Abstract
The development of Remote Sensing (RS) methods and data has brought new possibilities in evaluation of the changes in land cover. An evaluation of land cover changes of military training areas based on RS methods and data appears to be highly useful, because of the lack of traditional data on the landscape of these specific territories.
The main objective of this study is to evaluate land cover changes in the former military training area Brdy using freely available Landsat satellite data in the period of 1986-2017. The landscape of the former military training area Brdy underwent significant changes in this period.
Main trends were the increase of area of deciduous forest and the decrease of the share of built-up areas related to the decline of military activities. Land cover classification was performed using the Maximum Likelihood algorithm using Landsat 5 and 8 data.
The accuracy of the classification was verified based on aerial photographs, and always exceeded 90 % in the observed years 1986, 1998, 2011 and 2017. RS archive data have undoubtedly high information potential in the evaluation of landscape changes of the former military training areas.
For purpose of an increase of the accuracy of classification, it is desirable to link RS methods with in-situ measurements.