水系分形特征对流域径流特性的影响--以赣江中上游流域为例

INFLUENCE OF RIVER NETWORK FRACTAL CHARACTERISTIC ON RUNOFF——CASE STUDY IN THE BASIN OF UPPER AND MIDDLE COURSES OF GAN RIVER

以赣江流域中上游的8个水文站点所控制的流域为研究对象,以水系分形维数为指标来描述流域水系特征,以Hurst指数为指标来描述流域径流序列复杂性,研究两者之间的关系,并分析流域水系特征、水系变迁对流域旱涝灾害的影响。研究表明,流域水系分维数越高,水系越复杂,其径流深序列Hurst指数则越高,其径流过程则越简单,反之亦然。径流序列的复杂程度在一定程度上可以反映出洪涝干旱灾害发生的可能性,径流越复杂,则洪涝干旱灾害发生可能性越高。因此,一般来说水系分维越高,水系越复杂,则洪涝干旱灾害发生的可能性也就越低。各流域水系分维与径流深序列的关系,可以在一定程度上说明水系变迁对洪涝干旱灾害的影响:在气温、降水、地形、植被等要素变化不大的情况下,水系、河网消亡将增加洪涝灾害和干旱灾害发生的可能性。

The influence of river network fractal characteristic on runoff was studied. The study-areas are eight basins controlled by eight hydro-stations respectively, located in the upper and middle courses of Gan River. The river network characteristic is described by the fractal dimension, and the characteristic of run-off series is described by the Hurst exponent. Then the relationship between the river fractal dimension and the Hurst exponent was analyzed. And the influence of river network characteristic and the river network changes on the flood/drought are analyzed. The result shows that the higher the river network fractal dimension is, the more complex the river network will be, which is represented by higher Hurst exponent of runoff depth, so the runoff process is more simple. The complexity of runoff series can reflect the probability of flood/drought happening. Generally speaking, the higher the river network fractal dimension is,the lower the probability of flood/drought happening will be. The relationship between the river network fractal dimensions and the Hurst exponents of runoff depth indicates that when physical geography conditions changes are not obvious, the degeneration of river network can result in higher probability of flood/ drought happening.