Growth diffusion-limited aggregation for basin fractal river network evolution model

Abstract

Diffusion-limited aggregation (DLA) model has been widely used to simulate fractal aggregation processes. This DLA model and its improved model growth diffusion-limited aggregation (GDLA) can reasonably plan and manage rivers within the basin scale. In this paper, the growth and evolution of the river network at the basin scale were simulated using the Northwest China Basin as the study area. To achieve this research goal, this paper used gas injection system image processing technology and data analysis methods for overall processing and distributed processing of river basin remote sensing data. A new growth method, unit-gradient growth, was added to the standard DLA model, and a combination of the extracted Daxia River Basin geomorphological features and DEM data was used to optimize the Brownian motion of the model particle growth method. A computational model was proposed: GDLA model to predict the river network growth in the study area. The fractal dimension of the river network in the study area was extracted and corrected by the box-counting method. Finally, the predicted river network was compared with the extracted actual river network. The included angle cosine method was used for modeling evaluation and analysis. The results showed that the fractal dimension of the Daxia River Basin and sub-basin water systems was less than 1.6, indicating that the basin geomorphology was in its infancy. The structural similarity degrees were greater than 0.99, indicating that the GDLA model had a better simulation on fractal river networks, which can successfully reproduce and predict different morphological two-dimensional river networks. Studies showed that using GDLA to model fractal river networks can better predict the river network growth and evolution, ultimately promoting sustainable development of river basin water management.