Access to and control of surface and ground water were critical for Angkor’s success as a low-density city, particularly in terms of flood control in the urban core and agricultural productivity in the vast low-density sprawl that surrounds it. Yet, the city’s water distribution network was not constructed with a single integrative function (as a planned, unified single system for water management and distribution) but was a palimpsest created over several hundred years. As such, parts of the system may have ceased to function, have become redundant, or have been repurposed while the city was active [the early centers of Mahendraparvata and Hariharalaya are obvious examples of such; (48, 49)]. The systemic complexity and interconnectedness of the network emerged from its iterative growth, as well as the overprinting, modification, and redundancy of existing infrastructure. While this is widely understood, the complete network is often treated as a single, functioning, engineered system with a set of shared purposes [see, in particular, (32), but also 31)].

That the network is a diachronic artifact means that the functional part of the network at any given point in time may have been much smaller than our model indicates. From a network theory point of view, this means that the network is not topologically static either during its growth or in its operation. Equally, the remains of the network that can be observed (and mapped) in the modern environment represent only the endpoint of the network’s growth, not its evolution. However, irrespective of the intentions of the people who built and maintained the network, or the use of its parts over time, the components of the network are still capable of influencing the flow of surface water because they remain physical topographical features in the landscape. The age or original purpose of an individual component (edge) within the network is then irrelevant if its existence continues to influence water flow. In most cases, individual components can be linked mechanistically to other components in terms of flow connectivity, irrespective of the age of the components or the specific intentions that may have motivated their construction. These simple principles—that extant topographic features (canals, embankments, dikes, reservoirs, moats, and so on) influence the flow of surface water and that captured water flows downslope—form the basis of our approach to coding Angkor’s diachronic water management network.

Here, we took the network as it was represented in the archeological maps as an accurate representation of the water distribution network, as it was (conservatively) well after the end of its growth, in the middle of the 14th century. We were specifically interested in this “time slice” because it coincided with the “pluvial” episode associated with significant disruption to and reorganization of the water distribution network (37). At that time, the Siem Reap canal incised up to 8 m below the land surface, in its middle reaches. This is known to have occurred in the 14th century [no earlier than 1268–1423 2σ calibrated years CE using the IntCal13 database (50), with a median probability of 1345 CE; 1286–1440 2σ calibrated years CE with the southern hemisphere SHCal13 dataset following (51), with an offset of 21 ± 6 years following (52) and (53), with a median probability of 1356 CE]. That event then represents a fundamental reorganization of the water distribution network at the end of the “Angkor 1” drought (37) and represents the final fragmentation and collapse of the network. The frequency of flood events that cause damage to the network must be lower than the rate of response in the receiving system (that is, adjustment of the network topology through erosion or sedimentation) for the model to be applicable. The superdecadal frequency of seasonal rainfall events likely to lead to significant flooding [as reflected by the high Palmer Drought Severity Index values in tropical tree ring indices (37)] indicates that this minimal requirement can be met. Our simulated network incorporates all surface archeological features that can be associated spatially and functionally with the water distribution network and is therefore the best approximation of the network as it was in the middle of the 14th century.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.