Multiple Risk Trend Identification with Seesaw Methodology in Hydrometeorology Time Series

In this paper, trends are determined by population characteristics through a set of the risk levels (exceedence probabilities) from CDFs. This is a new method in which unique trend identification is possible at any risk level, not just the average. The entire approach is based on the cumulative distribution functions (CDF) risk levels for time series records and their two equal halves. At a given risk level the corresponding data amounts represent weights at two ends of a seesaw and the whole data amount the support. In this way, also extreme values such as low (dry, drought) and high (wet, flood) trend possibilities can be defined without any restrictive assumptions. The application of the methodology is given for annual discharge measurements of Danube River, in addition to the annual precipitation records from Istanbul meteorology station.