For rational use of soils it is necessary to possess exact information on soil properties. The traditional methods of monitoring of soils and (or) their separate properties based on local, one-time supervision don’t give an adequate assessment of a current state of a soil cover it should be noted. Transition to spatio-temporal methods with use of modern geoinformation and space technologies is necessary. Remote satellite methods of soil monitoring gain fast distribution, owing to the efficiency, a certain objectivism and relative low cost now, and also because of unique opportunities of one-time coverage by the images received from big height, enough territories, big on the area. For the development of remote monitoring chernozems southern used materials of multispectral scanning multispectral camera ETM + ( «Enhanced Thematic Mapper Plus»), which is on board the satellite «Landsat-7» (data of 2006, 2010, 2012) and OLI («Operational Land Imager»), which is on board the satellite «Landsat-8»(data 2015). The information obtained from them is unified from the point of view of preservation of geometry, calibration, a covering, spectral characteristics, quality of the image and availability of data, despite various carriers of devices ETM+ and OLI. The composite image which has been received from three cloudless satellite images of spring of 2012 (three terms of shooting – 21.04, 30.04, 05.05) has allowed to make the correlation analysis of extent of influence of maintenance of organic matter in a layer of soil of 0–10 cm at a brightness with various spectral channels of the camera ETM+. Such analysis has shown that the closest connection exists between the content of soil organic matter and brightness of the second (green), the third (red) and the fourth (the neighbor infrared) spectral channels. From them three, the greatest value of correlation has dependence between the content of soil organic matter (humus) and brightness of the red spectral channel (r=-0,32). For the purpose of spatio-temporal interpretation of the equation of multiple regressions, 20 agro landscapes in different parts of the Right-bank steppe of Ukraine (The Mykolayiv district and Zhovtnevy district of the Mykolayiv oblast) have been selected. For each agro landscapes was defined content of soil organic matter in the soil using Landsat 7 satellite images taken in 2006 and in 2010 and Landsat images 8 for 2015. The made estimates of maintenance of soil organic matter have shown on processes of fast loss of humus in all layers of soil. Annual losses of soil organic matter in a layer of 0–10 cm from 2006 for 2015 have made 0,16 % in a year, in a layer of 0–50 cm of about 0,13 % in a year, and in a layer of 0–100 cm at 0,10 % in a year. The irrational structure of sown areas and distribution of wind and water erosion processes is the reason of this sad process.
Semantically-Aware Retrieval of Oceanographic Phenomena Annotated on Satellite Images