Consistent linkage of distributed food, water, energy, environmental (FWEE) models: perspectives of data and modeling platform for integrated FWEE security NEXUS analysis and planning.

In this presentation we discuss methodologies, modeling tools and case studies on linking distributed disciplinary food, water, energy, environmental (FWEE) systems’ models into multi-systems multi-disciplinary integrated models for truly integrated analysis and managing of FWEE security NEXUS. Models’ linkage approaches enable to operationalize the concept of modeling and data platforms for distributed independent models’ “integration” and integrated FWEE security NEXUS management.

Local, national and global FWEE security in the presence of climate change and risks of various kinds depend on the consistent coordination between and within the interdependent FWEE systems regarding sustainable resource supply and utilization. Detailed independent sectoral and regional systems’ models are often used to address these challenges. However, the independent approaches overlook the close linkages and feedbacks between and within the systems and, therefore, possible cross-sectoral implications. Critical cross-sectoral FWEE systemic supply-demand imbalances can trigger a disruption in a FWEE systems network. Disruptions and failures can be induced by human decisions in combination with natural shocks. For example, overuse of water in one system, e.g., agricultural, can lead to drying up of wells, decrease of reservoir water level, shortage of water in other systems, e.g., for colling power plants or hydropower production; an extra load in a power grid triggered by a power plant or a transmission line failure can cause cascading failures with catastrophic systemic outages; a hurricane in combination with inappropriate land use management can result in a catastrophic flood and human and economic losses, similar to the induced by Hurricane Katrina. These are examples of systemic risks motivating the development of proper models’ linkage approaches and integrated systems analysis.

The linkage algorithms are becoming widely demanded in connection with the need for decentralized planning of distributed systems and technologies emerging in agriculture, water, energy, environmental systems, e.g., distributed precision agriculture technologies; hydro-economic models’ linkages; bio-physical crop modeling; distributed energy production.

In this presentation we define and illustrate the two main linkage methodologies:

• linkage of distributed FWEE optimization models (land use, water, energy systems models);
• linkage of simulation and optimization models (crop-yield meta-model from EPIC and a land-use GLOBIOM model).

Both methodologies are based on iterative sequential stochastic quasigradient (SQG) procedures of, in general, non-smooth nondifferentiable stochastic optimization, which converge to socially optimal solution maximizing an implicit nested nondifferentiable social welfare function. The linkage problem can be viewed as a general endogenous reinforced learning problem. The models act as “agents” that communicate with a “central hub” (a regulator) and take decisions in order to maximize the “cumulative reward". The procedure does not require models to exchange full information about their specifications. The distributed models can operate on distant computers of individual agents and “negotiate” with a central computer of a regional planner through the linkage procedure.