Nonlinear Chaotic and Trend Analyses of Water Level at Urmia Lake, Iran

Abstract: Urmia Lake is the largest habitat of Artemia Urmiana in the world located in an endorheic basin in northwest Iran. This shallow terminal lake has been designated as a Biosphere Reserve by UNESCO and a National Park under the 1971 Ramsar Convention. Currently, the entire lake’s watershed is threatened due to its abrupt desiccation of more than 50% of its surface area and the concomitant increase of salinity during the last two decades. As the rapidly declining eco-environmental conditions have serious impacts on the socio-economy of the whole region, the drying of the lake is considered as a national threat. In light of the above, within the present study a holistic hydroclimatic analysis of the lake is conducted. The principal hypothesis of this research is that ‘during the recent years (1996-2012), there has been an intrinsic change in the dynamic of the lake water level mainly due to human activities which has had deteriorating impacts on the lake and its entire basin’. Therefore, water level variations, as the eco-hydrological indicator of this basin, are compared for the two periods of natural and anthropogenically intervened i.e. 1966-1995 and 1996-2012, respectively. The result of Mann-Kendall trend test, which has been used to identify the changes, shows that there is a unique decreasing trend of about -0.25 m/yr in average during the recent period of 1996-2012. Further, the nonlinear chaotic analysis of Correlation Dimension Method is employed in order to investigate the underlying dynamic of water level oscillations. The correlation exponent has decreased from 3 during the period of 1966-1995 to 2 during 1966-2012 implying that 1 governing variable of the lake’s dynamic is missing! Pursuing the major cause of the recent desiccations, a multiple linear regression model is made based on the standardized monthly anomalies of the lake water level during 1966-1995 and the influencing teleconnection indices as predictors. For the months March and April as high season months and December as a low season month, water level variations have been modeled. Comparing modeled and observation time series shows that climate variability could explain about 35-40% of water level variations in the 1st period. For the 2nd period, nonetheless, only 11-14% of the variations in the high season (i.e. March and April) and 35% in December (low season) could be explained by climate variability i.e. natural causes. In view of the above and alluding to fact that extensive human activities in terms of climate change, land use changes, groundwater withdrawals, causeway, dam and other water diversion projects have occurred or have been intensified during the 2nd period, it could inductively be concluded that human activities are the major cause for the recent depletion of the lake. This result has been confirmed by previous paleolimnological studies as well. Afterwards, the salinity of the lake is modeled by means of regressions and the current salinity level of the lake is obtained to be more than 360 g/l. Eventual ecological consequences of the ongoing deteriorating conditions on the whole basin are discussed. Finally, after the critical review of previously proposed solutions, the project is closed by suggesting plausible possibilities for the restoration of the lake and its watershed towards a sustainable eco-socio-economic state, in terms of short- and long-term measurements.