The possibilities of using thermal infrared imaging data for detecting the main parameters of arable soil fertility

: The analysis of the possibility of using the thermal infrared images for detecting soil fertility parameters of gray forest and alluvial arable soils was carried out by the example of a test filed in Tula region of Russia. Together with the sampling of 25 soil probes from the 0–10 cm layer, the open surface of the soil was photographed using a FLIR VUE 512 thermal imager, and the spectral reflectance of the soil was measured. According to the results of the correlation analysis, it was found that the closest correlations for thermal images are observed with the following parameters of soil fertility: the content of humus, nitrogen, exchangeable magnesium and potassium. The correlation coefficient between the humus content and the reflectance in the visible and near IR-regions, as well as with the average value of the reflectance in thermal band exceeds 0.81. In different diapasons of the visible spectrum, the spectral reflectance correlation with the content of exchangeable magnesium and potassium is lower than in the thermal band, where the correlation coefficient with the content of exchangeable magnesium is 0.81, and with the content of exchangeable potassium is 0.65. Power regression equations were constructed for detecting such soil fertility parameters as humus content (R 2 = 0.74), exchangeable potassium content (R 2 = 0.68), and exchangeable magnesium content (R 2 = 0.72) by reflection in the thermal band of the spectrum. The regressions obtained with the thermal imager data and with the spectral reflectance data in the visible and near IR-bands are similar in quality for detecting humus and exchangeable potassium content, while for detecting exchangeable magnesium content they are a bit higher. The obtained results show that thermal infrared images are applicable for detecting the most significant parameters of soil fertility in the test field and can be used as a basis for their real-time remote sensing monitoring.