New Insights Into the Water Management System at Tetzcotzinco, Mexico

1.1 Background

Tetzcotzinco is one of the best-preserved sites with monumental architecture dating back to the Late Aztec Period (1400–1521 AD). The site is located east of present-day Mexico City, in the town of San Nicolás Tlaminca, municipality of Texcoco, State of Mexico (Figure 1), and is located on the hill of the same name. The archaeological park in Tetzcotzinco (Zona Arqueológica de Tetzcotzinco) was established in 2005 (Alcántara Onofre, 2002, pp. 52–53). Its boundaries also include the surrounding areas, including the nearby hill named Metecatl, together with the archaeological remains existing there (Lesbre, 2001, p. 325). This place is most often presented as the biggest pre-Hispanic palace-garden complex^[2] in the Basin of Mexico (Evans, 2000, p. 212). Nowadays, only the ruins of residential and ceremonial buildings and hydraulic features carved in rock (such as reservoirs and canals) are still standing. There is also a well-preserved complex of sculptures and petroglyphs contemporary to the period of the palace-gardens. A significant group of this type of objects is the one depicting the models of architectural structures carved in the rocks located on Metecatl (Domínguez Nuñez, 2007, p. 84). The time of largest growth of Tetzcotzinco falls in the fifteenth century and the reign of Nezahualcoyotl (1431–1472). He is attributed for the design and construction of the royal “garden” and ceremonial center after he became the ruler of the capital of the Acolhuacan territory, part of the Aztec Triple Alliance at the eastern part of the Lake of Texcoco (Anales de Cuauhtitlan, 1975, pp. 40–41) (Figure 1).

Figure 1

Tetzcotzinco archaeological site.

One of the most important projects of Nezahualcoyotl was the construction of a system for water management. It consisted mainly of a complex of canals and aqueducts leading fresh water from springs located around Mount Tlalocatepetl (Parsons, 1971, pp. 146–147). Archaeologists have also documented a system to control the quantity and speed of flowing water, as well as several water reservoirs, pointing out their religious significance in Tetzcotzinco (García García, 2007, pp. 136–204). No doubt the water management system had an important symbolic role, but the canals crossing the hills probably also played economic roles, supplying fresh water to farmlands and local settlements. Due to its location, hydraulic features, and historical sources, it has been suggested that the site was part of a regional network of canals connecting the most important cities throughout the Acolhuacan area (Barlow & McAfee, 1946, p. 110).

Tetzcotzinco’s monumentality has attracted scholars for decades. The first attempts to map the site were made at the turn of the 19th and 20th centuries when researchers prepared the first plans of Tetzcotzinco’s preserved architecture (Reyes, 1888, pp. 140–141; Mendizábal, 1946). In the subsequent decades, mainly thanks to the studies of Parsons (1971) and Townsend (1982), subsequent structures were discovered with new interpretations of their significance. The most extensive excavations so far, covering the central part of the site, were carried out by Hernández (1993) and García García (2007).

These studies have touched upon the water management system of the site. Parsons (1971, pp. 146–149) proposed new maps of canals based on the archaeological survey and studies of the local topography. This topic reached later discussions on Mesoamerican waterworks, like in the works of Palerm and Wolf (1955), Peña Santana, and Levi (1989, pp. 20–25). Particularly, the latter extended Parsons’ plans with aerial photos, and Doolittle (1990, pp. 127–135) and Rojas Rabiela, Martínez Ruiz, and Murillo Licea (2009, pp. 70–76) studied the construction of individual features of Tetzcotzinco’s water management system. Despite investigators’ interest in pre-Hispanic gardens in Mexico and their significant contribution to this field, there is still a lack of comprehensive and multifaceted research on Tetzcotzinco, primarily focused on the water management system, its construction, and its socio-economic importance.

2.1 Important Considerations in the Bibliographical Sources Used

When filling in the GIS database (WGS 84 UTM 14N) with the geographical data coming from bibliographical sources, there were some challenges encountered that are worth explaining. First, each of the following sources was contrasted and complemented with digital satellite imagery (from Google Satellite) showing georeferenced topographical, hydrological, and urban features together with orthophotomaps with an accuracy of 0.5 and 1.5 m provided publicly by the Mexican National Institute of Statistics and Geography (INEGI, https://www.inegi.org.mx/app/mapas/default.html?t=193&ag=15). Another data that helped building the database came from digitizing historical maps of the area, which were consulted and photographed in situ in the Map Library Manuel Orozco y Berra in Mexico City. In this way, the trajectory of canals, aqueducts, and sources of fresh water, which were proposed decades before when digital visualization and modeling were not possible, was therefore corroborated as it will be shown further.

The first watercourse that was examined and integrated into the database comes from the investigations of Parsons (1971, pp. 145–147). This lies at the eastern side of the site, reaching contemporary Santa Catarina del Monte. During his studies in the Texcoco region, Parsons mapped the remains of pre-Hispanic aqueducts and contemporary canals in the area, combining them into a single system. According to him, the still-functioning rural canals, to a large extent, reflect the course of the old system. With this reconstruction, he proposed a route of canals that could have supplied Tetzcotzinco with water in the past (Parsons, 1971, p. 145). Nowadays, this map is the most widely acknowledged reconstruction of the water management system in the region (García García, 2007, p. 211; Medina, 1997, p. 69). Furthermore, Parsons located a possible water source for Tetzcotzinco in the spring of Texapo, south of the town of Santa Catarina del Monte (Parsons, 1971, p. 146). Based on his analyses, he traced a water route running northwards from the spring, passing through San Pablo Ixayoc and Tequexquinahuac, and eventually reaching Tetzcotzinco by supplying the Caño Quebrado (Figure 2).

Although Parsons’ proposal is widely recognized, it should be noted that there is no historical data that would indicate such a course of the canals in southern Acolhuacan. The spring of Texapo (or rather “Techapan”) itself has been mentioned as a source of water for the local community only since the eighteenth century (González Rodrigo, 2006, p. 28). Also, the survey conducted by Parsons himself did not confirm the increased settlement activity in this area in the Late Aztec Period (Parsons, 1971, p. 95). Therefore, due to the emerging doubts, the canal route proposed by the scholar was revisited and confronted with the natural, topographical conditions of this area provided by the georeferenced data in GIS.

As with Parson, also the investigations of Victor Arribalzaga Tobón (2005) were examined. His research proved the hypothesis concerning the existence of a network of roads and pilgrimage paths, connecting the most important settlements in Acolhuacan with the temples and shrines on the Tlalocatepetl range (Arribalzaga Tobón, 2005, p. 166). He conducted a series of analyses that allowed him to develop a hypothetical road system of the Texcoco region. Remarkably, when his data were considered for this research, a clear correlation between these roads and the routes of canals in the vicinity of Tetzcotzinco emerged visibly. It is important to mention that he did not take into account the possibility that the same paths could also be water channels. However, the coincidences were so clear that became logical to include his work in the following proposed reconstructions of the southern part of the regional water management system.

The proposal of paths of Arribalzaga constitutes a second possible water supply system for Tetzcotzinco, running along the northeastern road as mapped by the author. It begins at the peak of Tlalocatepetl and runs along its northern slope, connecting several minor archaeological sites, and, like in the Parsons’ system, passing through Santa Catarina del Monte and finally reaching the Caño Quebrado and Tetzcotzinco (Figure 2). The trajectory of this route, however, has been slightly altered so that watercourses could feasibly run following the natural shapes of the land. Worth reiterating, this suggested second water network was completed with the assumption that the route of pre-Hispanic roads (in particular, pilgrimage paths) largely overlapped with the course of canals and aqueducts. The analysis of the topographic layer, the contemporary hydrological network, as well as the archaeological findings in Tetzcotzinco and its vicinity and the mentions of canals connecting pre-Hispanic altars and places of worship in the historical accounts (Obregón, 2009, p. 11; Pomar, 1971[1891], p. 3) bring support to this assumption.

The third important source that allowed for new insights into water sources and courses in the Tetzcotzinco area was a colonial historical account known as Titulos de Tetzcotzinco (1999[1898]). This document, written in Nahuatl, describes alleged land rights to Tetzcotzinco and surrounding settlements. This type of manuscripts has been crucial in diverse and controversial land disputes in Mexico since the arrival of the Spaniards. However, Titulos de Tetzcotzinco stands out from other documents of this genre because it deals predominantly with water control and the system of canals connecting individual towns in the Texcoco region. Even though this document states a date of 1537, it is probably a copy; its style indicates more an eighteenth-century creation (A. Brylak and J. Madajczak, personal communication, January 2021). Despite the late origin of the document, it can be assumed that some of the information provided is a legacy of actual knowledge about the region from the early colonial times, and at least some of the canal routes mentioned in the text were still functioning at the time of Titulos’ creation. Therefore, this document is an important source of knowledge about local water distribution systems. Other historical sources that were useful for this study and which helped to complement the information in Titulos were Relación de Tetzcoco (Pomar, 1971[1891]), containing a description of the lands in the Texcoco region and Proceso Inquisitorial del cacique de Tetzcoco (Obregón, 2009) which mentions the pilgrimage system of Acolhuacan.

Although the Titulos text undoubtedly refers to the studied area, its descriptions do not faithfully reflect the topography of the site and therefore, when placing the location of the sites, in the GIS platform, they ended up inaccurate. To overcome this, all the toponyms in the Titulos document were listed and searched on the contemporary region, while looking at the only available data, namely, the topography and hydrology of the terrain and possible courses along which the canals could have been routed. At the same time, other archaeological data were considered, some already mentioned in this work (Blanton, 1975; Brumfiel, 2019; Charlton, 1973; Parsons, 1971; Parsons, Brumfiel, & Hodge, 1996; Sanders & Evans, 2000). Descriptions and maps of these scholars allowed for analysis of the settlement patterns, which in turn served for the reconstruction of this third possible regional canal network in the region.

The most important information provided by Titulos was the reference to the natural water source for the entire system. According to the document, it was placed in Yeloxochitla[n] in the southern Sierra (Titulos de Tetzcotzinco, 1999[1898], p. 1). In this way, the exact location of the hill, today called Cerro Yeloxochitl, was placed; in fact, the hydric-geographical layer confirms an existing water spring (INEGI, 2010). By following the descriptions of the document, the possible course of water canals was suggested, starting at the hill and running northwards through the places and toponyms mentioned in Titulos (including Quetzaltepec, Ixayoc, Xochimanca, and Metecatl), and finally connecting the natural water source with the system in Tetzcotzinco (Figure 2).

These three proposed watercourses, the canal plan proposed by Parsons, the maps done by Arribalzaga, and the route of canals of Titulos, constitute the basis for the formulation of hypotheses about the water control system in the broader regional context (Figures 12–14). These projections were made manually, and the possible courses of canals were designed between the points mentioned in these sources, taking into account several geographical and topographical factors, namely, the elevation of the terrain, the slope of the land, the natural course of rivers and other waterways, dense forest cover, the openness of the land, and location of pre-Hispanic settlements and archaeological sites, that would potentially allow conducting such canals. The range of such features was also limited to the routes mentioned in historical sources (e.g., Titulos de Tetzcotzinco). As it will be shown further, the possibility of distributing water to downstream areas in the nearby valley, taking into account the water management system that was documented at the site, was taken into consideration. However, it is important to mention that there were variables that influence the capability and correct functioning of the proposed models which could not be part of this study. These variables include quantitative and qualitative data on soil permeability, seasonal precipitation, and groundwater status, just to name the most relevant. With these data, it could have been possible to run additional computational models in order to bring validation to the proposed water courses in this study. However, to obtain these data, interdisciplinary collaboration in the fields of geology, soil sciences, hydrogeology, and paleoclimatology is needed, requiring a detailed and financed research plan, perhaps even incorporating the sampling and testing on the field. For now, the proposed maps cannot be considered a definitive feasible outlook of the course of pre-Hispanic canals. Nevertheless, they constitute a first hypothetical approach, partially authenticated with historical, archaeological, and geographical data, aiming to advance in the viability of running water through the water management system at Tetzcotzinco and in its immediate regional area.

3.1 Water Management System Within the Site

Fieldwork allowed the documentation of all known archaeological structures and features at the site with particular attention to the remains of the water management system. The collected data led to metric three-dimensional models of this record (Figure 3), which constitute the basis for further analysis.

Figure 3

3D models of the water management system’s structures with dimensions of the scanned area: (a) Reservoir 1; (b) Reservoir 2; (c) Reservoir 3; (d) Reservoir 4; (e) Structure 5; and (f) Aqueduct 1.

The most impressive and well-known features of Tetzcotzinco’s water management system are monumental water reservoirs. Currently, three of these have been discovered on the hill of Tetzcotzinco (the so-called baños) and the one on Metecatl.

The largest reservoir is Reservoir 1 ^[4] (Figure 3a). It is located on the western slope of Tetzcotzinco, where it is connected with the canals surrounding the main path (Figure 4). It consists of the main reservoir and a terrace structure (García García, 2007, pp. 137–138). The reservoir itself has a circular shape and a floor made of a smooth, natural rock surface. In addition, the reservoir has a smaller, deeper pool carved into the floor. The walls of the reservoir were made of masonry and includes a staircase. Reservoir 1 is also connected with the terraced, three-level structure, probably used for ceremonial purposes. This structure was partitioned by a stone channel, which was formerly connected to the main canal accompanying the path. This channel brought water to the smaller pool that probably served as a filtration device and then filled the reservoir.

Figure 4

Map of Tetzcotzinco showing archaeological structures and features.

The second scanned reservoir is Reservoir 2 ^[5] (Figure 3b). It is located approximately 120 m in a straight line from Reservoir 1 to the southeast, below the main path (Figure 4). The entire structure is intriguing due to the almost exclusive use of monolithic architecture. The reservoir was carved entirely in the bedrock and it is circular in shape. The inner surface of the reservoir was polished and, in the past, it was covered with multi-colored stucco (García García, 2007, p. 141). In the eastern part, the reservoir has carved steps leading to the interior of the structure. The small size of this reservoir and the seat carved inside the reservoir seem to indicate that it played a much more private role than Reservoir 1 and could have been used for ritual or recreational baths reserved exclusively for the ruler.

The most poorly preserved reservoir of this type is Reservoir 3 ^[6] (Figure 3c), which is located on the route of the northern path (Figure 4). It is also a feature entirely carved in the bedrock, protruding a few meters above the level of the path. The circular reservoir itself is the smallest one in Tetzcotzinco. The original and precise depth of the reservoir is hard to determine due to extensive destruction and erosion in this part of the site. There is also a small terrace carved around the reservoir, limited on the south side by a smooth, vertical rock wall. Due to the immense damage, understanding the purpose of Reservoir 3 is currently unattainable.

The last and also the second-largest monumental reservoir is Reservoir 4 ^[7] (Figure 3d), located on the hill of Metecatl (Figure 4). This feature was built inside a larger building which has been interpreted as a place of water flow control (García Garcia, 2007, pp. 215–216). The whole complex consists of two parts. The first, located on the eastern side, is formed by the housing area, composed of a building divided into three small rooms. The second part is a small platform made of masonry with a huge, circular water reservoir. This structure was the first place for water collection at the site and directed the water to the canals within the complex (Structure 5 ^[8]; Figure 3e).

Structure 5 is one of the most complex and least comprehended complexes of Tetzcotzinco’s water management system. It is located directly at the beginning of the main Aqueduct 1 ^[9] (Figure 3f), on the northern slope of Metecatl (Figure 4). This monumental structure covers an area of approximately 1,065 m² and consists of seven terraces of various sizes. All terraces are connected by stairs and continuous channels with a rectangular cross-section that is then connected with the main Aqueduct 1. Three small water reservoirs have been documented within this complex, whose function is still not fully recognized. The first and the smallest reservoir is rectangular, while the two larger ones are in the form of low, circular pools. Moreover, it is possible that formerly there were buildings on some levels of Structure 5 terraces that have not survived to this day (García García, 2007, pp. 217–218).

Along with the documentation of the reservoirs, canals and aqueducts were also scanned. The largest of them is the main Aqueduct 1, which connects the hills of Metecatl and Tetzcotzinco (Figure 4). This structure is spread over a stone and earth mound, with a maximum height of 3 m, and slightly sloping sides in the form of a talud. The total length of the entire structure reaches 170 m, while the width at the base is about 3 m. Excavation studies have shown many phases of the construction of Aqueduct 1, which gradually raised and expanded in pre-Hispanic times. On the top of the mound, an open channel with a rectangular cross-section made with stone slabs and covered with stucco was documented.

Aqueduct 1 ends with a hydraulic “node” located near Structure 1. At this point, the water split into two small canals that circled the central part of the hill of Tetzcotzinco, along the main path, which ends in Reservoir 1. Depending on the section, this canal has a different form: an open canal sunk in the ground, a channel limited by low masonry walls, or a monolithic canal. Due to the form and length of this feature, its construction was documented with photogrammetry, which helped to recreate its course in a digital form.

3.2 WMS Archaeometrical Analyses

The obtained data, especially 3D models, were used for the analysis of the water management system in Tetzcotzinco. The high accuracy of the models up to 2 mm and the large size of the documented area (16,960 m²) allowed to carry out numerous studies in a digital way which not only provided high precision but also provided a faster and more accessible possibility to study architecture than traditional analyses conducted in situ.

One illustrative example that helped the reconstruction of the direction and technique of water distribution within the site was the analysis of height differences. To do this, multi-colored projections of buildings were generated in order to show height differences and surface anomalies both within structures and features. This enabled to observe differences in heights that suggest the direction of the water flow (Figure 5). On this basis, height measurements on individual sections were taken, for instance, the dimensions of canals were in this way obtained. In addition, this allowed the calculation of the water flow velocity at three stages of the system: on the hill of Metecatl, in Aqueduct 1, and in the main canal surrounding the hill of Tetzcotzinco. Determining the velocity of the flowing water can provide a wealth of information about the entire system. First of all, the connection of the water flow velocity with the natural topography of the hill may form the basis for a discussion on whether Tetzcotzinco was intentionally selected as the site for the construction of the water management system due to its natural form. Moreover, the information on approximate flow velocity can reveal parts of the entire system that were most exposed to erosion and damage from flowing water. Furthermore, this leads to finding spots where mechanisms were placed to control and slow down the water. Significant differences in velocity between the individual parts of the system also demonstrate the possible use of technologies for acceleration or deceleration of the flow (Farrington, 1980, pp. 290–292).

Figure 5

Plan of Structure 5 with the gradient of height differences generated from the 3D model and the direction of water flow.

The water velocity in Tetzcotzinco was determined based on the revised Gauckler–Manning formula of flow velocity for the open channels (Chanson, 2004, pp. 264–267):

This formula allows for determining the water velocity based on the values of the Gauckler–Manning coefficient (n), hydraulic radius (R), and hydraulic slope (I). The n coefficient provides information about the roughness and type of canal surface, thereby determining its construction’s influence on the water flow resistance. The hydraulic radius describes the size of the channel by the ratio of its cross-section area to the length of the cross-section perimeter. The hydraulic slope, in turn, determines the canal’s grade by the ratio of the drop-in elevation to the distance over which this drop occurred.

As a result of the measurements and calculations, the estimated velocity of water flow in the canals of Tetzcotzinco was determined. Of course, due to the condition of the water management system at the site and the varying dimensions of each canal (the width and depth of the canal vary between 1 and 5 cm at different sections), the given values are averaged and approximated and the actual speed could slightly differ from given values (Table 1). Due to the non-invasive nature of this project, it was not possible to perform physicochemical analyzes of the canal surfaces that would have allowed determining the type of material used in the construction of particular features and their roughness. These data could have allowed a better identification of the n coefficient for the calculations. In view of this limited in situ observations, this study adopted a coefficient of 0.02 which corresponds to stone, loess, or clay surfaces that appear to be most similar to the material of Tetzcotzinco’s canals (Chanson, 2004, p. 79).

Achieved results show a gradual decrease in the speed of the water within the site from about 3 m/s on the hill of Metecatl to about 0.8 m/s along the main path on Tetzcotzinco. The obtained values seem to correspond to the natural slopes of the terrain at the height of the main path of the site. Hence, this seems to confirm the hypothesis about the intentional choice of Tetzcotzinco as a place of water distribution. Changes in the height of the terrain along which canals had been led enabled control of the flow of water, which was supposed to reach the lowest speed in the central part of the site and slowly feed the most important reservoirs. Beyond the hill of Tetzcotzinco, significant differences are visible, primarily in the place between Aqueduct 1 and the main canal, where the hydraulic “node” likely played a role in slowing down the water. The above results indicate remarkable changes in water velocity in subsequent sections of the system, especially between the structures on Metecatl and those on Tetzcotzinco, which lead to look for other mechanisms that could be used to control the water flow, as will be explained further (Structure 5; Figure 5).

Another important factor in the analyses of the ancient water management systems may be critical flow. It is a concept in hydrological studies that defines how the speed and acceleration of water and the form of a canal can influence the state of water flow and good preservation of the structure (Farrington, 1980, p. 293). Critical flow is defined by the formula^[10]:

The values obtained in this way determine the state of the water flow. The optimal condition is achieved when the critical flow reaches the value of 1. The watercourse is then stable and harmless to the canal. Values less than 1 signify tranquil flow, meaning obstacles on the route or even a slight slope may cause the water to stagnate. Values above 1 indicate rapid flow when larger obstacles or bends can cause the spill of water, and the flow itself can damage the canal (Farrington, 1980, p. 294). The problematic issue with the critical flow is the fact that it is most effective for analyzing canals and watercourses with a relatively uniform shape and slope. Hence, the proposed calculations are only hypothetical and averaged.

Despite the relative inaccuracies, the obtained data (Table 1) may provide some important information. In particular, the calculations show the tranquil flow in Aqueduct 1 and the main canal of Tetzcotzinco. Such a state of water could indicate the need for frequent inspection of these canals and their constant purification due to high sedimentation in low slope structures. Otherwise, the water could stop flowing or even flood the earlier sections of the system.

The opposite situation can be seen on the hill of Metecatl, where the calculations provide us with interesting information about the possible existence of water control mechanisms. At this point, the results of calculations suggest an extremely high flow velocity of up to 3.84 m/s and supercritical flow with a value of 2. The speed and flow were certainly too high for the use of such a water management system, which made it necessary to slow down the water and control it. Then, the features within the Structure 5 became relevant (Figure 5). In this place, there are, at least, three uneven floor levels (sort of terraces), which allowed to reduce the slope of the canals, and three shallow reservoirs (one rectangular and two circular) enabling the drainage of excess water.

Therefore, the accuracy for measuring structures or buildings with water control features also became relevant. To do this, profile contour projections based on 3D models were prepared which allowed the visualization of the contours of individual building features in the form of metric lines. Such imaging, enabled to make measurement analyses on various sectors of the architectural structure. The sections made in this way also permitted calculations of the volume of water reservoirs which revealed the maximum amount of water stored and distributed in Tetzcotzinco (Table 2).

As a result of these calculations, it was estimated that the total capacity of Tetzcotzinco’s reservoirs was 35,488 L at one time. This amount may suggest that these structures did not function as permanent storage of potable water for the larger community. At the same time, it should be remembered that the system in Tetzcotzinco was open and had a constant water flow, which would mean that the reservoirs were constantly filled with water, and their actual long-term efficiency would be much higher. Although the form of the reservoirs (resembling small pools with stairs and benches) also does not support the hypothesis that they serve as water storage structures, they could be used as an emergency water source in special cases, e.g., during a drought (when the water stopped flowing through the canals). However, the exact function of these structures, especially for the consumption of the local population or agriculture, requires further analysis. At the same time, it should be remembered that these reservoirs could also have aesthetic and ritual functions, especially as water mirrors and ceremonial places. Then, the volume capacity of such pools would not matter as much as their form and location.

In addition, cross-section projections of the 3D models were analyzed. Thanks to the high accuracy of the scans, inaccessible places and structural features during the fieldwork, such as small water outlets placed at the bottom of water reservoirs (Figure 6), can be virtually reached. This enables tracking the exact directions of the water drainage.

Figure 6

Cross-section projection of reservoir 4 with a visible outlet below the bottom of the reservoir.

Irregularities and changes in the structure of architectural objects, particularly inclinations and collapses resulting from erosion and tectonic activity, were also observed with the aid of digital tools (Figure 7). These observations allowed noting the places most exposed to natural damage, which may be a key element in the maintenance and reconstruction of the hydraulic complex.

Figure 7

Example of wall collapses in the Reservoir 4 complex: (a) view of the inclination and (b) area that is exposed to damage.

Moreover, the 3D documentation of Tetzcotzinco also allowed to prepare new and precise metric plans for individual buildings, including those used in water distribution (Figure 8a). Based on two-dimensional projections, new architectural plans for the whole of Tetzcotzinco’s architecture were made which contributes to the list of maps made by previous researchers (Figure 8b). This documentation could point to amends for previous measurements but also place to scrutiny the topic of water management on the basis of architectonical analyses.

Figure 8

(a) Plan of the Reservoir 4 complex based on the 3D model and (b) Plan of the Reservoir 4 complex elaborated with traditional methods (García García, 2007, p. 156).

Based on the detailed analyses of 3D models of these structures and the plans of buildings, the exact water flow at the site was able to be traced (Figure 9). The directions of water flow were marked, not only with the differences in heights, especially in the canals, but also based on the construction of the reservoirs. Analyses of 3D models allowed finding water outlets in some reservoirs and tracing their directions. In this way, it was possible to connect the entire water management system and map a possible further course of the water, especially towards the southwestern slopes of Tetzcotzinco, with the direction of the outlet discovered in Reservoir 1. Although a precise analysis of this water flow still requires further studies and archaeological surveys, the observation of a water outlet in this structure seems to suggest a connection between the central water management system and the external canal, described in the next section.

Figure 9

Water flow with outlet directions. Due to the level of destruction and erosion of Reservoir 3, it is currently impossible to determine the exact water flow and location of outlets within this feature.

3.3 The External Canal

During the archaeological survey, there were indications for the discovery of a “new canal” at the lower slopes of Tetzcotzinco (Figures 4 and 10), which had not been mentioned in the literature before. The surveyed area is located approximately 100 m below the main path of the site, meant for pedestrian who visit the site. The area is currently overgrown with dense vegetation, and in many sections, it is very steep, making it difficult to access the area and using some of the documentation techniques. Therefore, the survey here focused on searching for archaeological remains, documenting them with digital photos, and measuring them using handheld GPS.

Figure 10

Construction of the external canal: (a) pre-Hispanic part and (b) cement-like pipe.

The canal spreads between 2,439 and 2,479 m a.s.l. and probably runs along the entire length of the hill of Tetzcotzinco and extends to the east (Figure 4). The archaeological record indicates the remains of this watercourse at a length of 1,430 m. The best-preserved parts of this canal are located in the northwestern part of the site. Contrary to the previously described features from the central part of the site, this canal is not accompanied by a distinct path. So far, it has not been able to locate the end of this canal due to its course through private properties that partially cover the northeastern parts of the archaeological site.

On the southern side, the canal is certainly worse preserved and more damaged. Nevertheless, it is still possible to capture most of its course. It surrounds the hill below the central part (both the central path and Structure 4 below), and it accompanies the formerly cultivated terraces along their entire length. In the eastern section, the canal probably passed towards the hill of Metecatl and was connected with one of the outlets coming from Structure 8 and Structure 9. The exact route of this connection was not observed due to the impossibility of conducting survey caused by the altitude and vegetation. Preliminary analyses and observations suggest, however, that the discovered canal has its origins in Metecatl.

The documented watercourse seems to coincide with the canal mentioned by Parsons (1971, p. 145) and described by him as a functioning structure. Despite the proximity of both watercourses and the similarity of their route, there is evidence that can confirm that they are two different canals. The newly documented object is not currently in use and its condition does not indicate that it could function over the last century. Moreover, the southern part of the canal does not follow the route marked by Parsons. This researcher documented a functioning watercourse, which at the height of Structure 7 changes its direction by 90 degrees and heads south towards the modern settlement of San Dieguito. The new canal does not change its route in the southern part and continues to the east, along the hill of Tetzcotzinco to the hill of Metecatl. It has been observed that the new canal contains various construction techniques. In particular, the southern part was made entirely of masonry (Figure 10a), with the bottom of the canals made of flat stone slabs surrounded by low walls. In some sections, this channel is located on the edge of the terrace, forming its outer border. In the northern part, there are differences in canal construction technology. The main part was also made with masonry; however, it was changed slightly. The canal walls are made of large, flat stones and slabs plunged vertically into the ground instead of the standard walls of small stones joined with mortar. There was also a construction technique within the external canal but used on a smaller scale, where it was carved in natural rock; both types alternate here, which proves the dependence of the use of different techniques on natural conditions.

In the northern part of the canal, there are visible “repairs” or “replenishments” (Figure 10b), which have not been noticed in the other constructions at the site so far. In several places, there are material modifications to form straight pipes or gutters. The analysis of this material shows a great resemblance to cement. The material used for the construction of the pipe was not discovered elsewhere on the site and does not appear to be of pre-Hispanic origin, which may suggest that it is one of the materials used in colonial architecture. The existence of cement-like fragments that probably served to complement the destruction in this channel would confirm the use of Tetzcotzinco in colonial times. Historical sources mention the destruction carried out by the Inquisition on this site and the official prohibitions related to Tetzcotzinco as a “ceremonial place” (Obregón, 2009, p. 24). However, confirmation of the canal repairs carried out during colonial times would suggest that Tetzcotzinco did not lose its importance completely after the conquest. Its practical and technological functions could still be important, including the supply of fresh water, both to arable fields and terraces, as well as drinking water. This canal hidden in plants, whose location and course seem to suggest its connection with human settlements located to the north of Tetzcotzinco, could therefore be the primary water source for humans in colonial times. The entire hydraulic system, destroyed in the central part as a symbol of the “old religion,” could still be used for a long time for practical purposes and was scrupulously repaired by the local community. It is also possible that despite the destruction of most of the site, its water management system was maintained with the consent of the Spanish authorities, which allowed the reusing of the old canals to supply water to local settlements. This discovery seems to change the view of Tetzcotzinco’s complex, previously understood primarily as a religious center, and it will draw attention to its technological and supply role. Tetzcotzinco itself could be understood then as the central hydraulic node and water storage center in the region.

3.4 Water Management System in the Surroundings of Tetzcotzinco

In this study, hypothetical courses of canals and aqueducts in the southern part of Acolhuacan were traced. The search for possible unidentified water sources for the system in Tetzcotzinco is still one of the most problematic issues related to the pre-Hispanic water distribution in the region. This stage of the research focused on the studies of maps previously explained (Figure 2). On the basis of the three proposed routes that delivered water from Tlalocatepetl, the elevation profiles of the terrain were generated. To do this, the Terrain Profile plugin in QGIS software was used, which produces a profile of a given route based on a bit-map with elevation data downloaded through the SRTM Downloader plugin in QGIS. The generated profiles show a graph that is embedded on the x-axis – representing the course of the road in meters, and on the y-axis – representing altitude data above sea level (Figure 11). This process was done separately for all three routes to verify whether the inclination of the land was steep enough on these trajectories, and therefore allow the water to run significantly downward; otherwise, the profiles would show significant doubts for the correct functioning and flow of each of the routes.

Figure 11

Terrain profiles of elaborated routes.

This constituted the first stage of analysis, before moving on to a preliminary hydrological study, including the calculation of the hypothetical water velocity and the state of critical flow, which represents an initial step for the validation of the formulated hypotheses and will allow further viability of such canals’ routes for future multidisciplinary studies. These analyses were conducted for two different types of watercourses: artificial canals (Figure 12) and natural streams and waterways (Figure 13).

Figure 12

Critical flow for canals’ routes.

Figure 13

Critical flow for natural waterways’ routes.

The first model was based on the assumption that each of the proposed routes runs entirely (from the source to Tetzcotzinco) over artificially constructed canals and aqueducts. Despite the lack of archaeological information on the form and dimensions of the canals on various sections of the route, it was assumed that this system could be similar in construction to the canal of Caño Quebrado and Aqueduct 1 in Tetzcotzinco. This assumption implies that the water would run along open channels with a relatively rectangular cross-section, 0.25–0.3 m wide and 0.2–0.25 m high. The hydraulic slope adopted in the calculations corresponded to the region’s natural topography. In this way, it was possible to estimate the speed and critical flow of different sections of each route. This, in turn, made it possible to identify – and therefore propose the lack of it – where such a canal construction would not ensure optimal water flow or the water velocity could destroy the canal.

The second type of watercourse that was analyzed was based on the assumption that the main part of the system would run through natural streams on local mountain slopes. Due to the lack of detailed data on the dimensions and form of individual streams in the region and insufficient knowledge of the hydrogeology of the region, uniform values were adopted for each of the proposed routes. As in the case of the first model, the obtained data allowed to identify and potentially eliminate the most problematic sections of each route. The analysis took into account both environmental problems (especially natural obstacles and terrain elevations) that would prevent effective water transport, as well as hydrological indicators that would suggest too high flow velocity.

The calculations and a comparative analysis of the results obtained for both models allowed combining them and achieving the most effective systems of flow (Figure 14). Therefore, the proposed routes take into account the share of both natural streams (on most parts of the routes, especially in the case of southern mountainous areas) and artificial canals and aqueducts (mainly in the inhabited areas near Tetzcotzinco). The analyses and calculations determined not only the most probable course of the studied routes but also their construction. Research on this issue is the first step in understanding the exact importance of Tetzcotzinco as a water distribution center for the entire region. Tracing potential long-range watercourses will help confirm or rule out the hypothesis about the existence of a regional canal network and understand how and by which routes water reached to Tetzcotzinco.

Figure 14

Critical flow for canals’ and natural waterways’ routes.

The modeling allowed recreating the sections where the water velocity and its critical flow were too high for water to run in a canal typical for the region (similar to Aqueduct 1 or Caño Quebrado). In these sections, it was assumed that the water ran along the natural riverbeds. In the future, this type of research would permit calculations for the time needed to supply Tetzcotzinco as well as volumes of water supplied.