This paper applies a procedure that identifies trends in hydrologic variables. The procedure utilizes the regional Mann–Kendall non-parametric test with and without both serial and cross-correlation to detect trends. The research investigates 15 streamflow variables including annual minimum, mean, maximum and monthly streamflows for a network of 75 streamflow gauging stations in seven geographical regions of Turkey. A considerable difference was obtained in the assessment of results with and without consideration of serial and cross correlation which might be due to a higher number of serial and cross-correlations among the sites in the geographical regions. Therefore, a quite different interpretation of these trend analyses would have been achieved if the temporal and spatial correlation of the streamflow series within the regions had been ignored. The application of the regional trend detection technique with both considerations has also resulted in the identification of significant decreasing trends in the Marmara, Aegean, Mediterranean and Central Anatolia regions. However, almost no evidence of significant change was observed with a general downward direction in the rest of the country. Besides, there are differences in the geographical regions of significant trends in the fifteen streamflow variables considered which implies that impacts on streamflows are not spatially uniform.
Research Article|April 01 2006
Regional trend detection of Turkish river flows
Fatih Topaloğlu; Regional trend detection of Turkish river flows. Hydrology Research 1 April 2006; 37 (2): 165–182. doi: https://doi.org/10.2166/nh.2006.0013
Download citation file: