Premium
Evaluation of LEACHM: II. Simulation of Nitrate Leaching from Nitrogen‐Fertilized and Manured Corn
Author(s) -
Jemison John M.,
Jabro Jalal D.,
Fox Richard H.
Publication year - 1994
Publication title -
agronomy journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.752
H-Index - 131
eISSN - 1435-0645
pISSN - 0002-1962
DOI - 10.2134/agronj1994.00021962008600050019x
Subject(s) - leaching (pedology) , manure , nitrate , environmental science , volatilisation , nitrogen , soil water , denitrification , chemistry , nitrification , fertilizer , agronomy , hydrology (agriculture) , zoology , environmental chemistry , soil science , geotechnical engineering , organic chemistry , biology , engineering
High NO 3 ‐N concentrations in groundwater resulting from agricultural production have increased the need for mathematical models to predict the concentration and mass of NO 3 ‐N leached from agricultural soils. Study objectives included evaluating the N version of LEACHM (LEACHMN) to predict mass of NO 3 ‐N leached from nonmanured and manured corn ( Zea mays L.), and to test two methods of model validation. Four treatments (no manure with 0 and 200 kg N ha −1 and a manure treatment with 0 and 100 kg N ha −1 ) from a NO 3 leaching experiment were modeled for 1988, 1989, and 1990. Model calibration involved adjusting nitrification, denitrification, and volatilization rate constants to minimize differences between predicted and observed data. When calibrated for each year, LEACHMN produced reasonably accurate predictions of NO 3 ‐N mass leached; however, LEACHMN tended to overestimate summer and underestimate spring leaching losses. The lack of a provision in LEACHM to allow solute diffusion out of water flow channels, and underpredicted corn N uptake early in the season probably created conditions conducive for high leaching losses early; following harvest, insufficient NO 3 ‐N in the soil profile resulted in the model underestimating leaching in the spring. Validating LEACHMN using 1988 rate constants for the 1989 and 1990 years was unsuccessful. When 3‐yr average rate constants were used, simulation accuracy improved somewhat; most accurate simulations were found if 3‐yr average values were close to the calibrated rate constant for that year. The model's predictive capability would probably improve if it contained a more complex corn N uptake routine and a dual‐pore water flow component.