Abstract
<jats:p>Using traditional methods for determining forest cover at high mosaic pattern of tree sets in cities may cause significant errors related to vagueness of the cover boundaries. The author evaluates the applicability of the channels R, G and B ranges boundaries dependence, characteristic of the main covering types (trees and shrubs; grassy cover; paved roads and building roofs; muddy- and dirt roads, and arable lands), on the scale of images for their quantitative assessment. Deviation of the decryption program’s results based on this model from those of determining the forest cover using a palette was 6–10 %, depending on the change in the linear scale of the images. Employing the program showed that the average forest cover in the administrative boundaries of Penza is 44,7 %, the specific landscaping is 413,3 m2/person, in the residential area – 25,1 %, 72,1 m2/person, respectively. The older neighbourhoods located on the city residential area’s edge have the highest forest cover, while the younger ones (not more than 30 years old), on the contrary, have the lowest. Non-compliance with the standards for forest cover and specific landscaping in specific areas of districts and microdistricts was revealed</jats:p>