Estimation of the Nitrogenized Fertilizer Content of the Agricultural Soil Using Hyperspectral Data

The amount of needed fertilizer in agricultural soils depends to the type of crop. Absence of enough fertilizer may cause in tensions in plants and consequently result in decrease in amount of product. According to the agricultural literature, soils containing total nitrogen amounts (N %) between 0.1-0.15 percent are considered as soils containing normal amounts of nitrogen. The values greater than 0.15% indicate rich amounts of nitrogen in soils. Accurate and timely information in this respect can be effective in increasing agriculture land gain. Applying experts and laboratories for measuring the amount of fertilizer can be time consuming and expensive. Remote Sensing technology, performing based on spectral behavior, sometimes produces acceptable data in large scale. To this end, spectral behavior of added material and agricultural soils should be used. In this study some data have been simulated by using a. spectral behavior of embedded nitrogen in soils containing nitrogen such as: Ammonium Nitrate (26% nitrogen), and Ammonium Sulfate (21% nitrogen), and b. spectrum of suitable soils for agriculture such as Sandy Loam soil between 1000 to 2500 nm. Then a regression model was produced and used. Evaluation of model shows the more than 80% correlation between real amount of materials in soil and measured amount which in present condition is valuable. In this paper infinities also were discussed and considered.