Journal of Rangeland

Using accurate and low-cost methods to monitor quantitative changes in vegetation cover play an important role in the efficient and sustainable management when the area is too large. In this study, the changes of vegetation cover were investigated using satellite data of TM (Landsat_5) and oli (Landsat_8) sensors for a 30-year period (1988-2018) in Chaharbagh rangeland of Golestan province. To detect the vegetation changes, the numerical values ​​of NDVI index were classified in to 4 different cover categories, good, medium, poor and very poor for the study period 1988 - 2018, and the changes were determined using the CROSSTAB technique. Accuracy was assessed by comparing satellite imagery and ground information using the kappa coefficient. The results well showed the changes, decreasing, increasing and unchanged vegetation classes so that the overall accuracy and kappa coefficient of the maps were 90% and 0.86, respectively for 2018. For the course of the study land areas with good vegetation were increased from 2025.5 ha to 2323.7 ha. The highest reduction of very poor cover was 22/7 hectares and the poor cover classes has decreased from 376.1 to 103.18 ha. The good cover class has decreased from 530.1 to 527.3 ha in 2018.  Evaluations showed that the application of remote sensing has the ability to detect minor changes in vegetation classes under environmental and managerial factors in semi-arid regions over time.