ВОПРОСЫ УСОВЕРШЕНСТВОВАНИЯ ПОЛУЭМИРИЧЕСКОЙ МОДЕЛИ ПОЧВЫ ДЛЯ ОПРЕДЕЛЕНИЯ ЕЕ ВОДНОГО СОДЕРЖАНИЯ

Abstract

The article is devoted to modeling the process of determining the water content of the soil. Monitoring of soil water content can be conducted using various contact methods, including gravimetric, electromagnetic, thermal, and other methods. The most promising method is considered to be hyperspectral, where the measuring signal is dependent on the color of the soil and its texture, as well as on the content of organic substances in it. In General, remote sensing methods with higher resolution rates both in time and in space allow us to achieve high efficiency in solving this issue. The proposed method of calibration of a known model uses as a starting position a known model relationship between the percentage of water content of the soil and quantity of reflected signal, averaged over all types of soil. Analysis of the well-known method of the inverse semiempirical model of the reflection signal showed that for the accurate realization of this method, used semiempirical model itself must be pre-calibrated. It is shown that an exponential generalized model developed on the basis of data from the XS1 SPOT channel can be used for preliminary calibration of the semiempirical model. Expressions are obtained for performing preliminary calibration of the semiempirical model and thus eliminating the error of measurement of the “miss” type when realizing this method using the inverse semiempirical reflection model. Despite the use of different wavelength ranges, obtaining the same mathematical expressions of the two models under consideration allows to make preliminary calibration of the semiempirical model used to determine the water content of the soil.