母质与土地利用类型对土壤光谱反演模型的影响

用可见光近红外(Vis-Near Infrared,VNIR)光谱建立的土壤反演模型可以快速高效测定土壤某些属性,但不考虑土壤自身特点的纯统计模型反演的精度会受到制约。本文研究了母质和土地利用类型对土壤光谱反演模型建立的影响。研究所用集合为SF(安徽宣城的林地样品)、SP(安徽宣城的水田样品)、DP(安徽定远的水田样品),结果显示:母质和土地利用类型的差异会显著影响异地模型的适应性,一个地区建立的反演模型不可随便用于母质和土地利用类型不同的其他地区;当异地模型不适用于反演时,可考虑采用精度稍低的全局模型进行预测。

Effects of Parent Materials and Land Use Types on Inversion Models by Using Soil Spectral Data

Using visible near-infrared (Vis-Near Infrared, VNIR) spectral data to build inversion model is a rapid and nondestructive potential method to measure soil properties. However, the models based on pure statistical methods are weak in explaining the information of soil properties. This paper studied the effects and strategies of parent materials and land use types on the established spectral inversion model. 3 sample sets of SF (forestry soils from Xuanchens), SP (paddy soils from Xuancheng) and DP (paddy soils from Dingyuan) were used, where, SF and SP with same parent material, SP and DP with same land use type. The results showed that the differences of parent materials and land use typse could affect the suitability of the off-site model significantly. It is possibly unreliable to use a model in a different region with different parent materials or land use types, however, the global model with expense of the prediction accuracy can be used when the off-site model does not work well.