Advantages and Limitations of Small-Scale Physical and Agent-Based Modelling for the Prehistoric Hohokam Canal Systems

by  A. Zoric

The concept of dry land irrigation involves the transport of water from a stream by means of gravity flow through canals. In order to understand the dynamics in an irrigation system, knowledge is required of short and long-term environmental fluctuations, the changing physical system and the consequential human interventions. Perhaps even more interesting are studies of ancient irrigation, where this water management meets archaeology. An example of a prehistoric irrigation culture is that one of the Hohokam peoples, who lived in Arizona for more than one and a half millennium.

Nowadays, a study of the Hohokam and their form of irrigation is not exclusively within the purview of the archaeologist, for it is an extraordinary complex operation, involving various sciences and analytical techniques. The research requires a specific approach to data recovery (stratigraphic analyses and dendrochronology), studies on cultural resource management (societal structure, politics, labour demands, and pottery exchange) and physical validations (hydrological and hydraulic models). 

Archaeological data recovery methods are therefore constantly modified and combined with other sciences to describe the different social and physical levels in ancient irrigation systems. However, there is a lack of models of the functioning of the physical and social components of present and ancient irrigation systems. The reason for this is that only a few projects have approached the ideal of reciprocal interdisciplinary science and the most did not provide a framework for integration of various studies.
This study emphasizes the need for an interdisciplinary approach between various sciences like archaeology, biological anthropology, hydrology, dendroclimatology, geography, geomorphology, botany, epidemiology, hydraulic engineering and many more, for a detailed understanding of ancient irrigation. It is the actual synergy and reciprocity among these disciplines that distinguishes this approach from the paradisciplinary one, for the latter often treats closely linked studies as additional information only. By integrating the interdisciplinary approach on one of the Hohokam’s canal systems and main villages – Snaketown – and by evaluating the relationships between hydrological patterns and irrigational rhythms, new theories might arise where future archaeological researches can find their basis.
Archaeology is the main provider of material evidence for studies on the Hohokam’s socio-political system, while anthropological comparisons with other irrigation cultures are used to strengthen or contradict these archaeological assumptions. The outcomes of these researches are used in this thesis, with the focus on the vicinity of Snaketown’s irrigation system. In this report, summaries of social and areal studies provide input for agent-based models, dealing with the construction, operation and maintenance of this canal system. These past studies are presumed to offer the full spatial extent of Snaketown’s canals, habitation areas and fields in four politically and climatically-related periods of its existence.

Hydraulic modelling of ancient irrigation systems still requires many details on this often complex system and its culture. An incomplete set of excavations offers a limited amount of cross-sectional data for models of irrigation rhythm and sediment accumulation. Moreover, the segmented character of the canal system and its regulatory structures cannot fully be comprehended in case the profiles between separate canal sections are unknown. Therefore new approaches in canal modelling are suggested. Concepts of quantifying water availability are elaborated by examining the cross-sections and associated discharges from previous studies. These are combined with approximations of irrigated land and visualized for two different periods (Pioneer and Colonial) with the aid of physical models. The results show a water distribution which appears to be insufficient for the tail end users. This suggests regulatory systems and rhythms to be prerequisites for well-managed irrigation water supply.
It is furthermore argued that water availability cannot solely be expressed in volumes. Stratigraphic cross-cuttings of the river bed have unveiled a bed level decline of Snaketown’s primary source of subsistence: the Gila River. The first main irrigation canal had a limited capacity, but could deliver enough water to support the lower-lying fields, just beneath Snaketown. If a bed level drop did occur, then the main canal could have been dug deeper to
provide the downstream fields of water. However, the water level might have been too low for irrigation on the upstream lower terraces. The concept of water level decline in canals has been elaborated in this report.

This kind of bed level decline, also referred to as river ‘downcut’, is a phenomenon which mostly occurs during peak floods with longer durations. Not only do floods affect the floodplain irrigators through the secondary effect of water level decline in the river and canals, but they also cause the low-lying fields to overflow during the whole irrigation cycle. In this report, floods of different magnitudes are analysed through new hydrological findings and associated flood models. These are based on translations of historical short-period river discharge rhythms to possible prehistorical events and give an indication of what could have happened during extreme weather conditions. An exceptionally large flood during the irrigation cycle, followed by a downcut and a water level decline would most certainly push the floodplain farmers to move from their primary focal village (Snaketown).

Extreme events like these, but also relatively stable river discharges, could bring sediment layers of different origins to the irrigation canals. How this happens and how it can be modelled, is explained in a qualitative way.

The discussion is a combination of the qualitatively described, agent-based physical water distribution model and the primarily physical flood model. The final comments lead to new insights in the decline and eventual disappearance of this irrigation-dependent culture.
In short, this thesis documents the social system, hydrological variability and physical properties of a prehistoric Hohokam irrigation system, located in the Middle Gila River Valley (AZ, USA). It provides new technical details in understanding the important physical and social struggles of the Hohokam. This is based on small and large-scale hydrological analyses and hydraulic models for flooding and water delivery during the whole agricultural cycle. An exploratory approach is used for the elaboration of important system changes and suggests the direction for future principles in interdisciplinary studies on ancient canal systems.

The content of this document has been edited and published as a scientific paper. Presentations of this work were given at the Landscape Archaeology Conference (September 18th, 2014; KNIR, Rome, Italy), a colloquium (February 26th, 2015; TU Delft, the Netherlands) and the IWHA Conference (24th of June, 2015; TU Delft, the Netherlands).

Student:       Ante Zoric

Committee:    Prof. dr. ir. N. C. van de Giesen (TU Delft), Dr.ir. M.W. Ertsen (TU Delft), Dr. ir. J. Timmermans (TU Delft), Dr. M.K. Woodson