Abstract: Eco-hydrological regionalization plays an important role in relaxing the contradiction between water resources development and eco-environment protection. Based on the theory of SOFM (self-organizing feature map) artificial neural network, this paper established an eco-hydrological regionalization model. Several integrated indices containing most original information were extracted from a large number of eco-hydrological indices by using principal component analysis (PCA), and on this basis, a cluster genealogy chart was obtained by hierarchical cluster analysis (HCA). Then, the SOFM artificial neural network was constructed to make a rational eco-hydrological regionalization. Taking Quanzhou City as an example, its eco-hydrology was regionalized. The city was regionalized into four eco-hydrological zones, and each of them had its distinct characteristics. Aiming at the different characteristics of each eco-hydrological zone, four strategic options (general, enhancive, strict, and strictest) were proposed to protect the eco-environment and develop the water resources.

Key words: Medicinal woody plants Ecological protection, Environmental regulation