Derelerde iki boyutlu taşkın modellemesi ve taşkın tehlike risk haritalarının oluşturulması = Two-dimensional flood modeling in streams and creation of flood hazard risk damage maps

Abstract

Canlılığın devamı için gereken su, en temel ve en önemli ihtiyaç olduğu için insanlar tarih boyunca suya yakın yerlere yerleşmişlerdir. Ancak suya yakın olmak taşkın tehlikesini de beraberinde getirmektedir. Taşkınlar, ülkemizde ve dünyada can ve mal kayıplarının en yüksek olduğu doğal afetlerden biridir. Son yıllarda iklim değişikliğinin etkisi ile de yaşanma sıklığı ve zararları artmaya başlamıştır. Bu nedenle insan hayatını maddi manevi etkileyen taşkınları anlamak, önceden tahmin etmek, taşkınların sebep olduğu zararları hesaplamak büyük önem taşımaktadır. Bu tez çalışması kapsamında Sakarya Alt Havzasında bulunan Dinsiz Çayı'nın iki boyutlu hidrolik modeller ile taşkın alanlarını belirlemek ve çalışma alanında bulunan Hendek 2. Organize sanayi bölgesi için taşkın zararlarını hesaplamaktır. Bölgede bulunan akım gözlem istasyonundan elde edilen veriler yeterli olmadığı için meteorolojik istasyonlardan alınan veriler ile yağıştan akışa geçiş yapılarak taşkın tekerrür debileri ve taşkın hidrografları elde edilmiştir. Bu hesaplamalar HEC-HMS programı kullanılarak yapılmıştır. İki boyutlu hidrolik hesaplamalar HEC-RAS programı kullanılarak taşkın modelleri oluşturulmuştur. İki boyutlu modellemelerde çalışma alanındaki taşkın yayılımları taşkın tekerrür debileri ile elde edilmiştir. HEC-RAS 2B ile oluşturulan 50, 100, 200 ve 500 yıllık tekerrür debilerinin taşkın modelleri, taşkın yayılım, taşkın su yüksekliği ve taşkın su hızı haritalarını oluşturmak için ArcGIS programına aktarılmıştır. Elde edilen haritalar yardımıyla DEFRA yöntemine göre taşkın tehlike haritaları, Dinh ve FEMA yöntemlerine göre de taşkın risk haritaları üretilmiştir. Taşkınlardan sonra meydana gelen zararları hesaplamak için literatürde sıklıkla kullanılan Pistrika ve Jonkman yöntemi ve Huizinga, van Eck ve Kok, ICBR derinlik hasar eğrilerine ek olarak HAZUS-MH fonksiyonları kullanılmıştır. Bu çalışmanın sonunda Dinsiz Çayı'nda 50, 100, 200 ve 500 yıllık taşkın debileri için taşkın tehlike, risk ve hasar haritaları elde edilmiştir. HAZUS fonksiyonları ile yapılan hesaplamalara göre sonuçlar van Eck ve Kok ve ICBR'den daha yüksek çıkmıştır. Ayrıca van Eck ve Kok ve ICBR sonuçları eşittir. Hesaplamalarda en yüksek sonuçları Pistrika ve Jonkman denklemi verirken, en düşük sonuçları ise van Eck ve Kok ve ICBR eğrisi vermiştir. Su derinliğinin artmasıyla birlikte binaların hasar oranlarında ve buna bağlı maliyetlerde artış gözlemlendi. HAZUS fonksiyonları, hem bina hem de içindekiler için hasar hesaplamalarına yardımcı olur. Sanayi tesisleri, kimya, metal, geri dönüşüm, otomotiv, gıda gibi kullanım amaçlarına göre, yerleşim yerlerindeki binalar ise bodrumlu/bodrumsuz, tek katlı, iki katlı gibi yapısal özelliklerine göre ayrılır ve her birinin farklı hasar değerleri vardır.

Since water is the most basic and important need for the continuation of life, people have settled in places close to water throughout history. However, being close to water also brings the danger of floods. Floods are one of the natural disasters that cause the highest loss of life and property in our country and in the world. In recent years, the frequency of occurrence and its damages have begun to increase due to the effect of climate change. Therefore, it is of great importance to understand and predict floods that affect human life both materially and spiritually, and to calculate the damages caused by floods. The aim of this thesis is to determine the flood areas of the Dinsiz Stream in the Sakarya Sub-Basin with two-dimensional hydraulic models and to calculate the flood damages for the Hendek 2nd Organized Industrial Zone in the study area. The first and most important step of the thesis study is data collection. The high resolution and detail of the data provides great convenience when using it in programs. For this reason, 1/1000 scale point data was obtained from the relevant institution and the study was started. After making the necessary adjustments to these data, the DEM data to be used in the programs was obtained. Statistical methods were used to calculate the flood recurrence rate with the data obtained from the flow observation station in the region. These methods; Gumbel, Normal, Log Pearson, Log Normal and Log Pearson type 3. However, the flow rates obtained from the flow observation station were not used because there was not enough data and the location was far from the stream. For this reason, the rainfall-to-flow transition method was used by obtaining data from rainfall stations in the region. There are 3 stations affecting Dinsiz stream in the study area, their names are Gölyaka, Akyazı and Hendek. The flows from these 3 stations were calculated statistically one by one in the same way. The most appropriate statistical method was chosen as Log Pearson type 3. Thiessen polygon method was used to determine the flow to Dinsiz stream from 3 stations. According to this method, the stations that affect Dinsiz Stream the most are Hendek, Akyazı and Gölyaka, respectively. Since the data obtained from the flow observation station in the region was not sufficient, flood recurrence flows and flood hydrographs were obtained by switching from precipitation to flow with the data obtained from meteorological stations. These calculations were made using the HEC-HMS program. The SCS CN method was used to generate flood hydrographs. Since the study basin did not have a previously calculated CN number, the CN value was calculated manually. Two maps were used to calculate the CN value. These maps are land use maps and hydrological soil group maps. These two maps were opened one under the other in a GIS-based program and calculations were made on an area basis. Cross sections are needed to better understand the study area and process it into the HEC-RAS program. However, since there were no cross sections taken before, these cross sections were taken by an expert and added to the program. Similar results were obtained in two different models that were established by adding both imported cross-sections and automatic cross-sections to the program. However, the models established in this way were not used because they were one-dimensional and not within the scope of the thesis. The flood propagation areas that emerged as a result of the 2D models established with the HEC-RAS program showed that the DEM data used was not sufficient. For this reason, DEM data was downloaded from open access internet channels to enlarge the DEM data used. The two DEMs obtained were combined (embedded) with the help of various GIS programs and a new data was created to be used in 2D modeling. By entering the new DEM data, hydrograph flows and other components into the program, different flood propagation maps were obtained for two-dimensional Q50, Q100, Q200 and Q500. In addition, flood velocity maps were created with HEC-RAS and these maps were transferred to the GIS environment to create flood hazard and risk maps. Flood hazard and flood risk maps are the most important processes to take precautions against floods. In this study, flood hazard maps were obtained by the DEFRA method, and flood risk maps were obtained according to the Dinh and FEMA methods. After the flood hazard and flood risk maps are created, it is time to calculate the damage. In addition to the van Eck and Kok, Huizinga, ICBR, Pistrika and Jonkman methods, which are frequently used in the literature, HAZUS-MH functions were also used in damage calculations. All methods used have some pros and cons compared to each other. While both water velocity and depth are used for damage calculation in the Pistrika and Jonkman methods, the van Eck and Kok, Huizinga and ICBR methods are depth-damage curves that give the depth-dependent damage rate. HAZUS has not been used in flood damage calculations in Turkey before, so it was preferred to use HAZUS as it was a first. It is a tool developed by FEMA and used to calculate the damage caused by natural disasters on a regional scale, although it was originally only for America but is now used almost worldwide. There are many hardware and software factors required to install and use the HAZUS program, and it is quite difficult to adapt the program for Turkey. For this reason, calculations were made manually using the functions used by the program instead of the program. The highest values for each depth were found in the Pistrika and Jonkman curves. This is mainly because Pistrika and Jonkman take into account floodwater velocity as well as floodwater depth. In the Pistrika and Jonkman method, damage analysis is done with the help of an equation, unlike other methods, this equation includes both flood speed and flood depth. Multiplication of these components can be effective at higher results than other methods because in other methods, damage rates are determined only by the water level. According to calculations made with HAZUS functions, the results were higher than van Eck Kok and ICBR. Moreover, the results of van Eck and Kok and ICBR are equal. While the Pistrika and Jonkman equation gave the highest results in the calculations, the van Eck and Kok and ICBR curves gave the lowest results. With the increase in water depth, an increase in the damage rates of buildings and the associated costs was observed. The accuracy rate in the calculation based on these features may be higher than other damage levels. Users can separate each building in the area they work according to its features. In addition, the ability to calculate damage by content for housing and industry can help property owners identify their losses more clearly. At the end of this study, flood hazard, risk and damage maps were obtained for 50, 100, 200 and 500-year flood flows in Dinsiz Stream. According to calculations made with HAZUS functions, the results were higher than van Eck and Kok and ICBR. Also, van Eck and Kok and ICBR results are equal. In the calculations, the Pistrika and Jonkman equation gave the highest results, while the van Eck and Kok and ICBR curve gave the lowest results. With the increase in water depth, an increase in the damage rates of buildings and the associated costs was observed. HAZUS functions help damage account for both the building and its contents. Industrial facilities are divided according to their intended use such as chemical, metal, recycling, automotive, food, and buildings in residential areas are divided according to their structural characteristics such as with or without basement, single-storey, two-storey, and each has different damage values.