论文部分内容阅读
在土壤普查中获得的大量的数据,包括野外观测和室内化验,从不同的角度反映了土壤环境状况和养分含量水平。但土壤是一个自然综合体,加之在操作过程中,人为误差也是难免的。这就需要通过一定的方法和手段,来取舍那些失真状态的数据。本文介绍格勒布斯法,讨论可疑数据的取舍问题,以期正确运用有关数据,提高土壤普查工作的质量。一、方法假设我们得到一组反映土壤理化特性方面的样本数据W是服从正态分布: W:N(μ,σ) 设W的一个随机样本数据,按其值的大小列成:
The large amount of data available in soil censuses, including field observations and laboratory tests, reflects the state of the soil environment and nutrient levels from different perspectives. But the soil is a natural complex, combined with the human error in the operation process is also inevitable. This requires the adoption of certain methods and means to choose those distorted data. This article introduces the Glebbus method to discuss the trade-offs between suspicious data so as to make proper use of relevant data and improve the quality of soil census work. First, the method Suppose we get a set of sample data that reflect the physical and chemical properties of the soil W is a random sample of data subject to a normal distribution: W: N (μ, σ) Let W be the value of its size: