Uzaktan algılama verilerinin yersel ölçümlerle entegrasyonu ile toprak tuzluluk haritalaması; Aşağı Seyhan Ovası, Adana, Türkiye

To properly respond to the global changes, which challenge the scientific community to make available the data for a decade or continuous data in order to monitor the changes of atmosphere, ocean, and land, the remote sensing science can be helpful. Only remote sensing from space, can provide the global, repeatable, continuous observations of processes, needed to understand the Earth system as a whole. Remote sensing data can be used in several applications such as, meteorological data collection, change detection and land cover mapping, disaster monitoring and so on. One of the important applications of remote sensing is to detect and monitor the soil salinity level in agricultural areas. Saline soils are present in many areas of the world. Moderate to severe salinity, which is more or less visible in the landscape, reduces the annual yields of most crops. In the lower Seyhan plate of Adana district in Turkey, soil salinity problem can be mentioned as one of the growing problems in the area. In this study the soil salinity detection of Seyhan plate from years, 2009 to 2010 were analyzed using remote sensing methods. Multitemporal data were acquired from LANDSAT 7-ETM+ satellite in four different dates (19-April-2009, 12-October-2009, 21-March-2010, 31-October-2010). The field electrical conductivity (EC) measuremnts, collected by Landscape Planning Department of Cukurova University during the years 2009 to 2010 were used as a ground-truth data together with the Landsat images. In the introduction part of the study, definition of remote sensing and a brief summary of the topic are given. In the second chapter, as an principles of remote sensing, the electromagnetic spectrum and radiation and electromagnetic interaction with Earth's features, spectral reflectance are explained. In the third chapter, general information about soil salinity and the role of remote sensing in soil salinity detection are provided alongside some literatures related to soil salinity detection using remote sensing. In the fourth chapter, different remote sensing satellites, which are suitable for soil salinity mapping, sensors and their characteristics, are given. In the fifth chapter, digital image is defined and main image processing steps and methods are explained. In the application chapter, the study area and the data used including satellite data and field EC measuremnts are defined. In this study, it is aimed to evaluate the soil salinity level in the study area, using the field EC measurements and remote sensing technology; and to produce the soil salinity map of Seyhan plate of Adana district. With regard to the above objectives, the following processing steps were applied. First, the map projection of all Landsat 7 ETM+ images used, were changed to European 50 in order to be compatible with the projection of field collected data. Then the radiometric correction was done in order to extract the top of atmosphere (TOA) value of each band. Different salinity and vegetation indices were applied to analyze the changes of salinity level in different soil conditions. Then in order to predict the soil salinity, the correlation between field EC measurements and remote sensing data were calculated. Correlation of EC value with DN or TOA value of each band and correlation of EC value with different indices were taken in to consideration in two regression models, the simple linear regression (SLR) and multiple linear regression (MLR). In simple linear regression, the correlation of EC value with DN or TOA of each band of satellite in sampled location was calculated, whereas in multiple regression, the correlation of EC value with DN or TOA value of all bands was computed. In the third approach, the combination of satellite bands and different vegetation and salinity indices were used as independent variables. Since the results of simple linear regression did not yield satisfactory results, the highest correlation (78.40%) was achieved using MLR method. In this correlation the all bands of satellite image dated on 21st March, 2010 and EC value was calculated. Finally the satellite data which shows the highest correlation (21st March, 2010) was chosen for producing the soil salinity map. In the final chapter, the results obtained in the application phase and the efficiency of remote sensing in soil salinity detection and mapping are discussed. Besides, some recommendations for the future research and the problems which encountered during analysis are outlined.