In-depth investigation of precipitation-based climate change and cyclic variation in different climatic zones.

Climate change and cyclic variation are investigated based on station data of 61 years (1951-2011), representing twelve climatic zones in Iran. Climate change is investigated by applying the non-parametric Mann-Kendall test and the three-dimensional loglinear model to the12-month SPI time series, and by applying the likelihood ratio test to precipitation time series. Cyclic variation is studied by applying the three-dimensional loglinear model to the 12-month SPI time series. Analysis included entire data period, two sub-periods [(1951-1981), (1982-2011)] and three sub-periods [(1951-1971), (1972-1991), (1992-2011)]. The Mann-Kendall test results indicated combinations of different trend behaviors, whereby climate change could not be evaluated. The likelihood ratio test did not confirm climate change (at 95 % confidence level), in most of the studied stations. However, the more in-depth analysis by the three-dimensional loglinear model, i.e., detection of significant differences among drought frequencies, did not confirm climate change (at 95 % confidence level), in any of the studied stations. Cyclic variation was not confirmed by the three-dimensional loglinear model (at 95 % confidence level), in any of the studied stations. The findings of this research illustrate the need for meticulous techniques like the three-dimensional loglinear model, as a necessary tool for climate change and cyclic variation studies. [ABSTRACT FROM AUTHOR]