Detection of Burned Areas by Remote Sensing Techniques: İzmir Seferihisar Forest fire case study
Journal Title: Doğal Afetler ve Çevre Dergisi - Year 2019, Vol 5, Issue 2
Abstract
The cause of damage to the environment, forest fires have a significant role in order to give way to natural disasters in the world. Forest fires not only affect ecosystems negatively, but also cause serious problems in economic and social life. As a result of forest fires that occurred between the years of 1937-2016 in Turkey 1,661,506 hectares of forest area was burnt. It is sometimes difficult and impossible to collect data from forests after forest fires. In addition, remote sensing techniques and algorithms are frequently used in determining the burning areas considering the long duration and the cost will be high. Different satellite images (Landsat, MODIS, SPOT, etc.) in various properties are important data sets of remote sensing technology which is frequently used in recent years to map fire intensity, fire damage and burnt zones. In this study, forest fire that occurred and continued 4 days in Seferihisar district of İzmir province on August 9, 2009 was analyzed by remote sensing techniques using Landsat 5 satellite images. Two Landsat images acquired in July 2009 as pre fire and in August 2009 as post-fire. In this study, the capacity of Normalized Burn Ratio (NBR) and Normalized Difference Vegetation Index (NDVI) indices and differenced Normalized Difference Vegetation Index (dNDVI) derived from Landsat 5 images have been analyzed in order to assess the fire severity. Besides NDVI and NBR indices results, maximum likelihood algorithm which is supervised classification method was applied to pre and post fire satellite images. The burnt area after forest fire was calculated as 711 ha with dNDVI, 695 ha with dNBR and 665 ha with maximum likelihood algorithm of pixel based supervised classification method. Based on the three different methods of results are compatible, rational and consistent with the results of the General Directorate of Forestry.
Authors and Affiliations
Aslı Sabuncu
Keywords
Related Articles
- EP ID EP664538
- DOI 10.21324/dacd.511688
- Views 137
- Downloads 0
Developing Flood Extent Map by using Numerical Models and Determination of Areas under Flood Risk: Türkköse Stream Case
Due to rapid and unplanned urbanization, natural disasters such as rising floods affect the daily life. In order to prevent losses and minimize damages, it is necessary to model the streams passing through urban areas an...
Uncertainties in Earthquake Risk Assessment for Disaster Planning
To obtain the sustainability of social life and to increase resistance to disaster that cause loss of life and properties; it is possible to raise awareness in the community about disaster preparedness and engross to tak...
Investigation of Soil-Structure Interaction for Design of Reinforced Concrete Structures under Earthquake Load
One of the most common natural disasters in our country is earthquakes with strong ground motion. Earthquake damages structures with the impact of ground shaking. Prediction of earthquake magnitude in advance is a great...
Thermal Shock Effect on Strength Loss Properties of Rock Materials: An Experimental Study on Thermal Fatigue Durability
Four different types of rock materials from Eastern Black Sea Region of Turkey were investigated to determine their change in porosity as well as the degree of cracking, fracturing, disintegration and strength loss under...
A Water Balance Model Based on Budyko Framework and Its Calibration through Particle Swarm Optimization Algorithm
In the study presented, to define monthly rainfall-runoff relation, a water budget model based on Budyko approach was used. Proposed five-parameter model requires only monthly mean precipitation and potential evapotransp...