Arazi yüzeyi iklim modellerinin anahtar elemanlarından olup, yüzeydeki enerjinin hissedilebilir ve gizli ısı olarak, yüzeydeki mevcut suyun ise buharlaşma ve akış olarak bölüşülmesini kontrol etmektedir. Ayrıca, arazi yüzeyinin pürüzlülüğü ve yüzeyin cinsine bağlı olarak ısı kapasitesi ve momentum tutulma miktarı değişmektedir. Bu nedenle arazi yüzeyinin doğru ve güvenilir bir şekilde ifade edilmesi iklim çalışmaları için önemlidir. Bölgesel iklim modellerinin çoğunda, global olarak hazırlanmış olan Global Land Cover Characteristics (GLCC) arazi örtüsü verisi kullanılmaktadır. Bu çalışmada, endüstrileşme ve nüfus artışı sonucunda özellikle 1980'li yıllardan sonra arazi örtüsü değişiminin meydana geldiği Marmara Bölgesi çalışma alanı olarak seçilmiştir. Çalışmanın ilk aşamasında, 2001-2005 tarihleri arasında elde edilen Landsat7 ETM+ görüntüleri radyometrik ve atmosferik olarak düzeltilerek atmosferik parçacıklardan kaynaklanan bozulma etkileri ve sistematik hatalar elemine edilmiştir. Geometrik distorsiyonları elemine etmek, piksel bağıl konum hatalarını düzeltmek ve görüntüleri ortak bir koordinat sisteminde tanımlayabilmek amacıyla her bir görüntü geometrik olarak düzeltilmiştir. Görüntüler farklı yöntemler kullanılarak sınıflandırılmış ve çalışma bölgesi için arazi örtüsü haritası oluşturulmuştur. Oluşturulan arazi örtüsü haritası, GLCC verisi ile kıyaslanarak bu verinin Marmara Bölgesi için doğruluğu araştırılmış ve verideki eksiklikler belirlenmiştir. Çalışmanın sonucunda GLCC veri setinin güncel olmadığı ve Marmara Bölgesinin önemli bir kesimini doğru temsil etmediği tespit edilmiştir. Bölgesel iklim modelleme çalışmaları için, bu veriye alternatif olarak daha doğru ve güncel olan Landsat ETM+ görüntülerinden üretilmiş arazi örtüsü verisinin kullanılabileceği gösterilmiştir.Anahtar Kelimeler: Arazi örtüsü, arazi yüzeyi, bölgesel iklim modelleme, uzaktan algılama, Landsat7 ETM+. ; In this research, utilization of Landsat7 ETM+ images in regional climate modeling was investigated and the accuracy of Global Land Cover Characterization (GLCC) data set used in regional climate modeling was assessed for the Marmara Region by comparing these data with Landsat ETM+ derived land cover data. Marmara Region was selected as study area because it faced with significant land cover changes as a result of rapid industrialization and population increase especially after 1980s. The region occupies the northwest corner of Turkey with a surface area of 67 000 km² and represents approximately 8.6% of the Turkish national territory. It is the smallest but most densely populated of the seven geographical regions of Turkey. This region includes eleven cities namely Istanbul, Bursa, Kocaeli, Edirne, Balikesir, Kirklareli, Tekirdag, Canakkale, Bilecik, Sakarya and Yalova, where first three cities are industrial and commercial centers of Turkey. Landsat7 ETM+ images obtained between 2001 and 2005 were used to derive land cover data of the Marmara Region. Since 2005 Landsat ETM+ frame includes the significantly changed areas of the Marmara Region like Istanbul and Bursa, it is assumed that created land cover data is representing the year of 2005. In addition to satellite images, forest maps, 1/25 000 scaled topographic maps, ground surveys and photographs were used to assist geometric correction and classification procedure. At the first stage of the research, all images were atmospherically and radiometrically corrected to minimize contamination effects of atmospheric particles (scattering and absorption effects due to the atmosphere) and systematic errors. Then, geometric correction was performed for each image to eliminate geometric distortions, correct errors in the relative positions of pixels, and define images in a common coordinate system. A new approach, semivariograms, was introduced to select appropriate band combinations for studying different land cover classes. After these corrections, images were classified using different classification methods to identify different land cover types. United Stated Geological Survey (USGS) Land Use and Land Cover Classification Legend was used in the study. Several pilot areas were created and classification employed separately for these areas to minimize the spectral mixing of various classes such as barren, crop and urban and increase the classification accuracy. The classification results were aggregated to 1 km to form final land cover data and classification accuracy assessment was performed on final land cover data. At the second stage of the research, 2005 land cover data obtained from Landsat7 ETM+ images was compared with the GLCC data set to analyze the accuracy of these data for the Marmara Region. These data have been used in many regional climate models like Regional Atmospheric Modeling System (RAMS), the Fifth-Generation NCAR/Penn State Mesoscale Model (MM5) and Weather Research and Forecasting (WRF). It was obtained from 1-km Advanced Very High Resolution Radiometer satellite images spanning April 1992 through March 1993 with an unsupervised classification technique and accuracy assessment of data set was not performed globally. Land surface is a key determinant in climate system and it controls the partitioning of available energy at the surface between sensible and latent heat, the partitioning of available water between evaporation and runoff. Therefore it must be represented accurately and precisely. Land cover products used in most climate models were initially compiled from maps and ground surveys till the global scale land cover products generated from remote sensing images became avaible. These remotely sensed derived global land cover products like GLCC, University of Maryland land cover classification and Global Land Cover 2000 have been implemented into various land surface schemes and climate models. However, no land cover data set is 100% accurate, even if developed from the most advanced satellite images. The results of comparison analyses between Landsat derived land cover and GLCC show that GLCC data is not up-to-date and have deficiencies and inaccuries in some parts of the Marmara Region. GLCC is not representing urban areas accurately in İstanbul, Adapazari, Bursa and İzmit. These data also have problems in coastal part of İstanbul European side and show some forest areas as crop areas. This research results show that land cover data obtained from Landsat ETM+ images can be successfully and accurately represent the study region therefore it is an alternative source of up-to-date and accurate land cover data for regional climate modeling.Keywords: Land cover, land surface, regional climate modeling,remote sensing Landsat7 ETM+.