Mr. ZHANG Xiaolang Department of Earth Sciences, HKU
The Badain Jaran desert (BJD) occupies parts of Alxa, China covering a big area of 49,000 km2 making it the second largest desert of China. It is attributed the tallest stationary dunes in our planet. Despite hyper arid conditions, another remarkable feature of BJD is the existence of nearly 140 permanent lakes that lies in the inter-dune depression. These mysterious lakes are believed to be fed by groundwater. Due to scarce meteorological and hydrogeological data in the desert, various hypotheses have been proposed on the recharge sources of lake water and groundwater.
This study focuses on the relationship between fractal behavior of lake areas and lake watershed indices. The time it takes for precipitation to travel through a lake watershed and reach the lake is a fundamental hydrological parameter. Surface water is well connected with surrounding groundwater and often an expression of groundwater. The timescales of watershed hydrologic response can be clarified using spectral methods, which decompose lake area variation signals into their component frequencies. The variations of lake areas that are extracted from Landsat images (from 1986-2017) exhibit fractal scaling 1/f^ over three orders of magnitude. The scaling exponent β of 1/f^β time series of lake area variations show close relationships with lake watershed topographic indices (e.g. watershed area, height difference). It suggests that local precipitation can recharges lake water after damping process of sand dunes, and hence local precipitation instead of remote source or paleo-precipitation is the main recharge of groundwater.