DISPLAYING:

Events happening on (October 23, 2026)

• 23
  Oct

  PHD-EE PROPOSAL DEFENCE "DATA-DRIVEN IDENTIFICATION OF SOCIO-ECOLOGICAL DYNAMICS FOR SUSTAINABLE RESOURCE MANAGEMENT"

  Date: October 23, 2026
  Time: 09:00 AM
  Location: EE Reading Room

  Speaker: Hassaan Ahmed
  Supervisor: Dr. Abubakr Muhammad
  

  Irrigation water use emerges from many individual decisions about when to irrigate, how much to apply, and whether to use surface water or groundwater. These decisions may be shaped not only by water availability, weather and institutions, but also by other farmers through imitation, shared information and local norms. Such influence networks could change how water stress, technologies and policies spread, yet they are rarely measured, and separating influence from a shared environment (the reflection problem) remains a central obstacle. This proposal develops a general framework for identifying farmer influence networks from water-use dynamics. Water use is modelled as an opinion-dynamics system coupled to the state of the water resource. The framework is tested in two contrasting settings. The first is monthly irrigation in 109 watersheds of the U.S. High Plains Aquifer, coupled to groundwater levels from a space-time Gaussian process. The second is daily irrigation on eight small farms in a canal roster (warabandi) system. DMDc and Hankel DMDc forecasting models are combined with constrained network estimators and identification tests. In both settings the estimators recover known networks in simulation, and the tests serve as diagnostics of data adequacy, showing which water-use data can reveal farmer influence. The proposed research will establish the data requirements and identification conditions under which the social network can be learned reliably, providing a foundation for network-aware, sustainable management of shared water resources. To this end, it will apply the framework to behavioural water-use records, address the reflection problem through network-based instruments and exogenous shocks, and couple the learned networks to resource models to evaluate how targeted interventions can steer water use towards sustainable outcomes.

  • All Events
  • Public Events
  • Student Events
  • Academic Activity
  • Thesis Defence
  • Departments
  • Electrical Engineering

  Read More