Premium
Ammonia Volatilization from Flooded Soil Systems: A Computer Model. III. Validation of the Model
Author(s) -
Jayaweera G. R.,
Paw U. K. T.,
Mikkelsen D. S.
Publication year - 1990
Publication title -
soil science society of america journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.836
H-Index - 168
eISSN - 1435-0661
pISSN - 0361-5995
DOI - 10.2136/sssaj1990.03615995005400050041x
Subject(s) - volatilisation , wind speed , wind tunnel , ammonia , ammonia volatilization from urea , soil science , environmental science , linear regression , paddy field , chemistry , analytical chemistry (journal) , hydrology (agriculture) , meteorology , environmental chemistry , mathematics , statistics , thermodynamics , geotechnical engineering , physics , geology , ecology , organic chemistry , biology
An NH 3 ‐volatilization model predicting NH 3 loss as a function of five input variables was validated using a wind tunnel to simulate rice paddy conditions and direct field experiments. A total of five variables in a central composite statistical design were compared to study the interactive effects of NH 4 ‐N concentration, pH, temperature, wind speed, and water depth. Experiments were also conducted in a flooded rice field with polypropylene basins placed at water level. Samples were collected every hour for determination of NH 4 ‐N concentration. Temperature, pH, and wind speed were recorded continuously, and water depth was constant. Wind‐tunnel data showed that the model predicted observed values with excellent accuracy in the range of conditions found in flooded rice systems. The regression of predicted NH 3 loss on observed losses resulted in an r 2 of 0.98 and a regression slope of 0.99. Field experiments also showed very close agreement between predicted and experimental values with 6‐, 12‐, and 24‐h averages of pH, temperature, and wind speed. The model validation confirmed the theory that NH 3 volatilization is a function of NH 3 (aq) concentration and the volatilization rate constant for NH 3 , which are dependent on five variables: floodwater NH 4 concentration, pH, temperature, water depth, and wind speed. The model is theoretically sound and predicts NH 3 loss with a high level of accuracy using a menu‐driven computer program with easily measurable variables, and can be used in comparison studies of NH 3 loss at the same site.