The multiple cases of flooding are known which are stimulated by processes of urbanization and wrong agricultural policy. Flooding leads to demolishing of spatial fertile layers of the earth which in its turn causes the erosion of land. Methods of remote sensing allows to obtain the information on humidity of the land cover in wide geographic regions. The soil humidity is measured usually using microwaves radiometers, because there is a sufficiently strong interrelation between soil humidity as far as depth of 5 cm and brightness temperature. The soil humidity is measured also by help of remote sensing meters operating in visible/infrared zone of spectrum, including measurements of NDVI in visible/near infrared band and the land spatial temperature in thermal band. In the article the analysis of uncertainty of remote estimates of soil humidity is carried out by the aim to predict the flooding authentically. It is noted that uncertainty calculation of the soil spatial cover humidity leads to uncertainty of prognosis of possible inundation. The questions on forming of uncertainty in measurements of spatial covering of soil, spatial temperature of soil an the soil humidity. The models for carrying out of series of measurements on relevant schemes are developed for obtaining data on the soil humidity with minimum uncertainty are developed.