Available nitrogen is the amount of this nutrient available to plants in the soil and the amount of nitrogen provided by fertilizers. Compared to total nitrogen, nitrogen availability is a more useful tool for determining how much fertilizer you need and when to apply it. Determining the level of nitrogen available in field soil is also a useful method to increase the efficiency of fertilizer. Most soil properties are time-consuming and costly to measure, and also change over time. Fast and accurate prediction of soil properties is a necessary to overcome the lack of measured soil property information. Satellite imagery provides contiguous spatial coverage of a field and can be used as a surrogate to measure soil attributes. In the past three decades, considerable progress has been made which prove the capacity and potential of remote sensing in soil science. The spectral characteristics of a number of nutrition content in soil have been studied and huge number of field nutrition mapping were implemented successfully.

Yet there still three major obstacles that prevent the wide application of remote sensing derived soil nutrition status map in precision farming, they are: 1) common remote sensing means cannot detect the entire soil body (“pedon”) that extends from the surface to the parent material, not mention that the thin, upper layer sensed by optical sensors may easily be affected by many factors such as dust, rust, crop residue, plowing and particle size distribution; 2) in most studies that mapping soil nutrition status based on its spectral characteristics, high spectral resolution data are required, yet ever since the failure of EO-1 Hyperion in 2009, there has been a period of more than 4 years that has no satellite-mounted hyper-spectral images at the resolution higher than 30 m. The acquisition of hyper-spectral satellite image cannot be guaranteed. 3) crop coverage in crop growing season make it difficult to obtain soil radiometric property directly, the short period of soil explosion between crop seasons make it difficult to obtain satisfactory satellite images.

To deal with these three obstacles in mapping field soil nutrition status with satellite images, a new method was put forward and tested in mapping available nitrogen content. The basic concept of the method put forward by this research is that soil nutrition deficiency is the primary limitation on crop yield when other conditions are favorable (water, temperature and radiation). Firstly we map the crop yield of three major crops (wheat, soybean and maize) in last 4 years with a light use efficiency (LUE) model –CASA, which can integrate remote sensing indicators and meteorological data to describe crop growth. Secondly, yields of different crops (wheat, soybean and maize) were normalized to make them comparable, a value (normalized yield index, NYI) between 0 and 1 were assigned to each pixel based on its place in the yield range of the crop type. Thirdly, the maximum NYI in the last 4 years was calculated for each pixel to represent the NYI in favorable crop growing conditions. At last the relationship between maximum NYI and observed soil available nitrogen content was identified through regression analysis, and then a map of field soil available nitrogen were produced.

In this study, taking HJ-1 CCD image as major data source and a farm in Northeast China as study area, the method proposed by the author was tested. The technical procedure, application and validation of this method were introduced in detail. After explore the potential of mapping field nutrition status with remote sensing derived crop yield, this paper provides some ideas on how to propel this technology forward to enable its widespread adoption in precision farming.