In-situ spectroscopic characterization of soil properties for precision agriculture**

SoilReader®是一家精密农业设备制造公司，正在革新土壤分析技术，利用土壤光谱学原理，对深度的多种土壤成分进行高精度和原位测量。SoilReader是一种附着在圆盘上的土壤光谱仪，它通过对多种土壤成分的高分辨率分析（约3500个样本/英亩），记录多个深度（顶部9英寸内）的实时土壤光谱响应，为作物投入的精确管理提供快速和非破坏性土壤表征解决方案。然而，在SoilReader圆盘的每次旋转期间，光谱仪在土壤内的黑暗环境和土壤外的明亮环境中连续记录光谱响应。这给土壤光谱的处理和土壤性质的预测带来了困难。圭尔夫大学和SoilReader之间的合资企业旨在开发分离光谱和改善土壤特性预测的策略。该项目旨在回答的更具体的问题是：1）光谱测量如何在现场（原位）条件下预测多种土壤性质？2）如何建立一个多元统计模型来处理地上/地下的光谱响应，以提高土壤性质的预测能力？通过该项目产生的研究、技术和信息将为种植者、服务提供者、研究人员和其他数据用户提供一种在短时间内进行土壤表征的廉价手段，以提供精准农业服务。此外，该项目开发的数学模型将指导SoilReader开发和加强分析服务，并将在未来改进设备，以通过优化使用农艺投入来支持更好的作物生长和发育，同时提高环境的可持续性。

SoilReader®, a precision agriculture equipment manufacturing company, is revolutionizing soil analysis techniques with highly precise and in-situ measurements of multiple soil constituents at depths in real-time utilizing the principle of soil spectroscopy. SoilReader, a soil spectrometer attached to a disc, records real-time soil spectral response at multiple depths (within top 9-in) as it moves through the field with high-resolution analysis (~3500 samples/acre) of multiple (9) soil constituents providing solutions for fast and non-destructive soil characterization for precision management of crop inputs. However, during each rotation of the SoilReader disc, the spectrometer records spectral response continuously both in dark environment within the soil and the bright environment outside of soil. This creates difficulty in processing the soil spectra and the prediction of soil properties. The joint venture between the University of Guelph and SoilReader aims to develop strategies to separate the spectra and improve prediction of soil properties. More specific questions that this project aims to answer are 1) how well spectroscopic measurements can predict multiple soil properties in the field (in-situ) conditions? and 2) how can we develop a multivariate statistical model to process above/below ground spectral responses for improved prediction of soil properties? The research, the technique, and the information generated through this project will provide an inexpensive means for soil characterization in short time for the growers, service providers, researchers, and other data users for precision agriculture services. Additionally, the mathematical model developed from this project will guide SoilReader to develop and strengthen the analytical services and will improve the equipment in future to support better crop growth and development from the optimized use of agronomic input while improving environmental sustainability.**