| Near infrared spectroscopy is a technique that shows many possibilities in the field of testing chemical and physical properties of materials. Soil and agricultural products are not excluded of this practical application of NIR. The aim of this research work was to study the potential of NIR spectroscopy technology to predict N, P, K, OM and pH in a loamy mixed soil. To achieve this purpose, 165 samples were analyzed, 30 of them at specific locations within the field, while the remaining 135 where randomly collected. The experiment was conducted at the Zhejiang University in the college of Biosystems. and Agricultural Engineering. The experimental site is located in Hangzhou County, China, at longitude 120°11' E, latitude 30°27' N, and altitude 25 m above sea level. The field covers an area of 0.36 ha. A soil profile to a depth of 50 cm at five plots, four of them located in each corner of the experimental area, and a central plot according to the diagonal northeast-southwest and northwest-southeast intersection, was excavated. According to each profile was analyzed and described the surface horizons in order to find out any information visually and soil type was classified as loamy mixed active thermic aeric endoqualfs. The test field boundary was performed with Trimble AgGPS 132 equipment and the normal grid method was used to sample the soil. The collected soil in each sample was divided in two portions, and was classified in A or B. Group A to be analyzed for pH, water content, total N, available P, available K, and OM using traditional chemical procedures and group B was taken to Remote sensing laboratory to be examined with NIR.The soil spectral data collection was performed with a spectrophotometer (FieldSpec Pro FR (350-2500 nm)/ A110070, Trademarks of Analytical Spectral Devices, Inc. (ASD). During these measurements, the soil was set in a petri dish and then was carefully flushed the surface. Reflectance measurements of monochromatic light were complete from 350-2500 nm, wavelength increment 2 nm. Three reflectance spectra were taken over the central area of a petri dish, rotating the sample in between each scan approximately 120 °. For each reflectance spectra the scan number were 20 at exactly the same position, a total scan for each sample were 60. All spectra recoded were checked visually and averaged using ViewSpec pro version 2.14 (from ASD) and exported to multivariate analysis software, The Unscrambler 8.0, (CAMO ASA, Norway). During the statistical analysis, over the spectral reflectance data, in units of log (1/R) different mathematical training and procedures were applied. Multivariate calibration techniques were used to relate the soil absorbance spectra (400-2400 nm) with the measured properties values.Focused into analyze the influence of moisture content, particle size, light source incidence angle and captator height on a loamy mixed soil spectra, as well as, in to propose the height, size and location angle for the capturer device with the detection fibers, according with the criteria of high absorbance resolution, several experiments were performed. To achieve the previous purposes five samples spectrum were recoded at three different height 100, 150 and 200 mm from the soil surface, while other five samples spectrum randomly selected were recoded at three different light source angles 20, 45 and 60 °. Later on, all the spectra recorded were averaged and visually checked, using ViewSpec pro 2.14. No representative differences were found among the spectra when the captator height above the sample surface was changed. In all cases, the spectra recorded fallow the same geometrical behavior, starting with high absorbance and decreasing with the increase of wavelength. A slight smaller absorbance with the smaller effect of scattering was observed when height 50 mm was used, while the absorbance spectra were almost the same when heights were 100, 150 and 200 mm. 100 mm most of the time showed the highest absorbance and after apply a splice correction was found to be the most representative. Ac...
|
PDF Full Text Request