3414"肥料试验模型拟合的探讨

Study on simulation of "3414" fertilizer experiments

以UNDP95-98年在河北唐山进行的27个"3414"冬小麦试验为例,研究了三元二次施肥模型与一元施肥模型对"3414"肥料试验结果的拟合情况。结果表明,三元二次施肥模型进行拟合试验成功率仅为56%,而采用一元施肥模型可以对全部试验进行数据处理。在同一试验中,与一元施肥模型相比,采用三元二次施肥模型计算的推荐施肥量偏高;根据一元施肥模型拟合计算分别有15%、44%、67%的点不需要施用N、P、K肥,而采用三元二次施肥模型则无法获得这一信息。因此,在"3414"试验中,一元施肥模型是三元二次施肥模型拟合的一种很好补充和优化手段。

This paper studied the effects of different fertilizer models on simulation for the'3414' fertilizer experiments, which were conducted from 1995 to 1998 in Tangshan City in Hebei Province, sponsored by UNDP (United Nations Development Program) The results showed that,the success ratio of simulation by threefactor and onefactor model was 56% and 100%, respectively; the recommended fertilizer amount by threefactor model simulation was higher than that by onefactor model simulation in the same '3414' experiment; and the ratio of the experiments without N, P2O5 and K2O recommended by onefactor model simulation was 15%, 44% and 67% , respectively But by the threefactor model simulation, such information was not available So, in '3414' fertilizer experiment, the onefactor model was a supplement and an optimization for the threefactor model 