New Methodology and Analytical System for Remote Forest Inventory

Abstract

Methodological solutions are proposed for conducting the state forest inventory, based on a statistically based multidimensional classification of forest areas of the Kostroma region according to 32 resource and environmental indicators, with the allocation of six contours of typical forest areas. On the example of a separate forest area, an algorithm for forming a sample set of sample areas with the presence of a quantity of trees in accordance with the «All-Union standards for forest taxation» is shown. To determine the required number of trees on the test areas, it is recommended to use the author's information and reference system of forest taxation standards (IRS FTS, copyright certificate № 2011615418). For the targeted orientation of inventory accounting units that provide representativeness of the sample, it is proposed to use strata, differentiated by types of forest growing conditions, prevailing breeds and age classes. To calculate the required number of sample areas for the individual strata, it is recommended to use the formula of Academician V.S. Nemchinov, which fully meets the requirements of mathematical statistics. On the example of a specific stratum, the entire methodological side of the formation of a representative sample is shown. The reliability of the compliance of the weighted average indicators of the samples in comparison with the indicators of the general sets is confirmed by the Student's t-test at the 95 % confidence level. For automated inventory of forests by means of remote sensing of the Earth and analytical interpretation it is recommended to use aerial photography from the UAV-«Geoscan 101». Photogrammetric processing of the survey materials and creating aerial photos were performed in the software environment-«Agisoft Metashape». As a result of aerial photography, ultra-high-resolution orthophotoplanes (10 cm/pixel) and digital vegetation models (DVM) were obtained. A comparison of the results of interpretation the taxational indicators of stands: stock, average height, average diameter, completeness, and the share of the predominant species in the stand with the actual data obtained on the test areas by means of a continuous enumeration of trees confirmed a fairly high accuracy of the taxation of stands by means of remote sensing of the specified resolution.