Response of farmers' decisions and stream water quality to price incentives for nitrogen reduction, carbon abatement, and miscanthus cultivation: predictions based on agent-based modeling coupled with water quality modeling

Ng, Tze Ling

Permalink

https://hdl.handle.net/2142/17069 Copy

Description

Title

Response of farmers' decisions and stream water quality to price incentives for nitrogen reduction, carbon abatement, and miscanthus cultivation: predictions based on agent-based modeling coupled with water quality modeling

Author(s)

Ng, Tze Ling

Issue Date

2010-08-31T20:31:20Z

Director of Research (if dissertation) or Advisor (if thesis)

• Eheart, J. Wayland
• Cai, Ximing

Doctoral Committee Chair(s)

• Eheart, J. Wayland
• Cai, Ximing

Committee Member(s)

• Braden, John B.
• Czapar, George F.
• Scheffran, Jürgen

Department of Study

Civil & Environmental Eng

Discipline

Environ Engr in Civil Engr

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Agent-Based Modeling
• Water Quality Policy
• Nitrate
• Carbon Trading
• Nitrogen Abatement Subsidy
• Bioenergy
• Miscanthus
• Second-Generation Bioenergy Crops
• Land Use Change
• Soil and Water Assessment Tool (SWAT)
• Watershed

Abstract

The present study develops an agent-based model of farmers' decisions under the influence of nitrogen, carbon and second-generation bioenergy crop prices. This study also estimates the effects in turn of these decisions on water quality, namely stream nitrate load. Given that agriculture is the single largest source of nitrogen in surface waters and is a major contributor to hypoxia in coastal ecosystems and eutrophication in streams and lakes, this study is motivated to explore insights for water quality protection in the context of the environment-energy nexus. In this study, the price of nitrogen is based on subsidies paid to farmers for reducing their fertilizer inputs, while the price of carbon is based on carbon trading and the price for second-generation bioenergy crops from a market demand for them. Due to climate change concerns, there is a real possibility for carbon emissions reduction trading to be implemented on a large scale in the near future, which will increase the demand for bioenergy crops, including second-generation ones (defined as high-yielding perennial grasses such as switchgrass and miscanthus). Even without carbon trading, it is likely that there will still be an increased demand for second-generation bioenergy crops due to energy independence concerns. Currently, the primary feedstock for biofuel production in the U.S is corn. However, as the technology to produce cellulosic ethanol improves, it can be expected that perennial grasses, which are high in cellulosic content, will take on larger and larger roles. All this may lead to large-scale changes in agriculture and consequently, stream nitrate load. This study takes a modeling approach to estimate those changes. The conventional approach to modeling water quality policy is to impose upon the system some least cost, or maximum utility, equilibrium. Inherent to this are the assumptions of “rational” behavior, perfect information, zero transaction costs and static conditions. This study explores agent-based modeling (ABM) as an alternative approach, which formulates the system from the perspectives of the individual agents within it. This gives ABM flexibility not found in the least cost equilibrium approach, such that it need not be constrained by the assumptions of the latter. Thus, an objective of this study is to demonstrate the applicability of ABM for water quality policy modeling. This is done using a semi-hypothetical case study of farmers in the Salt Creek watershed in East-Central Illinois under the influence of the nitrogen fertilizer reduction subsidy, and carbon and second-generation bioenergy crop (specifically, miscanthus) prices. An agent-based model of the system is developed and linked to an environmental-response model. The former is based on fundamental economic and mathematical programming principles, while the latter is based on the Soil and Water Assessment Tool (SWAT). The agent-based model is applied to fifty hypothetical farmers. The farmers are heterogeneous in terms of their initial perceptions of prices, costs, yields and the weather, and how they update those perceptions with time. They are heterogeneous in terms of their land areas, fractions of marginal land, economies of scale, yields, time discount rates, foresights, and risk aversions as well. The farmers are also interacting in terms of their knowledge of initially unfamiliar activities. Their uncertainties of the costs and benefits of these activities are reduced as their neighbors or they themselves experiment and gain experience. The farmers' decisions are dependent on their expectations and uncertainties of future conditions, which are updated with new observations according to a Bayesian algorithm. The Bayesian algorithm weights existing beliefs against new observations. The parameters in the Bayesian algorithm are set differently for different farmers such that each is unique in his processing of new information. In this study, two types of behavior are defined: cautious and bold. For cautious farmers, their Bayesian parameters are set such that they are slow to adjust their expectations in response to new observations but quick to reduce their forecast confidence when new observations fail to match expectations. On the other hand, bold farmers are quick to adjust their expectations with new observations but slow to reduce their forecast confidence when there are unexpected changes. Cautious and bold farmers are also dissimilar in their levels of risk aversion; cautious farmers are more risk averse than bold farmers. Results show that the different market instruments are not equal in their effectiveness in inducing large-scale land use changes with the ultimate purpose of reducing nitrate load in surface waters. For the scenarios examined, the most effective means of achieving a significant reduction in nitrate load is to have a market demand for miscanthus, followed by the nitrogen fertilizer reduction subsidy. However, carbon trading is unlikely to lead to any major change in nitrate load. The results are meaningful, which demonstrates the suitability of ABM to modeling water quality policy problems. ABM is also able to provide insights not possible using the least cost equilibrium approach. For example, it is able to predict how differences in the way farmers process new information affect their forecasts of future conditions and hence, decisions. It also appears able to predict patterns of their adoption of new technologies (in this case, conservation tillage and miscanthus cultivation). Its predictions, while empirically untested, appear plausible and consistent with general behavior by farmers. Further contributions of this dissertation include the parameterization of the crop growth component in SWAT for miscanthus. Even though SWAT comes with a database of default parameters for a number of crops, default values for miscanthus are unavailable as it is a relatively new crop of interest. Others may find the parameters for miscanthus useful for their purposes. Another contribution is the development of an economic model of a single farmer using dynamic programming that may be used to support the farmer’s decision-making regarding the possibility of miscanthus as a crop choice and carbon trading as a potential source of income. The model takes into account the sequential and multi-stage nature of the problem, as well as the initial unfamiliarity and learning process of the farmer. This work has also brought to the field of bioenergy development a systems perspective which views an aspect of the problem not in isolation but in the context of other aspects; specifically, in this study, the water quality outcomes of bioenergy crop cultivation is studied in the context of market policies targeted at farmers.

Graduation Semester

2010-08

Permalink

http://hdl.handle.net/2142/17069

Copyright and License Information

Copyright 2010 Tze Ling Ng

Owning Collections