Artificial Neural Network estimation of soil erosion and nutrient concentrations in runoff from land application areas

Authors

Minyoung Kim, National Institute of Agricultural Engineering, Rural Development Administration, 249 Seodun-dong, Gwonson-gu, Suwon 441-707, Republic of KoreaFollow
John E. Gilley, University of Nebraska-LincolnFollow

Date of this Version

May 2008

Comments

Published in Computers and Electronics in Agriculture 2008. Published by Elsevier B.V. doi:10.1016/j.compag.2008.05.021

Abstract

The transport of sediment and nutrients from land application areas is an environmental concern. New methods are needed for estimating soil and nutrient concentrations of runoff from cropland areas on which manure is applied. Artificial Neural Networks (ANNs) trained with a backpropagation (BP) algorithm were used to estimate soil erosion, dissolved P (DP) and NH₄–N concentrations of runoff from a land application site near Lincoln, Nebraska, USA. Simulation results from ANN-derived models showed that the amount of soil eroded is positively correlated with rainfall and runoff. In addition, concentrations of DP and NH₄–N in overland flow were related to measurements of runoff, EC and pH. Coefficient of determination values (R²) relating predicted versus measured estimates of soil erosion, DP, and NH₄–N were 0.62, 0.72 and 0.92, respectively. The ANN models derived from measurements of runoff, electrical conductivity (EC) and pH provided reliable estimates of DP and NH₄–N concentrations in runoff.