Center for Embedded Network Sensing

In an effort to balance water reuse and protect groundwater quality when irrigating with reclaimed water, an adaptive management scheme with feedback control has been developed and is currently undergoing testing. A wirelessly networked sensor array is being deployed at an agricultural research plot to provide system feedback. Several multi-level sensing arrays (pylons) equipped with soil moisture, temperature, and nitrate sensors are installed throughout the field site. The pylon is coupled to a simulation and management algorithm to optimize irrigation scheduling. Specifically, a nonlinear programming-based control algorithm, referred to as Receding Horizon Feedback Control (RHFC), is proposed to maximize water reuse and maintain nitrate concentration in groundwater below the regulatory threshold. Each pylon supplies the irrigation scheduling algorithm with real-time field information about water infiltration and distribution, nitrate propagation, and heat transport (in support of evaporation estimates). The simulators on which the management algorithm depends are a one- (vertical) dimensional form of the Richards equation coupled to energy and solute transport mass balances. The one-dimensional simulator is used to estimate key local soil hydraulic and transport parameters in near real time. The nitrate concentrations predicted with the updated parameter values are used as input to the management algorithm. The irrigation management problem associated with the RHFC algorithm can be constrained by soil moisture or nitrate concentration levels at specified depths. Additional constraints are associated with plant water needs, and avoiding runoff and spray drift from the experimental plot. Sample results from the adaptive management approach are presented in the context of an on-going field experiment.