Abstract:In order to reveal the characteristics of confluence network of different micro-topography types, field artificial rainfall simulation experiments were conducted. The confluence networks of seven micro-topography types were abstracted and analyzed by close-range photogrammetry. The rainfall events with 60, 90 and 120 mm/h rainfall intensity lasted for 90 minutes. The seven micro-topography types were continuous depressions, alternate depressions, continuous mounds, alternate mounds, continuous mounds and depressions, alternate mounds and depressions, and smooth slope. The results showed that:(1) The drainage density gradually decreased with the increasing of drainage area threshold. The optimal thresholds for alternate depressions and continuous mounds and depressions were 450 m2, those for smooth slope, continuous depressions and alternate mounds were 550 m2, and those for continuous mounds and alternate mounds and depressions were 650 m2. (2) The morphology of confluence network of the control group, continuous depressions and alternate depressions was dendritic pattern. The confluence network of alternate mounds and alternate mounds and depressions was mixed pattern, which was a mixture of dendritic pattern and parallel pattern. The confluence network of continuous mounds and continuous mounds and depressions was parallel pattern. In the micro-topography, depression played a role in collecting runoff and reducing the amount of low-level runoff, while the mound played a role in changing runoff direction and reducing convergence possibility of runoff. (3) The drainage density and stream frequency of the seven micro-topography types followed the order of dendritic pattern>mixed pattern>parallel pattern. With the rainfall intensity increased, the stream frequency generally increased while the drainage density changed only a little.