Abstract: In this study, a data sample has been established including 13 sorts of water quality index and 400 stations of hypothesis sampling according to the <National Water Quality Standard (NWQS)> for the evaluation model of coastal water quality. By calculating the KMO, sphericity test as well as the correlation matrix of all indices, we found that there was a good correlation among indexes, which indicated that the principal component analysis (PCA) could be applied to extract information. For 13 indexes of water quality, only the first two principal components with eigenvalues ??greater than 1 are effective, which could represent 81.25% information of the samples, as a result, a quality automatic classification layout (evaluation model) which could accurately classify four kinds of waters has been set up. A thematic map of water quality classification in the coastal waters of Leizhou Peninsula in 2010 was drawn according to the water quality automatic evaluation method. In this map, case IV waters were in the Zhanjiang harbor, Jianjiang Estuary and Tieshan harbor sea area, and case III waters were mainly in the southwest of Donghai Island, offshore area of Jianjiang estuary, northeast of Xuwen, Liusha Bay and Jianghong harbor sea areas, and on the other sea area of Leizhou Peninsula, they were case I and II waters in 2010. This study provides a better distribution in water quality environment of the coastal waters of Leizhou Peninsula, which offers some references for the comprehensive treatment and utilization in this area.

Key words: Leizhou Peninsula, coastal water quality, automatic evaluation model, principal component analysis