Simulating Irrigation Systems (Landscapes) in Ancient and Modern Times

Abstract

The human-water system, exemplified by irrigation systems, can be viewed as a highly intricate adaptive system that emerges from the dynamic and continuous interplay of environmental and societal elements over time and space. A thorough examination of the interconnections between humans, water resources, crops, and hydraulic infrastructure in the practice of irrigation management could yield profound insights into how irrigation systems function. Southern Mesopotamia boasts renowned, advanced irrigation systems that have supported the development of vast urban centers. Nevertheless, the historical workings and evolution of these irrigation systems in Southern Mesopotamia remain shrouded in mystery. It is imperative to explicitly address the interplay between human activities and the water system when investigating the development of irrigation systems and the landscapes they nurture.

In this thesis, I propose that a systematic exploration of the evolution of irrigation systems, progressing from small-scale to large-scale, from short-term to long-term, and from individual to collective, can provide a deeper comprehension of the intertwined environmental and societal dimensions of irrigation systems in Southern Mesopotamia. This exploration offers invaluable insights into understanding the co-evolutionary history of the environment and human society. To this end, I have developed an Agent-Based Model Framework with three model versions from the vantage points of human agents, hydrology, and hydraulics to simulate the irrigation systems in Southern Mesopotamia.

The first model version, the Irrigation-Related Agent-Based Model (IRABM), simulates the key functions of irrigation actions across various scenarios, infusing water realism and human realism into the agent-based model (ABM), thereby representing human-water interactions. This study primarily focuses on water distribution through the manipulation of hydraulic infrastructure and human-made strategies. The IRABM serves as a platform for the integration of human and non-human agents, facilitating actions and interactions among model agents. Furthermore, this theoretically and empirically informed computer model can offer fresh insights into the simulation of human-water systems, elucidating the emergence of irrigation patterns and yields from a dynamic environment.

The second model version, the Advanced Irrigation-Related Agent-Based Model (AI-
RABM), in contrast to the IRABM, incorporates learning behaviors, decision-making processes, and mechanisms at both the individual farmer and irrigation system levels. This model contributes to our understanding of the decision-making processes and mechanisms at both individual and collective levels, particularly concerning water conflicts among farmers in irrigation management. It also guides efforts to enhance communication and cooperation among farmers to optimize irrigation system performance. The model retains flexibility in the parameters, enabling its application to various irrigation systems worldwide.

In comparison to the IRABM and AIRABM, the third version of Irrigation-Related Agent-Based Model (IRABM3) maintains the core components of individual decision-making regarding farmland dynamics and collective decision-making in irrigation management. Through a comprehensive computational approach, including sensitivity analysis and Gini coefficient evaluation, I investigate the emergence of patterns in irrigation systems under diverse scenarios of water availability and the decisions made by heterogeneous agents. This allows for the discussion of the potential processes involved in the development of ancient societies in Southern Mesopotamia. Moreover, the IRABM3 offers adaptability to accommodate spatial and temporal variations within the irrigation system. This adaptability permits the exploration of irrigation-based societies in ancient Southern Mesopotamia on a larger scale, contributing to a broader understanding of the intricate dynamics at play in these societies. Furthermore, IRABM3 forms a foundation for future research by incorporating additional agents into the irrigation system, facilitating a more comprehensive grasp of the evolutionary dynamics of irrigation systems in ancient Southern Mesopotamia and providing researchers with a powerful tool for further investigation.

These three models, presented in this series, demonstrate the potential and reliability of using ABM to simulate the operation of irrigation systems. They enable the interaction, adaptation, and decision-making of agents in response to changing parameters, such as river discharge, gate capacity, various water allocation strategies, and learning behaviors. They make a significant contribution to the study of the development of irrigation systems, both in Southern Mesopotamia and in irrigation systems worldwide.