Assessing the Effect of Land Use Changes on Carbon Storage Using Forest Management Plan and Remote Sensing Data

OpenAlex İngilizce

This study evaluated the land use land cover (LULC) classes that were obtained from the forest management plans (FMP) of the Beykoz planning unit and generated from Landsat images for 1992 and 2012. FMPs increased in coniferous forest areas (18.64%) and broadleaved forest areas (1.03%), whereas a decrease (12.79%) was observed in other areas within 20 years. Eight different vegetation indices (VI) obtained from satellite imagery were used for land use classification. The results of the study showed that the RVI had the highest overall accuracy (71%-72%) and kappa value (0.56–0.59) in both study periods. While stored carbon estimated from the FMPs decreased by 46.2% within 20 years due to regeneration of mature forests with clear-cutting, stored carbon assessed from the NDVI map increased by 1.3%. A positive correlation was found for the stored carbon estimated from the FMP and NDVI maps at a rate of 95.2% (R2=0.90) in 1992 and 54.4% (R2=0.29) in 2012. Although there was a high correlation between the carbon storage data obtained using the two methods, there was also a significant difference (p < .05). In addition, the effect size was determined to be very large in 1992, and large in 2012.