Parametrik olmayan testler kullanılarak Sakarya havzası yağışlarının trend analizi
Yükleniyor...
Tarih
2004
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Selçuk Üniversitesi Mühendislik-Mimarlık Fakültesi
Erişim Hakkı
info:eu-repo/semantics/openAccess
Özet
Bu çalışmada Türkiye’nin önemli büyük havzalarından olan Sakarya Havzası’na ait aylık toplam yağışlardaki değişimi belirlemek amacıyla trend analizi yapılmıştır. Bu amaçla havzada bulunan 25 adet yağış gözlem istasyonuna ait 1960-2000 periyodundaki aylık toplam yağış verilerine parametrik olmayan Sen’in T, Spearman’ın Rho, Mann-Kendall ve Mevsimsel Mann-Kendall trend testleri uygulanmıştır. Trend analizleri iki bölümden oluşmaktadır. İlk aşamada aylık toplam yağışlar yıllık bazda değerlendirilmiş, ikinci aşamada ise her istasyona ait yağış serilerinin her bir aylık değişimi incelenmiştir. Trendlerin homojenliği de yine parametrik olmayan Van Belle ve Hughes Homojenlik testi ile test edilmiştir. Trendlerin lineer eğimleri (birim zamandaki değişim) Sen tarafından geliştirilen Sen’in Trend Eğim Metodu kullanılarak hesaplanmıştır. Her bir ay için trend başlangıç yılları ise parametrik olmayan Mann-Kendall Mertebe Korelasyon testi ile belirlenmiştir. Sonuç olarak incelenen istasyonların yarısında 0.05 anlamlılık seviyesinde azalan trendler bulunmuştur. 0.05 anlamlılık seviyesine göre toplam 300 ay içerisinde trend tespit edilen aylar dikkate alındığında toplam 44 ayda trend tespit edilmiştir. Bu trendlerin yaklaşık %20’si artan, %80’i ise azalan yöndedir. Artan yönde trendler 6 istasyonla Ekim ve 3 istasyonla Ağustos ayında bulunmuştur. Geriye kalan 35 aydaki trandler ise azalan yöndedir.
In this study, some trend analyses were made to determine the variations of monthly rainfall data belonging to Sakarya Basin which is one of the big and important basin in Turkey. Data records of monthly rainfall from 1960 to 2000 of 25 stations in Sakarya Basin were analyzed using four different non-parametric methods namely Sen’s T, Spearman’s Rho, Mann-Kendall and Seasonal Mann- Kendall to demonstrate the existence of any possible trends. Trend analyses were also divided into two phases. In the first phase, the monthly rainfall data was evaluated annually. In the second phase, variation of rainfall series belonged to 25 stations was evaluated for each month. Homogeneity of trends in monthly rainfall was also tested by using a procedure developed by Van Belle and Hughes. The linear slopes (change per unit time) of trends were calculated by using a trend slope method suggested by Sen. The starting points of trends for each month were defined by using the Mann-Kendall Rank Correlation test. Consequently, downward trends at the 0.05 significance were found at half of the investigated stations. Trends were found at total 44 months with in the months in which some trend exist at the 0.05 significance level in total 300 month. 20% of them trends is upward whereas 80% downward. The upward trends were found in October period at 6 stations and August period at 3 stations. Trends of remaining 35 months were downward.
In this study, some trend analyses were made to determine the variations of monthly rainfall data belonging to Sakarya Basin which is one of the big and important basin in Turkey. Data records of monthly rainfall from 1960 to 2000 of 25 stations in Sakarya Basin were analyzed using four different non-parametric methods namely Sen’s T, Spearman’s Rho, Mann-Kendall and Seasonal Mann- Kendall to demonstrate the existence of any possible trends. Trend analyses were also divided into two phases. In the first phase, the monthly rainfall data was evaluated annually. In the second phase, variation of rainfall series belonged to 25 stations was evaluated for each month. Homogeneity of trends in monthly rainfall was also tested by using a procedure developed by Van Belle and Hughes. The linear slopes (change per unit time) of trends were calculated by using a trend slope method suggested by Sen. The starting points of trends for each month were defined by using the Mann-Kendall Rank Correlation test. Consequently, downward trends at the 0.05 significance were found at half of the investigated stations. Trends were found at total 44 months with in the months in which some trend exist at the 0.05 significance level in total 300 month. 20% of them trends is upward whereas 80% downward. The upward trends were found in October period at 6 stations and August period at 3 stations. Trends of remaining 35 months were downward.
Açıklama
URL: http://sujest.selcuk.edu.tr/sumbtd/article/view/17
Anahtar Kelimeler
Trend, Sen'in T, İklim değişimi, Climate change, Mann-Kendall
Kaynak
Selçuk Üniversitesi Mühendislik-Mimarlık Fakültesi Dergisi
WoS Q Değeri
Scopus Q Değeri
Cilt
19
Sayı
Künye
Büyükyıldız, M., Berktay, A. (2004). Parametrik olmayan testler kullanılarak Sakarya havzası yağışlarının trend analizi. Selçuk Üniversitesi Mühendislik-Mimarlık Fakültesi Dergisi, 19, (2), 23-38.
