Page 99 - Çevre Şehir ve İklim Dergisi İngilizce - Özel Sayı

Mustafa Sert - Abdullah Emin Akay
Ayhan Ateşoğlu

Albedo = (−1,305 x NDVI) + 65,93 (6)
The CVA technique, ‘change-vector analysis’, was used to assess changes
in the direction and intensity of land surface change and to analyse all
these vegetation indices pixel by pixel. The results of the study showed
that the relationship between NDVI and BSI facilitated the characterisation
of desertification and that the pixel-by-pixel detection of the direction and
severity of change was effective. Moreover, the fact that remote sensing
enables both spatial and temporal analyses demonstrates that it is an efficient
and cost-effective tool.
In their study ‘Assessment of land cover change and desertification using
remote sensing technology in a local region of Mongolia’, Lamchin et al. (2016)
revealed the change in land cover and assessed the desertification situation
at the scale of Mongolia using Landsat satellite images. For this purpose,
vegetation biomass, land pattern and the land surface indicators- NDVI and
TGSI indices- derived from micro-meteorology and land surface albedo were
taken into account, and the Decision Tree (DT) approach was used to assess
land cover change and desertification.
• Topsoil Grain Size Index (TGSI): TGSI was developed by Xiao et al.
(2006) based on the fact that the NDVI method is insufficient to interpret
the actual degree of desertification since even a single rainfall can
affect vegetation change. TGSI is a remote sensing index used for the
detection of the topsoil layer by analysing the grain size composition
at the soil surface, which allows monitoring desertification, particularly
in arid regions. It is calculated with RGB (red, green and blue) bands of
satellite images.
NDVI = (Red − Blue) / (Red + Blue + Green) (7)
Decision Tree (DT): It is a diagram commonly used in machine learning to
forecast a future state using past data. In the study ‘Assessment of land cover
change and desertification using remote sensing technology in a local region
of Mongolia’, a combination of DT classification techniques was used and
maps were produced according to the frequency distribution of desertification
severity for each vegetation cover in a certain period with the aim of evaluating
the data obtained using NDVI, TGSI and land surface albedo variables. The
results of the study suggest that the combination of TGSI and albedo gives
better results than the combination of NDVI and albedo.
In his study ‘Estimate to Soil Losses in Catchment of Çankırı-Ekinne Dam
Using GIS/RUSLE Technology’, Özcan (2016) used the Universal Soil Loss
Equation (USLE) and Revised Universal Soil Loss Equation (RUSLE) empirical
models, which are widely used in the international literature and in Türkiye
to calculate the amount of soil lost due to water erosion. These models can

86 Journal of Environment, Urban and Climate