Geographical approach to erosion risk modelling in coastal Aegean Region

Erosion is a natural process which is necessary for soil formation and continues since the World exist. However, human intervention accelerates this process and causes land degradation. Infertile lands occures because of the moved fertile soils so that agricultural production effected negatively. For this reason, erosion has become one of the most important environmental problems. As well as direct studies such as field and laboratory analyzes to determine the risk of erosion, there are also models developed indirectly to perform risk analysis, especially in large areas. The most common models developed indirectly are ICONA, CORINE and LEAM. In these methods, without considering the geographical location of the study area; the effect of natural environment properties on erosion risk is questioned by non-modifiable coefficients. Each model uses different natural environment features. In CORINE method; soil properties, slope, climate, land use factors were investigated for erosion, in ICONA method; lithology, land use, vegetation density and slope characteristics are questioned. In the LEAM method, slope, precipitation and soil erosion sensitivity factors are used without considering the land use which is important for erosion risk. In addition, natural environment characteristics which are important for erosion risk such as slope shape and aspect are not used as parameters in any model. When all of these models are examined, it is seen that the most important natural environment characteristics that may cause erosion risk are not used together so a new erosion risk model is needed, which complies with all these requirements and where variables can be calculated with weighted values. In this study, the natural environment characteristics affecting the erosion risk were determined according to the literature review and worldwide used erosion risk models, the percentage of the impact of each natural environment on the erosion risk was calculated and a new erosion risk model was created. Coastal Aegean Region was selected as the study area. The reason for this area is the presence of large areas where the slope changes at short distances, the presence of areas containing clastic lacustrine units and the places where natural vegetation is destroyed and the irregular flow regime types of rivers. Analytical Hierarchy Process (AHP) was used to determine the effects of natural environment on erosion risk. In order to determine the weights of the natural environment properties that affect the risk of erosion, comparison matrices were formed and weights were determined as a result of the analysis of matrices. In this study, Geographic Information Systems (GIS) and Remote Sensing methods were used in all mapping and analysis. In the new developed model; it was determined that erosion risk is affected by geological / lithological features 34%, vegetation density 24%, land use 12%, slope factor 11%, Fournier Index 9%, Bagnouls-Gaussen Index 4%, aspect factor %4 and slope shape 2% rate. Geographical Information Systems (GIS) and Remote Sensing methods were used for the determination and mapping of erosion risk by using these determined weights. As a result of the developed model, 26% of the study area was found to be at high risk of erosion. It was also found that 55% of the olive groves in the study area and sloping terrain were under the risk of moderate erosion. The erosion risk of the study area was determined also by ICONA method to ensure the accuracy of the developed model and to eliminate the deficiencies in the model. The developed erosion risk model results were compared with the ICONA method results with literatüre review. In addition, it is supported by field observations in the areas where there are certain differences; the accuracy of the developed model has been questioned. In the analysis performed with the developed model, it was determined that the graben bases with accumulation areas were at high risk of erosion. Due to the accumulation in the graben bases, SQL codes (Structured Query Language) in the developed model, these areas were evaluated as accumulation areas. In addition, as a result of comparisons and field observations, the Bagnouls-Gaussen Index parameter was found to be ineffective in determining erosion. In the erosion risk model developed with this study, with the questioning of many natural environment features together; it is concluded that this study will support the emergence of new regional models instead of erosion risk models used worldwide.