Metamodels and nonpoint pollution policy in agriculture

Abstract

Complex mathematical simulation models are generally used for quantitative measurement of the fate of agricultural chemicals in soil. But it is less efficient to use them directly for regional water quality assessments because of the large number of simulations required to cover the entire region and because the entire set of simulation runs must be repeated for each new policy. To make regional water quality impact assessment on a timely basis, a simplified technique called metamodeling is suggested. A metamodel summarizes the inputoutput relationships in a complex simulation model designed to mimic actual processes such as groundwater leaching. Metamodels are constructed and validated to predict groundwater and surface water concentrations of major corn and sorghum herbicides in the Corn Belt and Lake States regions of the United States. The usefulness of metamodeling in the evaluation of agricultural nonpoint pollution policies is illustrated using an integrated environmental economic modeling system. For the baseline scenario, we estimate that 1.2% of the regional soils will lead to groundwater detection of atrazine exceeding 0.12 Mg/L, which compares well with the findings of an Environmental Protection Agency monitoring survey. The results suggest no-till practices could significantly reduce surface water concentration and a water quality policy, such as an atrazine ban, could increase soil erosion despite the conservation compliance provisions. Complex mathematical simulation models are generally used for quantitative measurement of the fate of agricultural chemicals in soil. But it is less efficient to use them directly for regional water quality assessments because of the large number of simulations required to cover the entire region and because the entire set of simulation runs must be repeated for each new policy. To make regional water quality impact assessment on a timely basis, a simplified technique called metamodeling is suggested. A metamodel summarizes the input-output relationships in a complex simulation model designed to mimic actual processes such as groundwater leaching. Metamodels are constructed and validated to predict groundwater and surface water concentrations of major corn and sorghum herbicides in the Corn Belt and Lake States regions of the United States. The usefulness of metamodeling in the evaluation of agricultural nonpoint pollution policies is illustrated using an integrated environmental economic modeling system. For the baseline scenario, we estimate that 1.2% of the regional soils will lead to groundwater detection of atrazine exceeding 0.12 •g/L, which compares well with the findings of an Environmental Protection Agency monitoring survey. The results suggest no-till practices could significantly reduce surface water concentration and a water quality policy, such as an atrazine ban, could increase soil erosion despite the conservation compliance provisions.

A Multi-objective Approach to Integrating Agricultural Economic and Environmental Policies

“…Careful data diagnostics were performed to avoid model misspecification and bias. Prior experience in building metamodels for herbicide leaching strongly suggests the need for data transformation to get a good fit (Bouzaher et al, 1993). A linear model for soil erosion indicated that the error terms are non-random, suggesting heteroskedasticity (non-constant variance).…”

“…A simple and scientifically valid statistical procedure for estimating site-specific attributes and aggregating them to regional levels called metamodelling is used. The metamodel is a simple response function fitted to the biogeophysical outputs from calibrated mathematical simulation models (Bouzaher et al, 1993). Lately, the strategy of combining simulation models with mathematical programming models in order to evaluate alternative resource policy scenarios has become the state-of-the-art technique for integrated assessment (Ellis et al, 1991;Wossink et al, 1992).…”

Atrazine and Water Quality: An Evaluation of Alternative Policy Options

“…Therefore, integration of the system, whereby an economic activity is temporally and spatially linked with physical phenomena, involves: (1) consistently and robustly linking component models; and (2) aggregating the results into regional indicators of risk and benefits. To ensure congruence of temporal and spatial scale, "experiments" with calibrated geophysical process models produced response surfaces (metamodels) that have statistical integrity and known experimental and sampling error (Bouzaher et al, 1993). Metamodeling abstracts away from unneeded detail for regional analysis by approximating outcomes of a process model through statistically validated response functions, which then allows alternative policy evaluations without the need for additional simulations.…”

ESTIMATING ATRAZINE LEACHING IN THE MIDWEST1

“…The use of previous versions of PRZi\1 (Bouzaher et al, 1993) demonstrated the need to evaluate the effectiveness of the model by comparing simulations with specific observations of herbicide contamination in grounchvater or soil water. The results of this study will ultimately be used to develop a rnetamodcl similar to that described by Bouzaher et al (1993).…”

“…The results of this study will ultimately be used to develop a rnetamodcl similar to that described by Bouzaher et al (1993). Metamodels, or regression models, will be developed for specific types of aquifers that have been classified and mapped to define potential vulnerability by hydrologic settings (Burkart and Feher, 1996).…”