The penetration of renewable resources, such as wind and solar energies, is increasing all over the world. Against the conventional thermal power systems, the intermittency and volatility of renewable energies are formidable challenges to the future power system operation. Therefore, future power systems need alternative forms of flexibility to hedge against intermittent power. Demand-side flexibility is a practical solution that attracted much attention in recent years. There is structural flexibility in electricity consumption, including residential, commercial, agricultural and industrial sectors, which can be integrated into future power systems. Against the literature on residential, commercial, and industrial sectors, agricultural flexibility is still a challenge in power systems. This paper aims to narrow this gap by proposing a responsive structure for agricultural irrigation systems. Achieve this aim, first of all, a mathematical model is proposed for responsive irrigation systems considering groundwater, surface and booster water pumps. After that, a multi-stage stochastic approach is addressed to schedule a time-oriented agricultural demand response program from 24 h ahead to near real-time. Then, a multi-layer structure is suggested to integrate the farm flexibility from the demand-side, i.e. responsive farms, into the supply-side, i.e. the dynamic electricity market, through a devised agricultural demand response aggregator. Finally, the approach is implemented in the Danish sector of the Nordic Electricity Market to show the applicability of the proposed framework. The results show that the proposed responsive irrigation system can provide great flexibility to power systems in comparison with traditional irrigation systems.