统计分区施肥系统模型的建立及应用

通过对一定区域内耕地产量和土壤养分含量进行统计分析,建立统计分区模型(Statistic Division Fertilization Model,简称SDFM),并依据模型参数划分高、中、低产量区间和土壤养分含量区间。在区域内进行多点实验,获取它们的区域适宜施肥量。然后在区域适宜施肥量的基础上,利用决策点实际产量和土壤养分值进行调整,得到最终施肥推荐结果,从而对施肥量进行决策。以农安县2006~2011年产量和养分采样点数据为原始数据,利用目标产量养分平衡法和统计分区施肥模型进行了决策对比。结果显示:统计分区模型决策结果出现异常值的概率只有2%,而目标产量法对氮肥、磷肥和钾肥的决策结果异常值出现概率分别为45%,47%和47%。可见统计分区施肥很大程度地提高了决策结果的精度,可在县级以下的测土配方施肥中进行推广。

Establishment and Application of Statistics Division Fertilization Model

In the paper,we established a fertilization method-statistic division fertilization model (SDFM) based on the statistical analysis of crop production and soil nutrient content in a certain area.Yield and soil nutrient content were divided into three levels (low,medium and high) based on the statistical parameters.We made the decisions for fertilization model after recommending fertilization amount according the actual yield and soil nutrient contents.We compared the nutrient balance method and statistic division fertilization model based on the data of crop production and nutrient contents from 2006 ～ 2011 in Nong'an County.The decision result acquired from the model of statistic division had less than 2％ of outlier only,while the probabilities of outlier were 45％,56％ and 47％,respectively,in the decision result of N,P and K fertilizers by using the method of target yield.Fertilization method according to statistic division greatly increased the accuracy of decision result and can be popularized in testing soil for formulated fertilization.