The presented article considers the problem of estimating the parameters of root crop biomass based on Earth remote sensing data. The underground commercial part of the biomass of this type of crops is inaccessible to optical remote sensing. The authors develop a classical approach to estimating the parameters of the state of dynamic systems based on mathematical models. In their previous works, this approach was implemented by the authors to assess crops with above-ground commercial biomass. Such crops are cereals and perennial grasses. To assess the biomass of crops with an underground commercial part, the authors proposed using three mathematical models. The first, main one, is the model of the dynamics of the biomass of a root crop, reflecting the relationship between the above-ground part of the biomass and the mass of root crops. The second is a dynamic model of the parameters of the soil environment, reflecting the removal of nutrients and moisture by the biomass of the root crop. The third is a model of optical remote sensing, reflecting the relationship between the reflectance parameters in the red and near infrared optical ranges with the parameters of the above-ground part of the biomass. Since underground biomass is inaccessible to Earth remote sensing, special requirements are imposed on the model of biomass parameter dynamics. This model must have the property of observability, which ensures the assessment of all components of the root crop biomass when probing its above-ground part. The presence of three mathematical models allows simultaneous assessment of the root crop biomass parameters and soil environment parameters with the closure of the assessment algorithm on real Earth remote sensing data. The proposed methodology and algorithms are quite applicable to other root crops, such as carrots, potatoes, etc.