Karun (iran) Üst Havzası’nda Taşkın Analizi

Abstract

Doğal afetler arasında insanlığa en fazla zarar veren doğa olaylarından biri taşkındır. Bu nedenle çok eski dönemlerden beri insanlar can ve mallarını korumak için çeşitli taşkın önlemleri almaktadırlar Bu çalışmada, Karun (İRAN) üst Havzası’nda 14 akım ölçme istasyonlarının debi değerlerini kullanarak taşkın frekans analizi yapılmıştır. Akımların kayıt süresi 18-36 yıl olan bu istasyonların her yılın anlık pik debilerine çeşitli yöntemler uygulanmıştır. İstasyonların verilerine L-moment, Kolmogorov-Smirnov ve Olasılık Çizgisi Korelasyon katsayısı (PPCC) testini uygulayarak verilere en uygun dağılım bulunmuştur. Her istasyon için 50, 100, 200 ve 500 yıllık dönüş aralıklarda taşkın değerini tahmin etmek için iki ve üç parametreli dağılımlar kullanılmıştır. Normal, Log- Normal, üç parametreli Log-Normal, Gumbel, GEV, Pearson Tip III ve Log-Pearson Tip III dağılımları kullanılan dağılımlarıdır. Verilerin zaman değişimi ile artış ya azalışlarını araştırmak için Mann-Kendall trend analizi testi yapılmıştır. Her istasyon için debileri, debilerin ortalamasına bölerek boyutsuz grafikleri çizilmiştir. Bu grafiklere doğrusal gidiş çizgisi (Trend) uydurarak görsel açıdan trendin olup olmadığı kontrol edilmiştir. Karun üst Havzasını homojen bölgelere ayırarak bölgesel analizi yapılmıştır. Bu çalışmada homojen bölgeleri belirlemek için Wiltshire yöntemi kullanılmıştır. Açısal mevsimsellik analizi ve göreceli frekans analiz yöntemleri ile Mevsimsellik Analizi yapılmıştır. İstasyonlarda gözlenen tüm taşkın tarihleri, ay bazında incelenmiş olup taşkınların gözlendiği tüm ayların frekans değerleri incelenirken, aylık dönemler, 30 günlük peryotlara dönüştürerek ayarlanmış frekans değeri hesaplanmıştır. Bu çalışmanın sonucu olarak, Normal dağılım çeşitli dönüş aralıkları için en düşük debi değeri tahmin etmiştir. Ayrıca, en yüksek debi değerleri LN ve LP3 dağılımları ile tahmin edilmiştir. Bu havzanın taşkın değerleri için L-moment testi GEV dağılımını, en uygun dağılım olarak belirlemiştir. PPCC ve K-S testleri Karun üst Havzası’na uygulayarak en uygun dağılım, LN3 ve LP3 olarak bulunmuştur. Yapılan Mann-Kendall trend analizinde 0,05 anlamlılık düzeyinde hiç istasyonda taşkın debi değerlerinde artış ve azalış görülmemiştir. Karun üst Havzası’nı 3 homojen bölgeye bölerek bölgesel analiz yapılmıştır ve her bölgenin kendi içinde homojen olduğu gösterilmiştir. Taşkın mevsimselliği, mevsimsellik ölçütleri kullanarak gözlenmiş taşkınların yılın 43. ile 83. günleri arasında değiştiği saptanmıştır.

Flooding is when the water level in a creek, river, lake or the sea rises and covers land that is usually dry. Whilst some floods occur without problem, others are devastating, causing large-scale destruction and significant loss of life. Flooding is experienced all over the world and in some countries flooding occurs regularly. Floods are the natural disasters that ruin humanity. For this reason, since very old times people have used certain flood precautions to protect their property. The aim of flood frquency analysis is to predict the flood discharge for a known period in a flood event for planning and desiening of reservior, river basins, bridge, dams, weirs, channels, water supply systems, culverts, etc. Frequency analysis is a technique of fitting a probability distribution to a series of observations for defining the probabilities of future occurrences of some events of interest, e.g., an estimate of a flood magnitude corresponding to a chosen risk of failure. The use of this technique has played an important role in engineering practice. The assumptions of independence and stationarity are necessary conditions to proceed with such analyses. The 18th largest country in the world in terms of area at 1,648,195 km2 , Iran has a population of around 78 million. It is a country of articular geopolitical significance owing to its location in the Middle East and central Eurasia. Iran is bordered on the north by Armenia, Azerbaijan and Turkmenistan. As Iran is alittoral state of the Caspian Sea. Iran is bordered on the east by Afghanistan and Pakistan, on the south by the Persian Gulf and the Gulf of Oman, on the west by Iraq and on the northwest by Turkey. Tehran is the capital, the country s largest city and the political, cultural, commercial and industrial center of the nation. In this study, flood frequency analysis has been observed in the upper Karun (IRAN) Basin and annual maximum discharge specification of 14 station from the basin have been studied. The largest river by discharge in Iran, the Karun River s watershed covers 65,230 square kilometres in parts of two Iranian provinces. The river is around 950 kilometres long and has an average discharge of 575 cubic metres per second. There are a number of dams on the Karun River, mainly built to generatehydroelectric power and provide flood control. Various methods has been applied to snapshot peak discharge values of this 18-36 year recording time stations. L-moment, Kolmogorov-Smirnov and Probability Plot Correlation Coefficient (PPCC) tests have been applied to each station’s data and found the most appropriate distribution. Estimation the value of the flood in each station for 50, 100, 200 and 500-year return intervals with two and three parameter distributions. The probability distributions are classified according to the number of parameters into two types. 2-parameters probability distributions which include: Normal (N) Log-Normal (LN2) Gumbel (EVI) 3-parameters probability distributions which include: Log-Normal (LN3) Generalized Extreme Value (GEV) Pearson type III(P3) Log-Pearson type III (LP3) The parameters of flood flows were estimated for each station. Tnese parameters are the mean x ̅, standard deviation Sx, coefficient of variation Cvx, .coefficient of skew Csx and coefficient of kurtosis ks. using probability weighted moments (PWMs) L- moments are easily computed, L-moment ratios are defıned as L-coefficient of variation, L- skewness and L- kturtosis. Various methods were used for estimation of the parameters of probability distributions. The parameters of the N distribution are the mean x ̅ aıd standard deviation Sx. Nonnormal distributed variables can be adjusted to the normal distribution by means of a suitable distribution. One of these transformation methods is computing the logarithms (y=lnx), In this case logarithmic mean y ̅ and standard deviation Sy, will be the parameters of the LN2 distribution. For the EVl distribution, α, u scale and location parameters were estimated by PWMs and L-moments. Just as the log-normal distribution represents the normal distribution of the logarithms of the variable x, so the 3-parameter log-normal distribution represents the normal distibution of the logarithms of the variable (x-xo) where xo is the third parameter corresponding to a lower boundary. The parameters of the GEV distribution is estimated by PWMs and L-moments, α, u and k are the scale, location and shape parameters, respectively. The parameters of the P3 and LP3 distributions can be estimated by the method of moments. Trend is a change (decrease/increase) of the values of a random variable. It is very important to determine the trend of the amount of water in the rivers in different periods of time for suitable planning and management of the water resources. There are different works to determine this change. For the determination of streamflow trends, parametric and nonparametric tests have been used. If data fit to normal distribution, parametric tests give good results. Nonparametric tests are independent of distribution and parameters of a random variable. These tests are related to the ranks in the arranged sample of the data. Generally, the distributions are not normal. So the use of nonparametric tests give good results. Investigatigation the increasement or decreasement of data exchange in time series, Mann-Kendall trend analysis test was conducted. For each station data, plotted non- dimensional graphic with divided by average of data. In addition to this, keeping up with the linear Trend line graphs that visually checked regardless the trend. Annual peak flow estimates of given return periods, T, are commonly used for design and economic evaluation of various water resources projects. Design quantiles, if it is possible, are often supplied from the at-site flood frequency analyses of the peak flow data available at the hydrometric station(s) as close as the project site. Recently, various regional flood frequeny analysis models for the estimation of design quantiles have been developed. The two major objectives of the regional analysis are to estimate the T-year event magnitude at a given location, mainly at ungaged sites, in a statistically homogeneous region, and to improve at-site estimates with large sampling errors because of the limited data. Regional flood frequency analysis for the upper Karun basin with divided into the homogeneous sub basin areas. In this study, has been used the Wiltshire method to determine the homogeneous regions. Seasonality Analysis has been conducted with Angular Seasonality Analysis and Relative Frequency Analysis methods. All the flood date per date observed at the stations have been studied monthly basis and during the study of frequency value of all the months adjusted frequency values have been calculated by converting the duration of each month into 30-day periods. As a consequence of this study, Normal distribution has been estimated minimum value to different return intervals. Furthermore, maximum values to different return interval, have been estimated by LN and LP3 distributions. In the L-moment diagrams the GEV distribution was acceptable to the 50% of the stations, the Pearson type III distribution was acceptable to the 36% of the stations and LN3 distribution was acceptable to the 14% of the stations. L-moment test has been identified GEV distribution, as the most appropriate distribution for the flood values of this basin. As a result of K-S test, it is seen that LN3 distribution, is the only one that accepted at all stations, and has the best-fit at 14 stations. N is rejected at 7 stations, LN2was rejected at 3 stations, Gumbel and GEV at 2 stations. All the distributions were acceptable at stations 21-233, 21-237 and 21-429. The K-S test could not be applied to the data at all stations for the P3 and LP3 distributions. By applying PPCC tests to the upper Karun basin, LP3 and LN3 have been the most appropriate distributions. LP3 and LN3 distributions accepted at the all stations. N and Generalized Extreme Value(GEV) distributions rejected at 9 stations, P3 distribution at 5 stations, LN2 at 2 stations. All the distributions were acceptable at stations 21-225 and 21-233. As a result of Mann-Kendall trend analysis at all stations flood flow values, not increased and decreased in the time series at the 0.05 significant level. Regional analysis was conducted by dividing the upper Karun basin to the 3 homogeneous region and shown to be homogeneous within each area. By using Flood Seasonality, seasonality measures, flooding varies observed between 43. and 83. days of a year