Quantitative assessment of 2014–2015 land-cover changes in Azerbaijan using object-based classification of LANDSAT-8 timeseries

Abstract

The main goals of this study are the objectbased land-cover classification of LANDSAT-8 satellite imagery of 2014 and 2015, the quantitative assessment of gross and net changes of agricultural land, built-up areas, forest, bare soil and forest between 2014 and 2015, the quantification of the Normalized Difference Vegetation Index (NDVI) rates within these land-cover classes, and the change detection analysis between the NDVIs. The achieved overall accuracies of object-based classification for the 2014 and the 2015 land-cover maps were 82 and 87 %, respectively. Therefore, the achieved accuracies were considered to be acceptable for quantified change detection analyses. For the gross areas of agricultural land, forest and built-up areas an increase was observed. The agricultural gross area was 30,911 km 2 in 2014 and 31,999 km 2 in 2015. The gross area of the built-up land increased from 12,550 to 13,548 km 2. The gross area of forest land changed from 8211 to 9175 km 2. A decrease was observed in the gross area of grassland from 28,229 to 24,925 km 2. This was primarily related to the land-cover shifts driven by agricultural activities. The gross areas of Keywords Land-cover Á Change detection Á Gross and net change Á LANDSAT-8 (OLI) Á Segmentation Á Classification & Emil Bayramov

The influence of land use and land cover change on landslide susceptibility: a case study in Zhushan Town, Xuan'en County (Hubei, China)

“…In the 1990s the highest spatial resolution of multispectral images was 30 m (Landsat TM), which allows optimal pixel-based classification. With the development of high-resolution RS images, object-oriented techniques, using a polygon entity as the basic unit, provide a widely used method for information processing (Blaschke, 2010;Bayramov et al, 2016;Ymeti et al, 2017). Therefore, for the present study, both the pixel-based and the object-oriented methods were chosen for the classifications of images obtained in 1992, 2002 and 2013.…”

“…LUCC often implies modifications in both the natural and social systems (Promper et al, 2015;Lopez-Saez et al, 2016), in particular to changes in vegetation cover (Tasser et al, 2003;Schmaltz et al, 2017), undercutting of slopes (Scalenghe and Marsan, 2009), surface sealing or changes of drainage system (Ghestem et al, 2011(Ghestem et al, , 2014, all of which can potentially influence landslide hazard processes -for example, the phenomenon that mountainous areas with forest cover typically appear to be less susceptible to shallow landslides than unforested mountain slopes as described in many studies such as Cruden and Miller (2001), Beguería (2006) and Galve et al (2015). Similarly, deforestation as a result of human activities, e.g., road and/or railway construction, undercutting of slopes and development of settlement areas in steep mountainous areas increases the vulnerability to landslide hazards (Glade, 2003;Bruschi et al, 2013).…”

Land Cover Changes in the Malewa River Basin, Kenya

“…However, findings from a study of the entire Lake Naivasha basin showed that except for cropland and built-up area, areas under forests, woodland, grassland and shrubland suffered a declining trend in [28] also differs with the findings of this study. In Azerbaijan, land cover changes (2014-2015) showed increase in areas under agriculture, built-up area and forest, but a decline in area under grassland [39]. In Hawalbagh block (India), changes observed (1990-2010) indicate an increase in area under vegetation and settlement, and decline in area under agriculture, barren land and water body [40].…”

Responses to reviewer #1

“…In the 1990s the highest spatial resolution of multispectral images was 30 m (Landsat TM), which allows optimal pixel-based classification. With the development of high-resolution RS images, object-oriented techniques, using a polygon entity as the basic unit, provide a widely used method for information processing (Blaschke, 2010;Bayramov et al, 2016;Ymeti et al, 2017). Therefore, for the present study, both the pixel-based and the object-oriented methods were chosen for the classifications of images obtained in 1992, 2002 and 2013. Three sets of RS images were prepared to obtain the LUC maps of different years: Landsat 4-5 TM images from 1992, and SuperView-1 images from 2002 and 2013.…”