Discussion on remote sensing estimation of soil nutrient contents

Remote sensing is the effective guarantee for quickly and accurately acquiring soil data needed by the precise agriculture. Soil nutrient´s measure and calculation method has already developed to spectrometry compared to the traditional geochemical method. This article systematically described the research progress and analyzed the key technologies in using remote sensing techniques for soil organic matter, N, P and K nutrient contents estimation. In the soil studies the spectrum data applied includes ground measured spectrum, multispectral remote sensing and hyperspectral remote sensing data. In the sensitive bands selection, the pretreatment mainly is the differential or derivative treatment on spectrum including more than ten kinds of transformation. There are linear and nonlinear models used in the estimation model establishment. In the model verification, four indicators are often used. This article also put forward the existing problems and future trends of the research domain.