Uydu görüntü verileri ve V-I-S model kullanılarak kentsel ekolojik fonksiyonun mekansal-zamansal analizi
Spatial temporal analysis of urban ecological function by using satellite image data and V-I-S model
- Tez No: 558893
- Danışmanlar: PROF. DR. ŞİNASİ KAYA
- Tez Türü: Yüksek Lisans
- Konular: Jeodezi ve Fotogrametri, Geodesy and Photogrammetry
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 2019
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Geomatik Mühendisliği Ana Bilim Dalı
- Bilim Dalı: Geomatik Mühendisliği Bilim Dalı
- Sayfa Sayısı: 141
Özet
Kentsel mekansal büyümeyi zamanında ve uygun maliyetli bir şekilde analiz etmek için uzaktan algılama verileri giderek daha fazla kullanılmaktadır. Ancak bu görüntülerin işlenmesinde düşük çözünürlüklü verilerin yol açtığı karışık piksel sorunu ve kent dokusunun karmaşıklığı sebebiyle yüzey örtüsünü sınıflandırmada karşılaşılan güçlükler bilim insanlarını bu ve buna benzer sorunları çözmek için yeni modeller geliştirmeye itmiştir. Ridd tarafından 1995 yılında geliştirilen V-I-S (Vegatation-Impervious-Soil) modeli yüzey örtüsünü sınıflandırmada oldukça sadeleştirilmiş bir araç sunmakta ve bu özelliği ile de kentsel ekolojik çalışmalarda çığır açan bir model olarak kabul görmektedir. Bu modelde arazi yüzeyi, bitki örtüsü, su geçirmez alan ve toprak şeklinde üç temel elemanın bir kompozisyonu şeklinde tanımlanmaktadır. Bu sayede kentsel alanların değişimini, gelişimini, arazi kullanımını, yapılaşmayı, kentsel büyümenin yönünü ve yoğunluğunu mekansal ve zamansal olarak analiz edilebilmesini kolaylaşmaktadır, böylece daha tutarlı sonuçlar elde edilmektedir. Uzaktan algılama verilerinin işlenmesinde yaygın olarak kullanılan birçok bilgisayar yazılımında bu verilerin V-I-S modeline uyarlanmasına imkan verecek özellikler bulunmaktadır. Bu çalışmada kentlerdeki mekansal zamansal değişim izlenirken, bitki ve toprak aleyhinde gerçekleşen yapılaşmanın kentsel ekolojik yapı üzerindeki etkisini ortaya koymak hedeflenmektedir. Bu değişimi gözlemleyebilmek için İstanbul'un 39 ilçesinin tamamı uygulama alanı olarak seçilmiş ve V-I-S model esas alınarak görüntü sınıflandırma işlemi yapılmıştır. Bitki örtüsü, su geçirmez alan ve toprak alanlardaki değişimi ortaya koymak için 1989, 1995, 2001 ve 2018 yıllarına ait Landsat Thematic Mapper (TM) ve Landsat Operational Land Imager (OLI) verileri kullanılmıştır. Benzer spektral özellik gösteren piksellerden alınan örneklemeler bitki örtüsü, su geçirmez alan ve toprak sınıflarına atanmıştır. Sınıflandırma sonucunda her sınıfa ait toplam piksel sayıları hesaplanmış, sınıflara ait piksel sayılarının toplam piksel sayısına oranıyla da her ilçe için bitki örtüsü, su geçirmez alan ve toprak yüzdeleri belirlenmiştir. Kentin yıllar içerisindeki vektörel değişimini ortaya koyabilmek ve izleyebilmek için yapılan sınıflandırma işlemi seçilen dört yıl için bütün ilçeler bazında tekrarlanmıştır. Kent morfolojisindeki zamansal ve mekansal değişim V-I-S diyagramları üzerinde vektörel olarak gösterilmiştir. V-I-S diyagramları ile kentin vektörel değişimi incelenirken üzerinde durulması gereken diğer bir konu da kent gelişiminin yönü ve kent gelişim yönüne etkileyen faktörlerin neler olduğudur. Kentsel gelişimin yönü ve büyüklüğünün belirgin olarak ortaya konulabilmesi için daha önce tarım ve hayvancılığın ağırlıkta olduğu, nispeten daha az nüfus yoğunluğu ve yapılaşmaya sahip ilçeler seçilmiştir. Çizilen grafikler yardımıyla bu ilçelerin 1989, 1995, 2001 ve 2018 yıllarına ait kentsel ekolojik fonksiyonun değişimleri ortaya konulmuştur. Doğal yapının hızla tahrip edilmesi ve buna yönelik alınan önlemlerin hızlı kentleşme karşısında yetersiz kalması kent ekolojisi ve sürdürülebilir kent kavramları doğrultusunda mevcut durumun gözden geçirilerek gerekli tedbirlerin süratle alınması ihtiyacını doğurmaktadır.
Özet (Çeviri)
Today, a significant portion of the world's population lives in cities. For this reason, having information about the internal composition and dynamics of urban environments is of great importance in terms of maintaining certain living standards. With the developments in remote sensing technologies, the introduction of multi-band image data with high spatial resolution offers important opportunities for studies in urban ecology. Numerical and objective identification of different surface components allows higher accuracy in urban morphological studies. Remote sensing data is increasingly used to analyze urban spatial growth in a timely and cost-effective manner. However, due to the complex pixel problem caused by low and medium resolution data and the complexity of urban texture in the processing of these images, the difficulties encountered in classifying the surface coverings led scientists to seek new models to solve these and similar problems. The V-I-S (Vegetation-Impervious-Soil) model, developed by Ridd in 1995, provides us with a highly simplified tool for classifying surface coverings and is recognized as an innovative model in urban ecological studies. This model, which defines the structure of the landscape as a composition of three basic elements as green vegetation, Impervious and exposed soil, makes it easier to analyze the spatial and temporal changes of the urban areas, and it is possible to obtain more consistent results. Many computer software that commonly used in the processing of remote sensing data have features that allow this data to be adapted to the V-I-S model. This study aims to analyze the temporal and spatial variation of urban ecological function in Istanbul province based on the V-I-S model developed by Ridd (1995) and using satellite image data. The VIS model recognizes that green vegetation, Impervious surface (artificial area) and bare soil trio are the most important components of the urban ecosystem. This model ignores water surfaces and emphasizes that due to the different nature of these three materials, they have a significant impact on ecosystem dynamics such as energy and moisture distribution. In this study, it is aimed to reveal the effect of urbanization against the vegetation and soil. In order to observe this change, all of the thirty nine districts of Istanbul were selected as the application area and the image classification process was made based on the V-I-S model. Landsat TM and Landsat OLI data for the years 1989, 1995, 2001 and 2018 were used to reveal the changes in vegetation, impervious and soil rate. Pixel based classification methods used in land use/land cover classification are divided into two main groups: supervised classification and unsupervised classification. In the supervised classification method, the data required for the classification is entered into the image processing software by the user. In the uncontrolled classification method, the image is processed by the software and the user evaluates the results in terms of classification. In this study, controlled classification method was used. Samples taken from pixels showing similar spectral characteristics are manually selected and assigned to vegetation, impervious and soil classes. Since the sample fields are selected manually, the classification process is called supervised classification. As a result of the classification, the total number of pixels for each class was calculated and the vegetation, impervious and soil percentages were determined for each district. In order to be able to reveal and observe the vector change of the city over the years, the classification process was repeated on the basis of all districts for four years. The temporal and spatial changes in urban morphology are shown in vector on V-I-S diagrams. It is only possible to determine the accuracy and consistency of the study by performing an accuracy analysis. Accuracy analysis informs us about the consistency and accuracy of the image classification results in the study. The method commonly used to determine the degree of accuracy of the classification is obtained by establishing an error matrix. In this study, positional accuracy was evaluated and standard deviation and kappa accuracy were calculated using error matrices. The obtained values were used to compare the classification results. Three districts have been taken as reference for accuracy analysis. Classified images of these districts belonging to 1989, 1995, 2001 and 2018 were included in the accuracy analysis. Standard deviation and kappa values were calculated and shown in tables.. Two of these districts are Bahçelievler and Güngören districts which have the highest impervious areas and the highest density. Çatalca district that has the most green area has been chosen as the third district. Another issue that needs to be considered when examining the vector change of the city with V-I-S diagrams is the direction of urban development and the factors affecting that. This issue has a complex structure and there are many factors affecting urban development. In order to reveal the direction and severity of urban development, districts with relatively less population density and settlements, which were predominantly agricultural and stockbreeding, were selected. With the help of the graphs drawn and the V-I-S diagrams, the major changes of these districts in 1989, 1995, 2001 and 2018 were revealed. The rapid urbanization process, which started in the 1980s and continues today, has brought many problems. The ecological future of the city is threatened by the rapid destruction of the natural structure and the inadequacy of the measures taken against this rapid urbanization. Within the scope of urban ecology and sustainable urban concepts, it is a vital necessity to review the current situation and take the necessary measures as soon as possible. The numerical information and diagrams we have obtained as a result of this study provide us with very simple and convenient tools for evaluating the satellite image data of this model and monitoring the urban morphological change. In particular, it is possible to achieve faster results by reducing the complex urban surface texture and land use patterns to three main components. The triple V-I-S diagram visually demonstrates the ecological status of the city and facilitates the combination of temporal and spatial analysis with a vector-based representation of the change over the years. During this study, there were some considerable difficulties in the application of the V-I-S model. The results of image classification are adversely affected by similarities and mixed pixels in spectral reflection values of land cover types. Mixed pixels are located at the border of different land cover types. The spectral values of these pixels are seen as a mixture of the values of these two different land cover. Since image classification is made only with the spectral values of the pixels, problems have arisen regarding to which class these areas will be included. Landsat TM and Landsat OLI data were used in the study. A resolution of 30 m reduces the spectral distinguishability between classes. It was seen that this situation caused some problems that were reflected in the classification results. In the application of the V-I-S model, it is revealed that important differences will emerge due to the natural structure of the applied regions, so that vegetation, impervious area and soil definitions may vary from region to region. Salt Lake City where this model was first applied and Istanbul have significant differences in terms of natural structure. Salt lake city is located on the shores of a large salt lake and is surrounded by mountains. In Salt Lake City, it is almost impossible to find green areas that are not created by human hands except for mountains, since salty soil does not naturally allow plants to grow. Istanbul is very different in terms of its natural features. Even the soil areas in Istanbul are covered with green grass for a significant part of the year and it is very difficult to distinguish the soil area and vegetation in low and medium resolution satellite imagery. This difficulty causes the necessity to look at the satellite images and aerial photographs taken in different seasons and and this leads to an increase in working time. As a result of the study, it is seen that the construction which we consider as a impervious area in all districts of Istanbul has greatly expanded against green vegetation and soil in 1989, 1995, 2001 and 2018. It has been shown that this result may adversely affect the distribution of moisture and energy and disrupt the ecological balance and bring many problems in the future. This study was applied separately as a district level and it was concluded that the plans for the future of Istanbul should be made as a whole from a single center. Especially in the old settlements which have become stable, the soil area was decreased dramatically. There is almost no soil area at this region. It is almost impossible to make large-scale ecological plans for this districts. The formation of corridors with the traditional ecological approach will be possible only with the support of the surrounding settlements. This will only be possible with a detailed and long-scale central planning.
Benzer Tezler
- V-I-S modeli ile uzaktan algılama verileri kullanılarak kentleşmenin izlenmesi
Monitoring the urbanization through V-I-S model using remote sensing data
EZGİ TOK
Yüksek Lisans
Türkçe
2006
Jeodezi ve Fotogrametriİstanbul Teknik ÜniversitesiJeodezi ve Fotogrametri Ana Bilim Dalı
YRD. DOÇ. DR. ŞİNASİ KAYA
- Elmalı ve Alibey su havzalarının uydu görüntü verileriyle izlenmesi ve bilgi sistemi oluşturma olanakları
Başlık çevirisi yok
ÇİĞDEM GÖKSEL
- Spektral indekslerin arazi örtüsü/kullanımı sınıflandırmasına etkisi: İstanbul, Beylikdüzü ilçesi, arazi kullanımı değişimi
Effect of spectral indices over land use/cover classification: İstanbul, Beylikduzu district, land use change
ÖZGE KAYMAN
Yüksek Lisans
Türkçe
2015
Jeodezi ve Fotogrametriİstanbul Teknik ÜniversitesiGeomatik Mühendisliği Ana Bilim Dalı
PROF. DR. AYŞE FİLİZ SUNAR
- Evaluating BFAST algorithm in landsat time series analysisof monitoring deforestation dynamics in coniferousand deciduous forests
Landsat zaman serisi ile iğne ve geniş yapraklı ormanlardaormansızlaşma dinamiklerinin izlenmesinde BFASTalgoritmasının değerlendirilmesi
NOOSHIN MASHHADI
Yüksek Lisans
İngilizce
2021
Bilim ve Teknolojiİstanbul Teknik ÜniversitesiCoğrafi Bilgi Teknolojileri Ana Bilim Dalı
Assist. Prof. Dr. UĞUR ALGANCI
- Görüntü sınıflandırması için yapay sinir ağlarının analiz ve optimizasyonu
Analysis and optimization of artificial neural networks for image classification
OZAN ARSLAN
Doktora
Türkçe
2001
Jeodezi ve Fotogrametriİstanbul Teknik ÜniversitesiPROF. DR. OĞUZ MÜFTÜOĞLU
PROF. DR. CANKUT ÖRMECİ