS positions. This process enabled the export of important stratigraphic layers and time slices in a geospatial format. Post-processing routines for the GPR data were performed in RADAN 7.0 software, including position correction, background removal, depth correction (migration), and high and low pass filtering.

The survey in the Central Plaza was conducted along its central east–west axis and was bounded on the west by a previous 2 m east–west by 4 m north–south excavation abutting Str. 5E-1, a large pyramidal platform. On the east, the survey was bounded by previously excavated units linearly investigated from 10 m north to south in the plaza center. Because the central axis and southern portion had been the target of investigations for several years, these areas were relatively cleared of large trees and dense vegetation that would otherwise have obstructed the survey. The area of the plaza north of the central axis, however, had not been extensively studied, as it was less linearly defined. Dense vegetation in the northern portion of the plaza obstructed the survey, resulting in an uneven distribution.

East Plaza Anomaly Detection and 2015 Excavation Results

In 2015, our GPR survey at Yaxuná focused on two areas within Central Plaza (Figure 3). The first area was located in the central-eastern portion of the plaza, along an exposed Middle Formative plaza floor, the sixth in sequence from the topsoil. Radiometric testing and ceramic analysis date the exposed surface of Floor 6 to between 650 and 350 BC (Collins Reference Collins2023:147). This block excavation took place within units 12N-4W, 14N-4W, 12N-6W, and 14N-6W in a flat area measuring approximately 4.3 × 3.5 m at a depth of 1.43 m below the surface, where an incised circular feature was uncovered on the dense sascab (a durable product made from pulverized limestone, often used to create flat, walkable surfaces on and around architecture, including plaza floors) floor (Collins Reference Collins2022:649; Stanton et al. Reference Stanton, Magnoni, Ardren, Burbano, Cariño, Carrillo and Casanova2016:257). Investigations initially paused on this surface due to time and budget constraints, allowing the prospective GPR survey to commence.

Figure 3.

Interpreted anomalies and depths below the surface detected in the initial results from the GPR survey in the Yaxuná central plaza during 2015. Map by the author and Peter Leach. (Color online)

Figure 3 Long description

A map detailing the West Plaza and Excavation Block surveys, indicating various depths below the surface. The map includes annotations for depths such as 30 cm, 50 cm, 60 cm, 70 cm and 103 cm below the surface. Specific areas are marked with different colors, representing GPR survey and previously excavated units. The map also includes labeled units: 12N-20W, 12N-30W and 13N-30W. A scale bar at the bottom indicates distances in meters and a north arrow is present for orientation.

The purpose was to use GPR to identify targets or anomalies of interest, then decide whether to backfill the unit or continue excavating. Survey data was interpreted in the field, and at least one line exhibited a discrete stratigraphic break filled with coarse materials. A total of 14 lines were collected in the vicinity of the excavation block in the eastern portion of the plaza. These profiles revealed a continuation of the floor construction sequences beneath the excavation base. Data were interpreted in the field, with at least one line showing a discrete stratigraphic break filled with coarse materials at a depth just below the floor surface and a second at 50 cm beneath the surface (Figure 4). This anomaly was consistent with results corresponding to a pit or trench excavation through uninterrupted stratigraphy. The locations of the detected anomalies were marked in the field for further investigation.

Figure 4.

Interpreted GPR Profile from the eastern Yaxuná plaza survey excavation block. Figure provided by Peter Leach.

Figure 4 Long description

Distance along Profile (meters). Depth below Surface (meters). The image A showing a grayscale subsurface profile section. The top axis is labeled Distance along Profile (meters) with tick labels 0.0, 0.5, 1.0 and 1.5. The vertical axis is labeled Depth below Surface(meters) with tick labels 0.0, 0.3 and 0.6. The plotted field contains continuous to broken, wavy, alternating light and dark bands across the full distance range 0.0 to 1.5 and depth range 0.0 to 0.6. Banding is more irregular and closely spaced at depths nearer 0.6 than at depths nearer 0.0. The image B showing an interpretive line plot of the same subsurface profile, using curved boundary lines to mark units. The top axis is labeled Distance along Profile (meters) with tick labels 0.0, 0.5, 1.0 and 1.5. The vertical axis shows tick labels 0.0, 0.3 and 0.6. Text labels identify regions: Filler Preparation Materials near shallow depths close to 0.0; Coarse Fill Units at shallow to mid depths around the 0.3 tick; Weakly Stratified Fill Units spanning mid depths around the 0.3 to 0.6 interval; and Small Steep Cut marking a narrow, steep-sided feature bounded by closely spaced curved lines. The Small Steep Cut label is placed between the 1.0 and 1.5 distance ticks and between the 0.3 and 0.6 depthticks.

To test the accuracy of GPR survey data, excavations were reopened in a confined area within unit 12N-4W, just south of the incised line feature. In a 1 × 1 m space in the northwest quadrant of 12N-4W, 2 cm of the floor and compacted sascab material were removed before encountering a thin layer of fill (no more than 10 cm) covering Floor 7. After this was removed, a portion of Floor 7 was revealed, along with a square cut into the surface (roughly 30 × 30 cm), which was filled with loose rubble stones up to the level of the floor. After removing these loose rubble stones, a rounded capstone (roughly 25 cm in diameter) sealed to the surface of Floor 7 was uncovered (Figure 5). The anomaly was exposed as a sealed primary deposit cutting through two floors, containing polished iron pyrite, a partial ceramic jar, and ample carbon, indicating in situ burning (Collins Reference Collins2023:151; Stanton et al. Reference Stanton, Magnoni, Ardren, Burbano, Cariño, Carrillo and Casanova2016:259). Due to this deposit and potential similar ones, excavations expanded to units 12N-4W, 12N-6W, 14N-4W, and 14N-6W, revealing several additional primary deposits. Furthermore, continued investigations in this area identified a linear course of stone intersecting with a circular course of stones, representing what I believe to be the partial foundation of a small Middle Formative private structure in an otherwise public place (Collins Reference Collins2023:151).

Figure 5.

Eastern Yaxuná plaza survey area excavation block; upper left, GPR survey block area on Floor 6 of units 14N-6W, 14N-4W, 12N-6W, and 12N-4W; upper right, an intentional cut in the surface of Floor 7, which has been resealed with rubble fill; lower left, a capstone sealed under the rubble of the resealed aperture in Floor 7; lower right, material contents beneath the capstone, including a large rim sherd, carbon, and a polished fragment of hematite. Photos by the author. (Color online)

West Plaza Stratigraphy, Anomaly Detection, and 2016 Excavation Results

Anomaly Detection

The widespread and somewhat consistent fill layers in the Central Plaza were disrupted in some areas by high-amplitude reflectors within the fill layers, often displaying clear geometrical shapes and alignments with nearby features of similar amplitudes. Such disruptions in otherwise consistent stratigraphy were interpreted in the field as possible anomalies (Figure 6). One detected possible anomaly was the clear and extensive continuation spanning east to west across the excavation block, seemingly coinciding with a space in alignment with a partially excavated highly compacted sascab causeway (or sacbe; ancient causeway, translating to “white road” in Yucatec), first observed in 2013 investigations at a depth of 10 to 15 cm below the surface.

Figure 6.

GPR time slices from surface to 110 cm showing changes in stratigraphic features with depth. Figure by the author.

Figure 6 Long description

Twelve separate ground-penetrating radar time-slice maps are arranged in a grid, each representing a horizontal depth slice at a distinct depth interval. The depth labels visible on each slice are 0 centimeters, 10 centimeters, 20 centimeters, 30 centimeters, 40 centimeters, 50 centimeters, 60 centimeters, 70 centimeters, 80 centimeters, 90 centimeters, 100 centimeters and 110 centimeters. Each slice shares the same plan-view spatial extent, representing the surveyed ground surface area. A scale bar labeled in the legend area indicates a reference length, allowing approximate spatial measurement of features within each slice. The legend is labeled Ground-Penetrating Radar Time Slice: Ground Surface. Pixel intensity in each slice represents radar wave amplitude, where lighter tones correspond to higher amplitude reflectivity and darker tones correspond to lower amplitude reflectivity. At 0 centimeters and 10 centimeters depth, the amplitude pattern is relatively uniform with diffuse variation across the survey area. From 20 centimeters to 50 centimeters depth, concentrated zones of elevated amplitude become visible, appearing as distinct lighter patches distributed across the slice area. These high-amplitude zones are most pronounced and spatially continuous across the 40 centimeters to 70 centimeters depth range, where lighter patches span approximately 3 to 4 meters in length based on the scale bar, suggesting persistent subsurface features across this interval. From 80 centimeters to 110 centimeters depth, the high-amplitude zones gradually reduce in spatial extent and intensity, with the 100 centimeters and 110 centimeters slices showing more diffuse and lower amplitude patterns overall. The spatial layout of each slice follows a consistent orientation, with a terrain profile visible along the upper edge of each map representing the ground surface relief.

Previous excavations identified portions of a buried sacbe that appeared to bisect the plaza along the central east–west axis (Collins Reference Collins2022:653). Investigations detected the sacbe throughout the Central Plaza along the 14N excavation grid, connecting the presumed western terminus at Str. 5E-1 (evidenced in unit 14N-60W), the plaza (evidenced in units 14N-26W, 14N-24W, 14N-22W, and 14N-20W), and the plaza’s eastern terminus at Str. 5E-6 (evidenced in units 14N-6W and 14N-4W), spanning a distance of 60 m (Stanton et al. Reference Stanton, Magnoni, Ardren, Burbano, Cariño, Carrillo and Casanova2016:120). As a result, the investigation team hypothesized that a GPR survey in the plaza focused on remotely sensing suspected subsurface features (like the previously identified sacbe) and unknown features (as anomalies) could enhance our understanding of the plaza with minimal investigative destruction while also informing future excavations of potential subsurface structures.

It should be noted that the level of Floor 3 lightly covered the height of the sacbe, an earlier Floor 5 construction. As a result, it was initially challenging to visually differentiate Floor 3 and Floor 4 from the sacbe feature through excavation (Stanton et al. Reference Stanton, Magnoni, Ardren, Burbano, Cariño, Carrillo and Casanova2016:120). However, the sacbe was distinctly composed of, and in some areas encased in, a deep reddish-orange sascab that had a much greater density than the surrounding floor surfaces. In fact, in some instances, the sacbe appeared to obstruct tree roots that seemed to follow the surface contours of the feature.

Approximately 2 m south of the suspected continuation of the sacbe, the GPR data indicated a linear arrangement of geometrical anomalies. At a depth of 30 cm below the surface, a circular anomaly was identified. At 75 cm below the surface, these anomalies were replaced by a roughly rectangular feature (Figure 7). The stratigraphy of these anomalies was documented in Profile 62, which coincided with later excavations in units 14N-30W, 12N-30W, and 10N-30W (Figure 8). Although the complex stratigraphy from successive construction efforts complicates the analysis of this plaza area, the strong high-amplitude anomaly patterns and clear stratigraphic breaks suggest the presence of significant buried archaeological features. Supporting this interpretation is the fact that previous excavations uncovered several architectural anomalies at 30 cm below the surface (Stanton and Collins Reference Collins2022:116) and between 70 and 150 cm below the surface along the Central Plaza’s axial line in the patterned stone fill under Floor 5 (Collins Reference Collins2022:652–653).

Figure 7.

Left, a 30 cm time slice above the field interpretation revealing a circular anomaly; right, 75 cm time slice above a field interpretation showing the circular anomaly transformation into a rectangular anomaly, along with the area surveyed with Profile 62, located at units 14N-30W, 12N-30W, and 10N-30W; note, the circular anomaly feature within the rectangular platform was confirmed through excavation. Figure by the author and Peter Leach. (Color online)

Figure 7 Long description

The image A showing a rectangular ground-penetrating radar slice labeled ′Plaza GPR Slice: 30cm′ with a grayscale texture field, a horizontal axis along the bottom with tick labels ′0′, ′50′ and ′100′ and a vertical scale on the right with tick labels ′0′, ′20′ and ′40′; a single vertical red line runs from the top edge to the bottom edge near the right side of the slice. The image B showing a second rectangular ground-penetrating radar slice labeled ′Plaza GPR Slice: 75cm′ with a similar grayscale texture field, a horizontal axis along the bottom with tick labels ′0′, ′50′ and ′100′ and a vertical scale on the right with tick labels ′0′, ′20′ and ′40′; a single vertical red line runs from the top edge to the bottom edge near the right side of the slice. The image C showing a plan-view line drawing labeled ′Current Excavation Area′ with a large rectangular boundary, an irregular rounded outline on the left side inside the rectangle and a cluster of smaller straight-edged shapes near the lower right inside the rectangle; a vertical red line passes through the rectangle near the right side and crosses the cluster of shapes. The image D showing a second plan-view line drawing with the text ′Profile 62 End′ near the upper center and ′Profile 62 Start′ near the lower center, with a large rectangular boundary, a long narrow slanted rectangle near the upper left inside the boundary and a stepped angular outline near the right side inside the boundary; a vertical red line passes through the rectangle near the right side and crosses the stepped angular outline.

Figure 8.

Profile 62 displays the survey anomalies observed in units 14N-30W, 12N-30W, and 10N-30W, demonstrating continuity with the excavation stratigraphy. Figure by the author and Peter Leach. (Color online)

Figure 8 Long description

Ground-penetrating radar profile with radar amplitude plotted against horizontal distance. The x-axis represents distance in meters from 0 to 13. The y-axis represents depth or two-way travel time, ranging from 0 to 1.25. The radargram displays wavy horizontal bands indicating subsurface reflections and stratigraphy. Key interpreted regions include a rubble area from 0 to 3 meters, a mounded area from 3 to 9 meters and another rubble area from 9 to 13 meters. An attenuation zone, marked by low amplitude, is noted beneath the mound, suggesting poor signal penetration. The lower panel provides an interpretation sketch, labeling finer-grained fill units, rubble areas and a stratigraphic boundary. The interpretation includes uncertainty, indicated by question marks and highlights the attenuation area as likely due to a buried prepared surface.

These observations, combined with other geometrical patterns of high-amplitude anomalies in the western portion of the plaza from the GPR survey results, made the anomaly south of the presumed sacbe continuation a strong candidate for ground-truthing. Since GPR data are remotely sensed, archaeological excavation is necessary to determine the origins of the anomaly and cross-reference complementary datasets. The likely sacbe continuation and nearby geometric anomalies detected through the GPR survey informed the targeted three-week excavation in the summer of 2016.

2016 Excavation Results

Excavations in 2016 targeted the possible continuation of the rectangular geometric anomalies across three 2 × 2 m units spanning north to south. Units 10N-30W, 12N-30W, and 14N-30W were chosen to maximize the lateral exposure of the anomalies observed between the surface and 150 cm below it. This approach began with excavations in 14N-30W, quickly confirming a portion of the suspected buried sacbe in the northern section of the unit, roughly 60 cm below the surface.

After confirming the presence of the sacbe, as visualized through GPR, investigations moved southward to test the rectangular anomaly anticipated by the survey. Shortly after removing brush and exposing the topsoil in units 10N-30W and 12N-30W, a break in the upper three floor surfaces was revealed. Once fully exposed, the feature measured roughly 130 cm east–west and 85 cm north–south, appearing to be an intentional cut spanning the first five floors of the plaza’s stratigraphy down to the limestone fill ballast composing Floor 5. Distinct from the surrounding sediments, the feature contained loosely compacted, baseball-sized limestone cobbles and ash. Finally, three evenly spaced postholes surrounded the feature, further suggesting its intentional construction.

Once debris and sediment were cleared from the circular cut that passed through the upper five floors of the plaza, a small pile of similarly sized cobbles, roughly 10 cm in diameter, was mounded in the center of the feature (Figure 9). Upon their removal, a clear circle of cut stones with a 50 cm diameter was revealed. Excavations showed that this feature had a depth of three masonry tiers, extending to the surface of Floor 6, giving it a well-like appearance (Collins Reference Collins2018:271). Taken together, the circular feature of a cut floor terminating on circular masonry perfectly matched the circular anomaly detected by GPR, as visualized on the 30 cm timeslice. However, further investigations were needed to account for the shift to a more rectangular anomaly detected in the same space at 75 cm below the surface.

Figure 9.

Stratigraphy of Yaxuná’s west plaza anomaly detected by GPR after confirmation through excavation. of units 14N-30W, 12N-30W, and 10N-30W. Note that GPR-detected meaningful stratigraphy (evidenced by Slice 62) is noted on the left, juxtaposed with the constructed floor sequences (occupational surfaces) on the right. While the GPR stratigraphy does not entirely align with occupational surfaces, it does distinguish meaning material differences, beginning with the topsoil (1), the highly compact Floors 2–5 (2), the ballast of small stones under Floor 5 (3), the loose dry core fill boulders underneath the finely packed ballast (4) and the surface of Floor 6, a 20 cm thick phase resembling bedrock (5). Figure by the author. (Color online)

Figure 9 Long description

A cross-section diagram illustrating archaeological strata from the ground surface to a depth of 140 centimeters. At the top, labeled ′Floor Surface,′ is a layer of topsoil, followed by an ashen soil layer. Below this, a post hole is visible, penetrating through the ashen soil. Several horizontal layers are marked, including architectural features and fill stones, which are differentiated by color. The diagram shows a feature matrix centrally located, flanked by an earthen platform on the right. The depth is marked on the right side in increments of 20 centimeters, from 0 to 140 centimeters. The bottom of the diagram includes a scale indicating horizontal distances of 0, 1 and 2 meters. The legend at the bottom left identifies architectural features and fill stones by color.

Once the cobbles and rubble were cleared from the area and excavations reached the surface of Floor 5, a presumably rectangular earthen platform was revealed 70 cm below the surface. This rectangular feature, composed of a dense earthen and limestone cobble matrix, was distinct from the composition and compaction of the sediment that encased it—loose dry core fill of large limestone cobbles and boulders serving as the ballast for Floor 5. While the northern boundary of the feature was documented through excavation, its eastern, western, and southern expansions remain unconfirmed. However, after the feature was fully excavated down to the surface of Floor 6 within unit 12N-30W, an incised line feature was pecked directly onto the sascab floor surface, representing the northern boundary of the rectangular earthen platform feature, roughly 135 cm below the surface (Collins Reference Collins2022:653).

While it is clear that the GPR-detected anomaly south of the sacbe is, in fact, an archaeological feature, its potential function and purpose remain open to discussion. It is possible that the earthen platform surrounding this hollow may have served as a temporary stage during the early construction of Floor 5, suggesting a possible ritual theatre function in the plaza (Collins Reference Collins2023:152). If paired with Str.5E-6, a north–south spanning eastern platform, the anomaly may have temporarily created a smaller E Group within the larger E Group.

However, the dense composition of the sascab and the presence of ash within the circular masonry of the feature suggest a potential architectural function. According to Soledad Ortiz (personal communication 2021), the open platform resembles Maya lime kilns utilized to produce sascab floors (see Seligson et al. Reference Seligson, Ruiz and Pingarrón2019). A lime kiln within the plaza to support civic planning functions is plausible, given that Floor 5 represented the largest construction episode within the Central Plaza. While the analysis of the well-like feature continues, the interpretation of architectural and ritual functions is not mutually exclusive.

While the GPR survey concludes with Floor 6, at roughly 150 cm, the circular aperture and presence of additional features continued in this space. First, the feature maintained its presence through the 20 cm thick Floor 6, and excavations revealed a floor cut within this location on each of the floor surfaces clearly represented in this area down to the bedrock surface beneath Floor 11 (Collins Reference Collins2023:144). While interpretations of these features are documented in more depth elsewhere (Collins Reference Collins2022:653; Stanton et al. Reference Stanton, Taube, Collins, Freidel, Chase, Dowd and Murdock2024:165), the circular floor cuts and tiered stone masonry, paired with the rectangular platform of compacted earth, clearly matched the surveyed anomaly (Figure 10). Methodologically, the GPR survey enabled our excavations of several important ritual and civic features, dating to the potential founding of Yaxuná, which were minimally invasive and provided a template for future investigations.

Figure 10.

The west plaza area selected for testing, with GPR-surveyed anomalies showing southeast views of 14N-30W and 12N-30W; upper left, a circular cut through Floors 2, 3, and 4, made in antiquity, featuring mounded limestone fragments surrounded by three postholes; upper right, a circular arrangement of masonry stones beneath the mounded limestone fragments; lower left, the surface of the floor area, coinciding with Floor 5, which contains circular masonry stones with three intact tiers encircled by an earthen matrix with limestone fragments; lower right, circular masonry on Floor 6, cleared of the earthen mound matrix. Photos by the author. (Color online)

Discussion: The Yaxuná Survey in Context

While some aspects of GPR surveys have remained relatively consistent over the last four decades (e.g., the frequency and power of the antennas), other elements, such as digital interfacing and the ability to georeference three-dimensional data, have advanced significantly. For these reasons and due to the associated equipment costs, rather than soil composition, moisture, or forest coverage, Aitken (Reference Aitken2008:29) asserted in a master’s thesis that GPR in the Maya Lowlands should be regarded as an emerging technology. I concur with Aitken and recognize that considerable technological progress has been made in the last two decades.

As an emerging technology, GPR does not provide a completely clear view of the buried subterranean landscape, and much of the data obtained from a survey must be validated through physical investigation. The success of the GPR survey at the Yaxuná Central Plaza was demonstrated by detecting anomalies and confirming features through excavation rather than through their outright, remotely sensed identification. Despite limitations, the Yaxuná Central Plaza GPR survey successfully detected both anticipated and novel anomalous features that guided immediate and follow-up investigations. We must consider the factors that may have influenced the conditions producing the results at Yaxuná. In other words, which subterranean elements contributed to successful remote detection? While this remains speculative, we can hypothesize four factors that contributed to the positive feedback: the chronological, cultural, and material contexts, along with the composition of the anomalous feature confirmed in the excavation.

Underscoring the entire Yaxuná survey is the chronological context related to the Middle and Late Formative periods. These phases, particularly between 600 and 200 BC, begin with a significant degree of architectural experimentation, culminating in what is arguably a standardization of Late Formative construction traditions and techniques at the site (Collins Reference Collins2022:649). The increased density of plaza floor surfaces, combined with the Maya’s tendency to build directly over previous phases of construction, contributes to enhanced preservation and may facilitate easier stratigraphic detection through GPR.

Contributing to the cultural context is that large public plazas, like those in Middle and Late Formative E Groups throughout the Maya Lowlands, tend to have consistent human activities along their central axes (e.g., Aoyama et al. Reference Aoyama, Inomata, Pinzón and Palomo2017; Estrada-Belli Reference Estrada-Belli2006:59; Inomata et al. Reference Inomata, Triadan, Vázquez López, Fernández-Díaz, Omori, Belén Méndez Bauer and García Hernández2020:531). While the precise locations within plazas may vary, the forms of human activities remain fairly consistent, often involving the caching of important materials in intentionally resealed floor surfaces. These episodes of construction, destruction, and repair alter the compaction of the local stratigraphy and may be more easily detected through GPR survey when working under conditions of good preservation.

Earlier Middle Formative floors, such as those on levels 8, 7, and 6 (roughly dating between 600 and 400 BC), surveyed in the eastern section of the Central Plaza, display similar construction patterns featuring 5–20 cm of dense sascab surfaces poured over porous stone ballasts ranging from 10 to 35 cm thick. This relative uniformity of early plaza floor construction at Yaxuná, consistently composed of dense sascab, establishes nearly ideal conditions for GPR surveys to detect and distinguish occupation surfaces, concealed intrusions, and anomalies from a distance.

In contrast, Late Formative floor constructions, including Floors 5, 4, and 3 (dating roughly between 400 and 200 BC), surveyed in the western part of the plaza, were thinner, more durable, and stacked atop one another with minimal ballast separating them. The compaction of three thinner floor surfaces (approximately 5 cm of sascab each) likely contributed to their appearance as a single stratum in the GPR data. However, the large stone ballast beneath the sascab surface of Floor 5 appeared as two separate layers, likely due to the compaction separating large, loose limestone fill cobbles from the increasingly finer, granular fill necessary to support the smooth sascab surface above them.

Without excavation and understanding of the plaza stratigraphy, the relationship between the GPR survey data and the context of occupation surfaces might go unrecognized. Attesting to this observation, Barba and colleagues (Reference Barba, Ligorred, Blancas and Ortiz2019:253) have repeatedly documented the presence of three distinct meaningful strata beneath Mérida in Yucatán, Mexico, using GPR. Barba and colleagues define these layers as “the contemporary and colonial city, the Classic and Formative period Maya ruins, and the sterile limestone bedrock” (Barba et al. Reference Barba, Ligorred, Blancas and Ortiz2019:250).

Like the results from Mérida, the GPR survey in the western portion of the Central Plaza at Yaxuná revealed distinct stratigraphy that requires interpretation and understanding of subsurface construction. To a depth of approximately 1.5 m, the Yaxuná data showed a clear distinction among five strata throughout the western portion of the Central Plaza. It is important to note that five meaningful strata were observed in the survey data, revealing differences in material composition and layering processes related to density rather than distinct occupation surfaces.

As with the need to ground-truth features revealed through lidar, anomalies revealed through GPR also require verification. As Aitken (Reference Aitken2008:102) put it in the context of the plaza stratigraphy of Maya sites in Belize, without some knowledge of layering and site phases, there is some “guesswork” to remote sensing. This critique, however, should not imply that stratigraphy is not discernible through subsurface remote sensing. Rather, clarity and fine precision may be sacrificed for sensing small-scale distinctions in favor of larger-scale differences.

In addition to the confirmed stratigraphy and features in the plaza, several more geometric patterns of high-amplitude anomalies were observed and remain unexcavated. The strong patterning of high-amplitude anomalies and clear stratigraphic breaks hint at significant buried archaeological features that can inform future investigations. The GPR survey revealed a complex stratigraphy of successive construction efforts confined to different strata, complicating our understanding of the plaza’s use and function over time. While identifying archaeological features through remote sensing does not always provide a clear or straightforward path to interpretation, the results from our survey certainly challenge and expand our knowledge of the early development and transformation of urban life at Yaxuná. Although small-scale, these findings demonstrate the potential to enhance large-scale remote sensing techniques with GPR for uncovering concealed landscapes.