Automated image processing methodology is proposed for all-weather satellite monitoring of floods based on C-band radar data, which allows to determine the boundaries and areas of flooded areas when assessing the magnitude, dynamics and consequences of floods. Processing results comparison of medium spatial resolution scanner and radar images from Sentinel-1 and Sentinel-2 satellites is made. The advantages of a radar survey with cloudiness in the monitoring area are shown.

A technique for the operational monitoring of vegetation and water objects in the territory of cities by satellite images of ultrahigh spatial resolution of the visible and IR ranges is proposed. In the process of processing multispectral images from the satellite DubaiSat-2, vector layers of recognized vegetation and water objects were obtained. The analysis of the results of the automated processing of satellite images showed a sufficiently high accuracy of distinguishing the boundaries of recognized objects and a good separation of classes. The developed methodology allows to significantly increase the efficiency and reliability of updating maps of large cities while reducing financial costs. Due to the high degree of automation, the proposed methodology can be implemented in the form of a geoinformation web service, functioning in the interests of a wide range of users.