Author: Nadire öZEL, Serdar KALAYCI, Mehmet F. SEVİMLİ, Meral BüYüKYILDIZ

Publishing Date: 2004

E-ISSN: 2147-9364

Volume 19 Issue 2


In this study, monthly streamflow data of eleven stations operated by EIE (Electrical Power Resources Survey and Development Administration) located at Sakarya river basin were investigated. Three different non-parametric trend tests (Sens T, the Spearmans Rho and the Seasonal Kendall tests) were selected to determine linear trends of monthly flow data. The linear slopes of trends were calculated by using a non-parametric estimator developed by Sen. The beginning of trends was determined by using the Mann-Kendall rank correlation test. In addition, homogeneities in monthly trends were tested by using a method developed by Van Belle and Hughes. According to the analyses carried out; meaningfull trends are determined (at the 0.05 significance level) in 64% of total 132 month of 11 stations. Generally, the determined changes are in decreasing trend. The decreasing trend is maximum in the stations 1221 and 1243 and they are located in northwest of the basin. Maximum negative slopes are determined in March for all stations except for stations 1216 and 1226. March is the one in which maximum number of trend is determined (10 station) and all of trends are downward. Also March is the month in which maximum slope is observed. April is the month in which least trend is determined (5 station). The beginning of trends was observed in the early 1980s. According to the Van Belle and Hughes Homogeneity test, there is no global trend in the basin.

Key Words: Akım veriler, Mevsimsel Kendall, Sen T, Spearman Rho, Van Belle ve Hughes

Full Text