Li Hui-lin, Han Yong, Cai Zu-cong
State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing 210008, China.
Huan Jing Ke Xue. 2008 Apr;29(4):1045-52.
The ammonia volatilization on the Typic Gleyi-stagnic Anthrosol with application of common urea and controlled release urea (LP-S100) fertilizers in the rice seasons in paddy soil of Taihui region of China was modeled by Jayaweera-Mikkelsen model. Results showed great difference of ammonia volatilization from two type fertilizers was detected with lysimeter experiment in the rice season. Nitrogen loss via ammonia volatilization after common urea application with conventional ways was 29%-35%, while only 5% of controlled release urea-N was volatilized. The Jayaweera-Mikkelsen model was over estimated the total amount of ammonia volatilization in the whole season, and great deviation from the measured data was obvious for the higher volatilization from common urea fertilizer. The estimated data were 2.95-4.19 times of the measures one for common urea treatments, while they were 1.19-1.40 times of those measured for LP-S100 treatments. The order of magnitude quotient was one of the indicators to evaluate the model estimation. The value of it was 0.8, which indicated the estimation of the model need improvement. Though sensitive analysis for the five parameters in the model was tested and amended the parameter of the concentration of NH4+ -N, a limited term was inducted in the model operation. The amended model got better results as the ratio of estimation to measured data was decreased to 1.12-1.28. The alga activity in the paddy field influenced ammonia volatilization and might make the failure of the model estimation of the original model.
采用Jayaweera-Mikkelsen模型对中国太辉地区稻田土壤典型潜育潴育水稻土在水稻季施用普通尿素和控释尿素（LP-S100）肥料时的氨挥发情况进行了模拟。结果表明，通过蒸渗仪试验发现在水稻季两种肥料的氨挥发存在很大差异。常规方式施用普通尿素后通过氨挥发造成的氮损失为29%-35%，而控释尿素氮只有5%挥发。Jayaweera-Mikkelsen模型高估了整个季节的氨挥发总量，对于普通尿素肥料较高的挥发量，其与实测数据存在明显偏差。普通尿素处理的估算数据是实测数据的2.95-4.19倍，而LP-S100处理的估算数据是实测数据的1.19-1.40倍。数量级商是评估模型估算的指标之一。其值为0.8，表明该模型的估算需要改进。尽管对模型中的五个参数进行了敏感性分析并修正了NH4+-N浓度参数，但在模型运算中引入了一个有限项。修正后的模型取得了更好的结果，估算值与实测数据的比值降至1.12-1.28。稻田中的藻类活性影响氨挥发，可能导致原模型的估算失败。
