Page 106 - Çevre Şehir ve İklim Dergisi İngilizce - Özel Sayı

The use of Geographic Information Systems and Remote Sensing Technologies in
Combating Desertification and Erosion

problems caused by climate change and affecting the existence of living
things. The determination of desertification and erosion status using ground-
based measurements is both costly and time consuming at national and
international scale. Therefore, both researchers and decision makers utilise
technological means to obtain accurate and effective results in a short time.
With the developing technology, geographical information systems and
remote sensing technologies make this possible. This review has analysed
the scientific studies conducted in the last ten years, and tried to present the
potential of remote sensing technologies in combating desertification and
erosion. The scientific studies on the subject have been increasing significantly
in recent years. Although Sentinel and Landsat satellite images are widely
utilized in the determination of the change in land cover and the sensitivity
and severity of erosion, different satellite images (such as MODIS, ASTER)
are produced. In addition, temporal, spatial and positional resolutions that
facilitate analyses increase every year. The studies have used many indices
such as NDVI, TGSI, BSI, SAVI, SWI, Albedo to determine plant and soil
condition. For the last few years, the use of high-resolution satellite imagery
and machine learning methods has been focussed because the indices alone
don’t provide effective results. An artificial intelligence-based approach has
been adopted to utilise soil organic carbon and soil texture maps.
An AI-based approach has been applied to generate soil organic carbon
and soil texture maps from a digital soil map. Satellite images, land and climate
data have been considered as input to the model. The XGBoost algorithm has
been used for the optimization of the model.
The most widely-used method for determining the soil erosion condition
from past to present is RUSLE. This method is based on the interaction of
the factors affecting soil erosion such as annual soil loss, rainfall erosivity, soil
erodibility, slope length, slope steepness, cover-management factor, support
practices factor and sediment transport rate.
All these methods and indices offer effective results, but they are
not sufficient at some points. The plant indices may not clearly reflect
desertification and land degradation, for example vegetation may be affected
by rainfall. In such cases, a combination of multiple indices should be made,
including the influencing factors. Field measurements should be focused to
support validation and ensure reliability in the models. The use of artificial
intelligence algorithms for determining the soil surface condition should be
extended. The harmonised vegetation maps and soil maps of Türkiye should
be produced for the values of factors K and C.

93
Special Issue / 2024