Ecohydrological zoning based on the Gaussian mixture model clustering method

ABSTRACT Aiming at the spatial heterogeneity of ecohydrological features in Ziya River Basin, this study constructed a regional index system based on entropy weight method, and proposed an innovative probability-density-driven Gaussian mixture model (GMM) clustering method. In this study, the GMM clustering method was used to calculate the posterior probability values for each unit, which utilised the 1 km×1 km grid of the Ziya River Basin as the calculating unit and selected indicators related to the underlying surface characteristics, human activities, and hydrological and ecological features. The weights of each indicator are as follows: elevation 0.263, slope 0.218, precipitation 0.053, temperature 0.048, relative humidity 0.055, runoff 0.224, vegetation normalised index 0.030, and gross domestic product per capita, 0.108. By calculating the posterior probability, the basin is divided into 5 ecohydrological regions. The zoning results reflect the similarities and differences in the ecohydrological situation of the basin. After the Kruskal-Wallis testing and analysis, all eigenvalues were p = 0.000<0.05, indicating that the eight indicators were not the same for each type of distribution. This study provides a minimum management unit for basin water resource management departments, conserves resources, and improves social benefits.