The nonlinear hydro-climatic process: A case study of the tarim headwaters, NW China

Some researchers have suggested that the hydro-climatic process is a complex system, with nonlinearity as its basic characteristic. But there is still a lack of effective means available to thoroughly discover the dynamics of hydro-climatic process at different time scales. Therefore, more studies are required to explore the nonlinearity of hydro-climatic process from different perspectives and using different methods. Based on the hydrologic and meteorological data in the areas of the Tarim headwaters, this chapter investigated the nonlinear hydro-climatic process by a comprehensive method including correlation dimension, R/S analysis, wavelet analysis, regression and artificial neural network modeling. The main findings are as follows: (1) The hydro-climatic process in the Tarim headwaters presented periodic, nonlinear, chaotic dynamics, and long-memory characteristics. (2) The correlation dimensions of the attractor derived from the AR time series for the Hotan, Yarkand, Aksu and Kaidu rivers were all greater than 3.0 and non-integral, implying that all four headwaters are dynamic chaotic systems that are sensitive to initial conditions, and that the dynamic modeling of hydro-climatic process requires at least four independent variables. (3) The Hurst exponents indicate that a long-term memory characteristic exists in hydro-climatic process. However, there were some differences observed, with the Aksu, Yarkand and Kaidu rivers demonstrating a persistent trait, and the Hotan River exhibiting an anti-persistent feature. (4) The variation pattern of runoff, temperature and precipitation was scale-dependent with time. Annual runoff (AR), annual average temperature (AAT) and annual precipitation (AP) at five time scales resulted in five variation patterns respectively. (5) The nonlinear variation of runoff is resulted from regional climatic change. The variation periodicity of AR is close with that of AAT and AP. The multiple linear regression (MLR) and back-propagation artificial neural network (BPANN) based on wavelet analysis reveal the correlations between annual runoff (AR) with annual precipitation (AP), annual average temperature (AAT) at different time scales.