State Water Resources Research Institute Program

Project Id: 2010IL202B
Title: An Agent-Based Model of Nitrogen and Carbon Trading at the Watershed Scale
Project Type: Research
Start Date: 3/01/2010
End Date: 2/28/2011
Congressional District: 15th
Focus Categories: Management and Planning, Models, Water Quality
Keywords: carbon trading, nitorgen trading, nutrients, agent-based modeling, environmental policy
Principal Investigators: Eheart, J. Wayland; Cai, ximing (University of Illinois at Urbana-Champaign)
Federal Funds: $ 33,472
Non-Federal Matching Funds: $ 67,460
Abstract: Excessive nutrient loads in surface waters are a major cause of hypoxia in coastal ecosystems and eutrophication in streams and lakes. Agricultural non point sources are the single largest contributor to the problem in the Midwest. One promising approach to controlling agricultural nonpoint source pollution is point/nonpoint source nutrient trading. Under such a policy, farmers may generate nutrient credits through a variety of means, such as reducing their fertilizer inputs, installing filter strips, etc. They may then sell these credits to point sources as direct substitutes for the latter's own reductions.

Another form of trading that affects water quality is carbon trading. Carbon trading gives farmers incentives to adopt certain land management practices such as conservation tillage and cropland conversion to grassland. These practices enhance soil organic carbon, turning soils from a net source of carbon emissions to a net sink; and at the same time reducing soil erosion and surface runoffs, thus enhancing water quality.

Numerous modeling studies have been carried out to estimate the cost-efficiency and environmental effectiveness of discharge permit trading programs (of which nutrient and carbon trading programs are a subset). The common approach used in these studies is to suppose that trading leads to some least cost equilibrium. Inherent to this approach are the assumptions that all trading occurs simultaneously; and that traders have access to perfect information. However, these assumptions are not always valid.

A modeling approach that does not incorporate these assumptions would be more realistic and an improvement over the current least cost approach. It is proposed that agent-based modeling (ABM) is on such approach. ABM is a modeling method that simulates the actions of the individuals, or agents, in a system. ABM possesses a flexibility and level of realism no found in the least cost approach such that it is unconstrained by the aforementioned assumptions of simultaneous trading and perfect information.

To meet these objectives, an agent-based training model will be developed and linked to a watershed-scale hydrologic model to computationally simulate nitrogen and carbon trading under various scenarios. This will focus on the effectiveness of such trading in controlling nutrient runoff from nonpoint agricultural sources. For this study, the Salt Creek watershed in East-Central Illinois will be used as a case study.

The agent-based trading model will incorporate agent learning, where agents' perceptions of its environment, and thus their decisions, may change with time based on past observations. To model agent learning, a Bayesain inference method will be used to update gent's perceptions by weighing new observations against initial prior beliefs.

Simulation results will be analyzed for three elements: water quality, cost and state of equilibrium. It is expected that the results will answer fundamental questions pertaining to the cost efficiency and environmental effectiveness of nitrogen and carbon trading programs; as well as give some insights unobtainable using the leas cost approach. The understanding derived here should be of interest to regulators and stakeholders in formulating effective trading policies to control nutrient runoffs to surface water.

Progress/Completion Report, 2010, PDF

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo logo U.S. Department of the Interior | U.S. Geological Survey
Page Contact Information: Earl Greene
Page Last Modified: Tuesday, 15-Jan-2013 00:50:43 EST