Skip to main content
ARS Home » Research » Publications at this Location » Publication #107017
Title: PREDICTION OF PESTICIDE TRANSPORT THROUGH THE VADOSE ZONE USING STOCHASTIC MODELING

Author
item COYNE, KEVIN - UNIVERSITY OF MARYLAND
item SHIRMOHAMMADI, ADEL - UNIVERSITY OF MARYLAND
item MONTAS, HUBERT - UNIVERSITY OF MARYLAND
item GISH, TIMOTHY

Submitted to: Agricultural Engineering International Conference
Publication Type: Proceedings
Publication Acceptance Date: 11/30/1999
Publication Date: 1/3/2000
Citation: N/A

Interpretive Summary: The critical process that is limiting our ability to develop agricultural management practices which conserve water quality is our inability to accurately describe chemical movement through the root zone to underlying groundwater. Although many models have been proposed, they lack the ability to account for the spatially and temporal dynamics associated with water movement through a myriad of pore size distributions encountered at the field scale. With a stochastic approach, a distribution of parameters, instead of a single value, is used to describe this dynamic processes. In this study, field data was used to evaluate the trends in the spatial pesticide concentration distributions associated with each depth. The changes in these distributions were then used to develop a probabilistic approach to predicting pesticide concentrations below the root zone. The results indicate that such a method improved our ability to simulate observed pesticide concentrations below the root zone and thus may be helpful in developing management mode models that could be used to develop environmentally friendly management practices.

Technical Abstract: Prediction of pesticide concentrations through the vadose zone has not been without difficulty. Existing prediction models seem to produce reasonable results for the upper portion of the root zone but fail in accuracy at lower depths. Additionally, measurement techniques are expensive, especially when the sampling of deeper depths is desired. This study has developed a stochastic approach whereby either measured or predicted concentrations of a given pesticide in the root zone area may be used to predict the concentrations of that compound at deeper depths in the soil profile. Data collected at six depths (30, 60, 90, 120, 150, and 180 cm) were used to develop stochastic properties such as log normal probability density function (PDF), mean, and variance for each respective depth. Then, a relationship between mean and variances obtained for each of the six depths was developed using the method of moments. Finally, using the PDF for the 30 cm depth and adjusted mean and variances for the deeper depths, pesticide concentrations were predicted at lower depths for a completely different time period. Results indicate a great promise for accuracy and feasibility of using a stochastic approach in predicting pesticide transport through the vadose zone.